USF Libraries
USF Digital Collections

The relationship between technology integration and achievement using multi-level modeling

MISSING IMAGE

Material Information

Title:
The relationship between technology integration and achievement using multi-level modeling
Physical Description:
Book
Language:
English
Creator:
Hohlfeld, Tina N
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
Florida
Longitudinal research
Public schools
Secondary data
Student outcomes
Dissertations, Academic -- Instructional Technology -- Doctoral -- USF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: The purpose of this longitudinal study was to examine the relationship between technology integration indicators and school level achievement. Four years of school level secondary data from publicly available databases maintained by the Florida Department of Education were combined for all public elementary, middle, and high schools in the state. This study examined approximately 2300 schools that participated each year in the Florida Innovates Survey about technology integration between 2003-04 and 2006-07. Complexity theory supported the use of multi-level modeling to examine the relationships between technology integration and outcomes. Three achievement outcomes (reading, mathematics, and writing) and two mediating behavioral outcomes (attendance and misconduct) were investigated. Moderating variables controlled in the model included school level, demographics, and learning environment.^ After data preparation, all composite variables were developed using factor analysis. Models were progressively built with significant variables at each level retained in subsequent levels of the study. A total of 94 models were estimated with maximum likelihood estimation using SAS 9.1.3 statistical software. The integration of technology is only one of the many factors that impact student learning within the classroom environment. Results supported previous research about the relationship between the moderating variables and school level achievement and confirmed the need to include moderating variables in the model. After controlling for all the other moderating variables, technology integration had a significant relationship with mean school achievement.^ Although the percent of teachers who regularly use technology for administrative purposes was consistently significant in the models for four out of five outcomes studied, the interactions with time, time2, and time3, resulted in curvilinear trends with inconsistent results. These inconsistent significant findings make drawing conclusions about the integration of technology within Florida's public schools difficult. Furthermore, the small changes observed in mean school achievement over the span of this study support the concept that time is a critical factor for school level learning and change. Therefore, continued analyses of the longitudinal trends for Florida schools in the relationship between technology integration variables and school achievement, while controlling for moderating variables, are recommended.
Thesis:
Dissertation (Ph.D.)--University of South Florida, 2007.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Tina N. Hohlfeld.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 433 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001970779
oclc - 276841333
usfldc doi - E14-SFE0002333
usfldc handle - e14.2333
System ID:
SFS0026651:00001


This item is only available as the following downloads:


Full Text

PAGE 1

The Relationship Between Technology Integration And Achievement Using Multi-Level Modeling by Tina N. Hohlfeld A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Secondary Education College of Education University of South Florida Major Professor: Ann E. Barron, Ed.D. James A. White, Ph.D. Jeffrey D. Kromrey, Ph.D. Elizabeth Shaunessy, Ph.D. Date of Approval: December 12, 2007 Keywords: Florida, longitudinal research, public schools, secondary data, student outcomes Copyright 2008 Tina N. Hohlfeld

PAGE 2

Dedication To my parents, who have always encouraged me to be a lifelong learner by their example and with their support. Thank you for always encouraging me to pursue my dreams. To my children, who encouraged me to finish this Ph.D. program and dissertation, because it is personally meaningful to me. I love you. To my friends, who provided the emotional support that kept me focused even when obtaining the goal was difficult. I am truly blessed.

PAGE 3

Acknowledgements This dissertation is the culmination of many learning experiences and supports provided by very special people at the University of South Florida. Dr. Ann E. Barron, my major professor, provided both personal and professional guidance. Dr. Barron has the keenest insight into finding the most advantageous and expeditious path to accomplishing tasks or solving problems. She also has the knack for providing that just-in-time encouragement that makes all goals possible. Dr. James A. White offered positive encouragement, professional guidance, and feedback. Dr. White always promotes the thor ough examination of all perspectives and possibilities for every research qu estion. He consistently provides that right mix of challenge and assistance that supports each learner’s deepest understanding. Dr. Jeffrey D. Kromrey inspired me with his joy in conduc ting research. Dr. Kromrey is the most patient teacher that I have ever observed both as a student and as a teacher. He ma kes all students believe that they are capable of accomplishing important work. Dr. Kromrey always provid es the support and assistance needed to solve every research question or SAS programming problem. Indeed, without Dr. Kromrey’s guidance the research questions in this dissertation and my other research projects would never have been answered. Dr. Elizabeth Shaunessy supported and encouraged my desire to learn how technology can be used to empower and to support the learning of all students. In particular, she models the best practices for promoting and supporting a productive on-line learning community that stimulates high levels of inquiry through student discussions. Other people at USF also need special recognition. Dr. John M. Ferron and Dr. Robert F. Dedrick introduced me to multi-level modeling statisti cal analysis for educati onal research as well as provided on-going support as I applied this method during my dissertation. All of the students, staff, and faculty in the College of Education with whom I have had the opportunity to interact and learn have made my experiences at USF enjoyable an d worthwhile. Last, Dr. Kate Kemker, Bureau Chief of Instruction and Innovation at the Florida Department of Education, provided the vision and resources to conduct this research with Florida data.

PAGE 4

i Table of Contents List of Tables................................................................................................................. .................................iv List of Figures................................................................................................................ ................................vi Abstract ...................................................................................................................... ...................................ix Chapter 1: In troduc tion........................................................................................................ ...........................1 Background In formation......................................................................................................... ..........2 Theoretical Framework.......................................................................................................... ...........2 Technology Integration......................................................................................................... ............4 Essential Organizational Conditions for Technology Integration.......................................4 School Level Factors that Imp act Technology In tegration.................................................5 Uses of Technology............................................................................................................6 Instruction delivery...............................................................................................7 Product creation....................................................................................................7 Aassessmen t.........................................................................................................8 Student Achievement............................................................................................................ ............9 School Level Factors that Impact Achievement.................................................................9 Teacher ................................................................................................................9 Classroom.............................................................................................................9 Student demographics........................................................................................10 Research Questions............................................................................................................. ............11 Research Plan.................................................................................................................. .11 Variables...................................................................................................................... .....12 Definitions.................................................................................................................... ....14 Technology.........................................................................................................14 Technology integration.......................................................................................14 Socio-economic status (SES)..............................................................................15 Teacher qualif ications........................................................................................15 Positive student learni ng environment................................................................15 Content software.................................................................................................15 Tool-based software...........................................................................................15 Office/ producti on software................................................................................15 Advanced produc tion software...........................................................................15 Florida Comprehensive Achi evement Test (FCAT)...........................................15 Florida School Indicat ors Report (FSIR)............................................................16 Average Yearly Pr ogress Repo rts.......................................................................16 Digital divide......................................................................................................16 Delimitations.................................................................................................................. .................16 Limitations.................................................................................................................... ..................17 Educational Significance....................................................................................................... ..........18 Chapter 2: Lite rature Review................................................................................................... .....................19 Theoretical Frameworks......................................................................................................... ........19 Carroll Model of School Learni ng and Student Achievement..........................................19 Complexity Theory and Organizational Change..............................................................26 Changes in School Organi zation and Instruction............................................................................28 Accountability an d Standards...........................................................................................28 School Reform.................................................................................................................. 29 Essential Conditions for Integra tion of Technology Initiatives......................................................30 District and School Level Factors.....................................................................................31 Communication..................................................................................................31

PAGE 5

ii Leadership and vision.........................................................................................31 School-level differences.....................................................................................32 Classroom Leve l Factors..................................................................................................33 Teacher ..............................................................................................................33 Professional development...................................................................................34 Access to computers...........................................................................................36 Access to software..............................................................................................37 Technical and technology integration support....................................................37 Student-Level Factors.......................................................................................................38 Socio-economic status........................................................................................38 Research Evidence for the Relationship of Technology and St udent Outcomes............................40 Student Achievem ent Outcomes.......................................................................................40 Meta-analy sis......................................................................................................41 Research sy nthesis..............................................................................................50 Large-scale longitudi nal research.......................................................................53 Multi-level model research.................................................................................56 Summary of research on student achievement...................................................61 Student Behavioral Outcomes..........................................................................................63 Summary........................................................................................................................ .................66 Chapter 3: Methods............................................................................................................. ..........................67 Data sources................................................................................................................... .................67 Master School Identifi cation (MSID) files.......................................................................67 Instrument: Florida Comprehe nsive Assessment Test......................................................68 Instrument: Florida School Indicators Repor t (FSIR).......................................................69 Instrument: Average Year ly Progress Reports..................................................................69 Instrument: System for Technology Acco untability and Rigor (STAR) Surveys.............69 Predictor Variables............................................................................................................ ..............70 School Level Predictors at the Schoo l Level....................................................................70 Elementary..........................................................................................................70 Middle ..............................................................................................................70 High ..............................................................................................................70 Demographic Predictors at the Schoo l Level....................................................................70 Free or reduced lunch status students.................................................................71 Minority ..............................................................................................................71 Limited English Prof iciency ( LEP)....................................................................71 Student with disabilitie s.....................................................................................71 Gifted ..............................................................................................................72 Learning Environment Predicto rs at the Sc hool Level.....................................................72 Positive student learni ng environment................................................................72 Absence ..............................................................................................................72 Stability rate.......................................................................................................72 Student conduct..................................................................................................73 Teacher qualif ications........................................................................................73 Technology Integra tion Measures.....................................................................................74 Student access to software..................................................................................74 Percent of teachers who regula rly use computer technology..............................74 Frequency that students use software.................................................................75 Support for t echnology.......................................................................................75 Outcome Measures............................................................................................................... ...........76 Student Achievement........................................................................................................76 Reading achi evement..........................................................................................76 Mathematics ach ievement..................................................................................76 Writing achiev ement..........................................................................................77 Mediating Behavior al Outcomes......................................................................................77 Absence ..............................................................................................................77 Student misconduct............................................................................................77

PAGE 6

iii Data Preparation Procedures.................................................................................................... .......77 Merging Data Files...........................................................................................................77 Exploratory Fact or Analysis.............................................................................................78 Sample ........................................................................................................................ ....78 Descriptive Statistics fo r Outcome Variables...................................................................79 Descriptive Statistics for Predictor Variables...................................................................81 Correlations of Technology Indicat ors with Predictor Variables......................................81 Data Analys is Plan............................................................................................................. .............81 Multi-level Models...........................................................................................................81 Research Qu estion 1.........................................................................................................82 Research Qu estion 2.........................................................................................................83 Chapter 4: Resu lts............................................................................................................. ............................85 Research Qu estion 1............................................................................................................ ............85 Hypothesis 1................................................................................................................... ..85 Hypothesis 2................................................................................................................... 122 Hypothesis 3................................................................................................................... 159 Research Qu estion 2............................................................................................................ ..........190 Hypothesis 1................................................................................................................... 190 Hypothesis 2................................................................................................................... 229 Chapter 5: Discu ssion.......................................................................................................... ........................266 Limitations.................................................................................................................... ................266 Achievement Outcomes Re search Ques tion 1............................................................................268 Mediating Outcomes Re search Ques tion 2.................................................................................279 Variance Explained............................................................................................................. ..........287 Instrumentation................................................................................................................ .............290 Conclusions.................................................................................................................... ...............291 References..................................................................................................................... ..............................293 Appendix A: IRB Application for Exempt Ce rtification........................................................................... ..307 Appendix B: Items from STAR Survey............................................................................................. .........311 Appendix C: Data Prep aration Pr ocedures........................................................................................ ..........318 Appendix D: Permissions........................................................................................................ ....................431 About the Author............................................................................................................... .................End Page

PAGE 7

iv List of Tables Table 1. Variables by Catego ry, Type, Measurement, and So urce in the 2-Le vel Model.............................12 Table 2. Meta-analysis Research Studies about the Integration of Technology and Student Achievement.................................................................................................................... ...............49 Table 3. Large Scale Research Studies about the Relationship of Technology Integration with Student Achievement............................................................................................................ ..........62 Table 4. Research Studies about the Relationship between Technology Integration and Student Behavioral Outcomes............................................................................................................ ..........65 Table 5. Internal Consistency Re liability for Predictors by Year Measured with Cronbach’s Alpha...........76 Table 6. Number of Schools Used in the Analysis by Outcome...................................................................79 Table 7. Descriptive statistics for outcome va riables by school leve l and school year.................................79 Table 8. Model 3: Time, Time2, Time3, and School Level as Predictors of Reading....................................87 Table 9. Model 4a: Reading predicted by Time, School Level, and Demographics Variables No Gifted......................................................................................................................... .....................89 Table 10. Model 4b: Reading predicted by Time, School Level, and Demographics Variables for Elementary and Middle Schools with Gifted..................................................................................91 Table 11. Model 5a: Reading Predicted by Demographics and Student Learning Environment by School Level (All School Le vels without Gifted)...........................................................................92 Table 12. Model 5b: Reading Predicted by Demographics and Student Learning Environment by School Level for Elementary and Middle School with Gifted........................................................95 Table 13. Model 6a: Technology Integration with Demographics and Student Learning Environment by School Level (All Sc hool Levels with out Gifted)................................................98 Table 14. Model 6b: Technology Integration with Demographics and Student Learning Environment by School Level for Elementary and Middle School s with Gifted..........................102 Table 15. Final Model 7a: Significant Technology Integration with Demographics and Student Learning Environment by School Level (A ll School Levels w ithout Gift ed)...............................105 Table 16. Final Model 7b: Significant Technology Integration with Demographics and Student Learning Environment by School Level for Elementary and Mi ddle Schools with Gifted...........108 Table 17. Model Fit Indices for Models Predicti ng FCAT Reading Scores for All School Levels (without Gi fted)............................................................................................................... .............110 Table 18. Model Fit Indices for Models Predicting FCAT Reading Scores for Elementary and Middle School Levels (with Gi fted)............................................................................................ .111 Table 19. Model 3: Time, Time2, Time3, and School Level as Pr edictors of Math.....................................125 Table 20. Model 4a: Math predicted by Time, School Level, and Demographics Variables without Gifted......................................................................................................................... ...................126 Table 21. Model 4b: Math predicted by Time, School Level, and Demographics Variables for Elementary and Middle Sc hools with Gifted................................................................................128 Table 22. Model 5a: Math Predicted by Demographics and Student Learning Environment by School Level (All School Levels without Gifted and LEP)..........................................................130 Table 23. Model 5b: Math Predicted by Demographics and Student Learning Environment by School Level for Elementary and Middle School w ith Gifted......................................................132 Table 24. Model 6a: Math Predicted by Technology Integration with Demographics and Student Learning Environment by School Level (All Sc hool Levels without Gifted and LEP)................135 Table 25. Model 6b: Math Predicted by Technology Integration with Demographics and Student Learning Environment by School Level for Elementary and Mi ddle Schools with Gifted...........139 Table 26. Final Model 7a: Math Predicted by Significant Technology Integration with Demographics and Student Learning Environment by School Level (All School Levels without Gifted and LEP)........................................................................................................ .......143 Table 27. Final Model 7b: Math Predicted by Significant Technology Integration with Demographics and Student Learning Environment by School Level for Elementary and Middle Schools with Gi fted..................................................................................................... .....146

PAGE 8

v Table 28. Model Fit Indices for Models Predic ting FCAT Math Scores for All School Levels (without Gifted and LEP)....................................................................................................... .......148 Table 29. Model Fit Indices for Models Predicting FCAT Math Scores for Elementary and Middle School Levels (with Gifted).................................................................................................... ......148 Table 30. Model 3: Time, Time2, Time3, and School Level as Pr edictors of Writing.................................162 Table 31. Model 4a: Writing predicted by Time, School Level, and Demographics Variables No Gifted......................................................................................................................... ...................163 Table 32. Model 4b: write predicted by Time, School Level, and Demographics Variables No High School in cludes Gifted.................................................................................................... .....165 Table 33. Model 5a: Demographics and Student Learning Environment by School Level (All School Levels w ithout Gi fted).................................................................................................. ....167 Table 34. Model 5b: Demographics and Student Learning Environment by School Level for Elementary and Middle School with Gifted..................................................................................169 Table 35. Model 6a: Technology Integration with Demographics and Student Learning Environment by School Level (All Sc hool Levels withou t Gifted)..............................................172 Table 36. Model 6b: Technology Integration with Demographics and Student Learning Environment by School Level for Elementary and Middle School s with Gifted..........................176 Table 37. Final Model 7a: Significant Technology Integration with Demographics and Student Learning Environment by School Level (A ll School Levels w ithout Gift ed)...............................180 Table 38. Model 8a: Magnet Schools with Significant Technology Integration Demographics and Student Learning Environment by School Level (All School Le vels without Gifted)..................182 Table 39. Model Fit Indices for Models Predicti ng FCAT Writing Scores fo r All School Levels (without Gifted and LEP)....................................................................................................... .......184 Table 40. Model Fit Indices for Models Predic ting FCAT Writing Scores for Elementary and Middle School Levels (with Gi fted)............................................................................................ .185 Table 41. Model 3: Time, Time S quared, and School Type as Pred ictors of Student Absences.................192 Table 42. Model 4a: Student Absences Predicted by Time, School Type, and Demographics Variables (N o Gifted).......................................................................................................... ..........194 Table 43. Model 4b: Student Absences predicted by Time, School Level, and Demographics Variables for Elementary and Mi ddle Schools with Gifted..........................................................195 Table 44. Model 5a: Absences Predicted by Demographics and Student Learning Environment by School Level (All School Leve ls without Gi fted).........................................................................197 Table 45. Model 5b: Absences Predicted by Demographics and Student Learning Environment by School Level for Elementary and Middle School w ith Gifted......................................................199 Table 46.Model 6a: Technology Integration with Demographics and Student Learning Environment by School Level (All Sc hool Levels withou t Gifted)..............................................201 Table 47. Model 6b: Technology Integration with Demographics and Student Learning Environment by School Level for Elementary and Middle School s with Gifted..........................205 Table 48. Final Model 7a: Significant Technology Integration with Demographics and Student Learning Environment by School Level (A ll School Levels w ithout Gift ed)...............................208 Table 49. Final Model 7b: Significant Technology Integration with Demographics and Student Learning Environment by School Level for Elementary and Mi ddle Schools with Gifted...........210 Table 50. Model 8a: Magnet Schools with Significant Technology Integration Demographics and Student Learning Environment by School Level (All School Le vels without Gifted)..................212 Table 51. Model 8b: Magnet Schools with Significant Technology Integration Demographics and Student Learning Environment by School Level for Elementary and Middle Schools with Gifted......................................................................................................................... ...................214 Table 52. Model Fit Indices for Models Predicting FCAT Student Absences Scores for All School Levels (without Gifted)........................................................................................................ .........215 Table 53. Model Fit Indices for Models Predicting FCAT Student Absences Scores for Elementary and Middle School Le vels (with Gifted)......................................................................................21 6 Table 54. Model 3: Time, Time Squared, and Sc hool Type as Predicto rs of Misconduct..........................231 Table 55. Model 4a: Misconduct predicted by Time, School Type, and Demographics Variables (No Gifted).................................................................................................................... ................232 Table 56. Model 4b: Misconduct predicted by Time, School Level, and Demographics Variables for Elementary and Middle Schools with Gifted...........................................................................234

PAGE 9

vi Table 57. Model 5a: Misconduct Predicted by Demographics and Student Learning Environment by School Level (All School Levels without Gifted)....................................................................236 Table 58. Model 5b: Misconduct Predicted by Demographics and Student Learning Environment by School Level for Elementary an d Middle School w ith Gifted.................................................238 Table 59. Model 6a: Technology Integration with Demographics and Student Learning Environment by School Level (All Sc hool Levels withou t Gifted)..............................................240 Table 60. Model 6b: Technology Integration with Demographics and Student Learning Environment by School Level for Elementary and Middle School s with Gifted..........................243 Table 61. Final Model 7a: Significant Technology Integration with Demographics and Student Learning Environment by School Level (A ll School Levels w ithout Gift ed)...............................247 Table 62. Final Model 7b: Significant Technology Integration with Demographics and Student Learning Environment by School Level for Elementary and Mi ddle Schools with Gifted...........249 Table 63. Model 8a: Magnet Schools with Significant Technology Integration Demographics and Student Learning Environment by School Level (All School Le vels without Gifted)..................251 Table 64. Model 8b: Magnet Schools with Significant Technology Integration Demographics and Student Learning Environment by School Level for Elementary and Middle Schools with Gifted......................................................................................................................... ...................252 Table 65. Model Fit Indices for Models Predicting Student Misconduct Scores for All School Levels (without Gifted)........................................................................................................ .........254 Table 66. Model Fit Indices for Models Predicting Student Misconduct Scores for Elementary and Middle School Levels (with Gi fted)............................................................................................ .255 Table 67. Variance for Each Model for Achievement Ou tcomes by Dataset..............................................289 Table 68. Variance by Each Model for Ea ch Mediating Outcom e by Dataset...........................................289 Table C 1. Master School Identification Files: Schools by Type for each School Year..............................320 Table C 2. Schools Participat ing in the FC AT each Year......................................................................... ..321 Table C 3. Missing Mean School Level FCAT Test Scores for Mathematics by School Level and Year........................................................................................................................... ....................322 Table C 4. Number of Schools by School Level included in the FSIR and Merged with the MSID by Year........................................................................................................................ ..................324 Table C 5. Missing Demographic Indicators in th e FSIR and AYP by School Level for 2003-04.............326 Table C 6. Missing Demographic Indicators in th e FSIR and AYP by School Level for 2004-05.............327 Table C 7. Missing Demographic Indicators in th e FSIR and AYP by School Level for 2005-06.............328 Table C 8. Schools in Original STAR Data and Merged with MSID by School Level and by School Year.................................................................................................................... ...............329 Table C 9. Missing Response for Items in the STARS Survey for each School Year.................................330 Table C 10. Schools with Missing Response s for All STAR Item s in 2003-04..........................................334 Table C 11. Number of schools only Missing Responses for Level of Support Items for 2003-04 to 2005-06........................................................................................................................ .................335 Table C 12. Descriptive Statistic s for FCAT Outcome Scores....................................................................3 37 Table C 13. Descriptive Statistics of Demographic Variables in the Florida School Indicators Reports........................................................................................................................ ..................339 Table C 14. Descriptive Statistics of the Technology Integration Variables from the Florida Innovates (STA R) Survey........................................................................................................ .....346 Table C 15. Common Factor Analysis with Oblique Rotation: Student Lear ning Environments...............385 Table C 16. Common Factor Analysis with Ob lique Rotation: Teacher Qualifica tions.............................386 Table C 17. Common Factor Analysis with Obliq ue Rotation: Student Ac cess to Software......................387 Table C 18.Common Factor Analysis with Oblique Rotation: Teachers Regularly use Technology.........392 Table C 19. Common Factor Analysis with Oblique Rotation: Frequency St udents Use Software............395 Table C 20. Common Factor Analysis with Ob lique Rotation: Suppor t for Technology............................397 Table C 21. Descriptive Statistics of Predicto r Variables for FCAT Reading Outc ome.............................398 Table C 22. Descriptive Statistics of Predic tor Variables for FCAT Math Outcome..................................404 Table C 23. Descriptive Statistics of Predic tor Variables for FCAT Writing Outc ome..............................410 Table C 24. Descriptive Statistics of Predic tor Variables for Absences Outcome......................................416 Table C 25. Descriptive Statistics of Predicto r Variables for Student Conduct Outcome...........................421 Table C 26. Correlations and P-values of Pred ictor Variables for Lear ning Environment and Technology Indicators for FC AT Reading Outcome....................................................................426

PAGE 10

vii Table C 27. Correlations and P-values of Pred ictor Variables for Lear ning Environment and Technology Indicators for FC AT Math Ou tcome.........................................................................427 Table C 28. Correlations and P-values of Pred ictor Variables for Lear ning Environment and Technology Indicators for FC AT Writing Ou tcome.....................................................................428 Table C 29. Correlations and P-values of Pred ictor Variables for Lear ning Environment and Technology Indicators for Absences Outcome.............................................................................429 Table C 30. Correlations and P-values of Pred ictor Variables for Lear ning Environment and Technology Indicators for Conduct Outc ome...............................................................................430

PAGE 11

vi List of Figures Figure 1. Model of Technology Integration with Contextual Variab les in MultiLevels................................6 Figure 2. Degree of Learning Carr oll Model of Scho ol Learning.................................................................2 1 Figure 3. Optimal Learning in Ca rroll Model of School Learning................................................................2 2 Figure 4. Quality, Appropriateness, Incentiv e, and Time (QAIT) Model of instructional effectiveness relating alterable elements of instruction to student achievement (Slavin, 1987).......................................................................................................................... .....................24 Figure 5. Relationship between Frequency Students Use Tool-base Software and FCAT Reading in High Schools................................................................................................................ .............113 Figure 6. Relationship between Frequency Students Use Tool-base Software and FCAT Reading in Middle Schools.............................................................................................................. ...........114 Figure 7. Relationship between Frequency Students Use Tool-base Software and FCAT Reading in Elementary Schools.......................................................................................................... ........115 Figure 8. Relationship between Percent of Gifted Students on FCAT Reading by School Level (Gifted Included).............................................................................................................. .............116 Figure 9. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and FCAT Reading in Middle Schools (Gifted Included)....................117 Figure 10. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and FCAT Reading in Elementary Schools (G ifted Included)..............118 Figure 11. Relationship between Frequency that Students Use Content Software and FCAT Reading in Middle Schools (Gifted Incl uded)..............................................................................119 Figure 12. Relationship between Frequency that Students Use Content Software and FCAT Reading in Elementary School s (Gifted Included)........................................................................120 Figure 13. Relationship between Technical Support for Hardware and FCAT Reading in Middle Schools (Gifted Included)...................................................................................................... .......121 Figure 14. Relationship between Technical Support for Hardware and FCAT Reading in Elementary Schools (G ifted Included)..........................................................................................1 22 Figure 15. Relationship between the Percent of Teachers Who Regularly Use Technology for Administrative Purposes and FCAT Math in High Schools.........................................................150 Figure 16. Relationship between the Percent of Teachers Who Regularly Use Technology for Administrative Purposes and FCAT Math in Middl e Schools. .....................................................152 Figure 17. Relationship between the Percent of Teachers Who Regularly Use Technology for Administrative Purposes and FCAT Ma th in Elementary Schools...............................................153 Figure 18. Relationship between Percent of Gifted Students on FCAT Math by School Level (Gifted Included).............................................................................................................. .............154 Figure 19. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and FCAT Math in Middle Schools (Gif ted Included).........................156 Figure 20. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and FCAT Math in Elementary Schools (G ifted Included)..................157 Figure 21. Relationship between the Percentage of Teachers Who Regularly Use Technology for Delivery of Instruction and FCAT Math in Middle Schools (Gif ted Includ ed)............................158 Figure 22. Relationship between the Percentage of Teachers Who Regularly Use Technology for Delivery of Instruction and FCAT Math in Elementary Schools (G ifted Included).....................159 Figure 23. Relationship between Frequency that students use content software and FCAT Writing in high schools................................................................................................................ ..............187 Figure 24. Relationship between Frequency that Students Use Content Software and FCAT Writing in Midd le Schools...................................................................................................... ......188 Figure 25. Relationship between Frequency that students use content software and FCAT Writing in elementary schools.......................................................................................................... ..........189

PAGE 12

vii Figure 26. Relationship between Percent of Gifted Students on FCAT Writing by School Level (Gifted Included).............................................................................................................. .............190 Figure 27. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Student Absences in High Schools.................................................218 Figure 28. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Student Absences in Midd le Schools.............................................219 Figure 29. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Student Ab sences in Elementa ry Schools......................................221 Figure 30. Relationship between Percent of Gifted Students on Student Absences by School Level (Gifted Included).............................................................................................................. .............222 Figure 31. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and St udent Absences in Middle Sc hools (Gifted Included)................224 Figure 32. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Absences in El ementary Schools (Gifted Included).......................225 Figure 33. Relationship between Technology Support – Human and Student Absences in High Schools........................................................................................................................ ..................227 Figure 34. Relationship between Technology Support – Human and Student Absences in Middle Schools........................................................................................................................ ..................228 Figure 35. Relationship between Technology Support – Human and Absences in Elementary Schools........................................................................................................................ ..................229 Figure 36. Relationship between Percent of T eachers Who Regularly Use Technology to Deliver Instruction and Student Miscon duct in High Schools...................................................................256 Figure 37. Relationship between Percent of T eachers Who Regularly Use Technology to Deliver Instruction and Student Miscon duct in Middle Schools................................................................257 Figure 38. Relationship between Percent of T eachers Who Regularly Use Technology to Deliver Instruction in Elem entary Schools.............................................................................................. ..258 Figure 39. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Student Misconduct in Hi gh Schools.............................................259 Figure 40. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Student Mi sconduct in Midd le Schools ..........................................260 Figure 41. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Student Misc onduct in Elementa ry Schools...................................261 Figure 42. Relationship between Percent of Gifted Students and Level of Student Misconduct by School Level (Gif ted Included)................................................................................................. ....262 Figure 43. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Student Misconduct in Middle Schools (Gifted Included).............263 Figure 44. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Misconduct in El ementary Schools (Gif ted Included)...................265 Figure 45. Relationship between Frequency that Students Use Content Software and FCAT Reading at All School Le vels without Gifted...............................................................................271 Figure 46. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and FCAT Reading at All School Levels without Gifted.....................276 Figure 47. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and FCAT Math at All School Levels without Gifted........................277 Figure 48. Relationship between U.S. Magnet Sc hool Status and Mean FCAT Writing Scores in All Sc hools.................................................................................................................... ................278 Figure 49. Relationship between Percent of T eachers Who Regularly Use Technology to Deliver Instruction and Student Misconduct at A ll School Levels (without Gifted).................................282 Figure 50. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Percent of Stud ents Absent More than 21 Days in All Schools withou t Gifted......................................................................................................... .........284 Figure 51. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Student Misconduct at All School Levels (without Gifted)...........285 Figure 52. Relationship between U.S. Magnet School Status and Student Absences at All School Levels......................................................................................................................... ...................286

PAGE 13

viii Figure 53. Relationship between US Technology Magnet School and Student Absences at All School Levels without Gifted................................................................................................... .....287

PAGE 14

ix The Relationship between Technology Integration and Achievement Using Multi-Level Modeling Tina N. Hohlfeld ABSTRACT The purpose of this longitudinal study was to examine the relationship between technology integration indicators and school level achievement. Four years of school level secondary data from publicly available databases maintained by the Flor ida Department of Education were combined for all public elementary, middle, and high schools in the st ate. This study examined approximately 2300 schools that participated each year in the Florida Innovates Survey about technology integration between 2003-04 and 2006-07. Complexity theory supported the use of multi-level modeling to examine the relationships between technology integration and outcomes. Thre e achievement outcomes (reading, mathematics, and writing) and two mediating behavioral outcomes (attendance and misconduct) were investigated. Moderating variables controlled in the model included school level, demographics, and learning environment. After data preparation, all composite variables were developed using factor analysis. Models were progressively built with significant variables at eac h level retained in subsequent levels of the study. A total of 94 models were estimated with maximu m likelihood estimation using SAS 9.1.3 statistical software. The integration of technology is only one of the many factors that impact student learning within the classroom environment. Results supported prev ious research about the relationship between the moderating variables and school level achievement and confirmed the need to include moderating variables in the model. After controlling for all the other moderating variables, technology integration had a significant relationship with mean school achievement. Although the percent of teachers who regularly use technol ogy for administrative purposes was consistently significant in the models for four out of five outcomes studied, the interactions with time, time2, and time3, resulted in curvilinear trends with inconsis tent results. These inconsistent significant findings make drawing conclusions about the integration of technology within Florida’s public schools difficult. Furthermore, the small changes observed in mean school achievement over the span of this study support the concept that time is a critical factor for sc hool level learning and change. Therefore, continued analyses of the longitudinal trends for Florida schools in the relationship between technology integration variables and school achievement, while controlling for moderating variables, are recommended.

PAGE 15

1 Chapter 1: Introduction Due to the multiple demands on schools for allocating their resources, the return on investment of technology for improving student achievement is an extremely important consideration for responsible decision-making. However, the variety of factors th at impact achievement within the complex school environment makes the study of educational phenomena difficult. The many variables at the individual, classroom, and school levels that moderate student ac hievement cause the assessment of the isolated effect of technology to be problematical (Bryk & Hermonson, 1993). Multilevel models and special statistical computer programs allow for simultaneous analysis of the disaggregated impact of several levels of contextual factors on achievement (Luke, 2004; Raudenbush & Bryk, 2002). However, scant research has been conducted using these methods on longitudinal data to examine the relationship of the integration of technology on student achievement. Indeed, the St ate Educational Technology Directors Association, a national organization of all state technology directors, has called for this type of research to be conducted using state-wide data (Lemke, Wainer, & Haning, 2006) With 61% of states now requiring [Local Edu cation Agencies] LEAs that receive competitive grants to “report findings based on improvements as compared to baseline data,” it is only a matter of time before states will be able to report statewide summaries of correlational results. In addition, nearly 25% of states report that they have commissioned or funded research studies on the impact of technology on lear ning (Lemke et al., 2006, p. 5). Florida is one of the first states to create a da ta warehouse to help inform state policy makers, district and school planners, and other stakeholders about the current status and progress of the state’s educational initiatives. Within this system, trend data on both student performance and technology indicators has been systematically coll ected from all of the schools by their districts in order to study the impact of specific programs on student performance (Bureau of Instruction and Innovation, Florida Department of Education, 2007a; Technology Counts, 2006). This study used multilevel statistical analysis with longitudinal data collected by the Florida Department of Education to investigate the relationships between technology integration and overall school achievement (related to mathematics, reading, and writing).

PAGE 16

2 Background Information According to the 2006-07 Florida State Governmen t Technology Investment Forecast Florida allocated over 17.5 million dollars to the State Board of Education to spend on educational technology initiatives during fiscal year 2007 (pjmathison, 2006). In addition to receiving state funding, the Florida Department of Education also obtained over 88 million dollars over the past three years from the No Child Left Behind, Title II, Part D E nhancing Education Through Technology program (Lemke et al., 2006). While many reformers of education believe that tech nology will transform the way students are educated, provide equitable learning opportunities for all students, and improve children’s competitive advantage (Dede, Korte, Nelson, Valdez, & Ward, 2005; Ringstaff & Kelley, 2002), others believe that any benefits from the purchase of technology will not outweigh the tremendous investment (Cuban, 1986, 1998, 2001; Oppenheimer, 2003; Tyack & Cuban, 1995). Conseque ntly, it is crucial to determine if Florida’s investment in technology has had a positive impact on student outcomes. Noteworthy, despite the huge increases in access to computers and the Internet both at school and at home for most students, the digital divide has continued for many students based on their socioeconomic status and ethnicity (e.g., DeBell & Chap man, 2006 and Parsad & Jones, 2005). It was imperative that the investigation of technology integration on student outcomes also examined the equity of the results across various subgroups. Thus, socio-economic status and ethnicity variables were included in the analysis of this study. Theoretical Framework Complexity theory can be used as a framework to explain the dynamics that occur during the schools’ change process as they integrate technology into their curricula. Complexity theory has been used to study and explain the workings of complex systems in many disciplines, including physical sciences, biology, business, and sociology (Jacobson & Wilensky, 2006; Morrison, 2002). All of these systems have multiple levels of organization and heterogeneous components (Caldwell, 2005; Jacobson & Wilensky, 2006). Complexity theory perspective allows the simultaneous examination of phenomena on both the micro and macro level (Caldwell, 2005; Jacobson & Wilensky, 2006). These modeling techniques and new computer statistical programs allow educational systems to be examined to understand the effect of policy

PAGE 17

3 decisions and to inform subsequent policy (Jacobson & Wilensky, 2006; Luke, 2004; Raudenbush & Bryk, 2002; Wenglinsky, 2005). Applying complexity theory to educational organizations fits well on all levels (Morrison, 2002; O’Day, 2002). On the micro level or within the classroom, the students and teachers are independent agents that interact with each other. So metimes individual students or teach ers change and are replaced by new agents; however, the dynamics of that particular classroom environment co ntinues to impact all participants or agents within it. On the macro level or at the sc hool or district level of the organization, there are multiple variables that impact the activities of the organization such as the mission and goals, strategic plan, budget, curriculum, and management style. On the micro-level, the process of teaching and learning is also complex and has many confounding variables from both the individual students and teacher as well as from the classroom and even the school. Carroll’s Model of School Learning (1963, 1989) can be used to explain the dynamics of the learning process within this complex environment. This model has two main categories of factors that impact school learning: time needed in learning and time spent in learning. By examining Carroll’s Model of School Learning through the lens of complexity theory, the factors can be separated into two levels – the individu al level and the classroom level, which Carroll calls individual and external conditions (Carroll, 1963). The factors at the individual level are aptitude, ability to understand directions, and perseverance, while the factors at the classroom level include quality of instruction and opportunity for learning. These factors at both levels interact through the dynamic process of teaching and learning. Thus, an im portant contribution of this model is the explanation of the importance of time in learning. Carroll presents a formula that calculates the degree of learning as being the function of the time actually spent learning divided by the time that is needed (Berliner, 1990; Carroll, 1963). Maximum degree of learning occurs when a student actively engages in learning for the time that the student needs when all other conditions are optimal. When conditions are not optimal, such as when the instruction delivered is not organized in the most accessible manner for the student or when the student does not have the prerequisite learning required for understanding, then the time needed to learn increases. If the time allowed for leaning is not equal to the time needed, then the amount of learning is decreased.

PAGE 18

4 The dynamic process of teaching and learning within the classroom resides within yet another level, the macro level of school. Within this level, there are also agents (teachers and administrators) who interact with and have impact on th e classes and staff within this macro level, as well as have interactions with students on the micro level. These nested levels continue to expand, because schools interact within the district, thereby creating another set of impacting variables within the school organization. The organization as a whole has a common mission of educating students and communicating the methods through the strategic plan at the district level and school improvement plans at the school level. Communication and interactions are reciprocal and iterativ e; that is, the micro level components impact the macro level components and vice versa. Ultimately, the organization exchanges information with the outside community and other school organizations, and, in turn, responds and adapts as an organization. When new technology is added to the environment at multiple levels, technology becomes one of the agents that stimulates change and adaptation (Cough lin & Lemke, 1999; Culp, Honey, & Maninach, 2005; Wenglinsky, 2005). When the Carroll Model of School Learning is applied to the organization level of school, with multiple teachers learning to integrat e technology into their curriculum, it is apparent that the opportunity to learn or time will be the critical factor required for change to occur. Consequently, the technology integration change process must be continuous, as it re quires extended time for teachers to progress through several stages -entry, adoption, adaptation, appropriation, and invention (Apple Computer, Inc., 1995; Coughlin & Lemke, 1999; Dwyer, Ringstaff, & Sandho ltz, 1990), while they lear n to use and incorporate technology into their lessons. Because teachers need time to progress th rough the stages of integration of technology, the impact of their curricular plans that integrate technology into their daily instructional practices on their students’ learning also will ta ke time. Thus, the relationship between technology integration and student achievement must be studied over time. Technology Integration Essential Organizational Conditions for Technology Integration Fundamental conditions have been identified among schools that have suc cessfully integrated technology and improved student outcomes. These factors at the organizational level include broad-based educational reform efforts and long-range plans, while the influential factors at the school level include

PAGE 19

5 readily accessible technology, adequate and appropriat e staff development, ongoing support, changes in teachers’ beliefs about teach ing and learning, and techno logy integrated in to the curricula along with other teaching methods (Ringstaff & Kelley, 2002). The In ternational Society for T echnology in Education (ISTE) designated these components as essential conditions for technology-enriched learning environments (ISTE NETS Project, 2007). These essential conditio ns include access to contemporary technologies, technical assistance, skilled educators, community partners, and political and financial support for technology. In all of these recommendations, accessibility and use of technology are crucial ingredients within the context of the other nested levels of school variables. School Level Factors that Impa ct Technology Integration Within the school learning environment, the impact of the integration of technology on student achievement is complicated by many dynamic factors on multiple levels. The variables overlap, interact, and moderate each other on a daily as well as longterm basis during teaching and learning. These multilevels of the technology integration factors and contextu al variables are depicted in Figure 1. The sections that follow delineat e these variables.

PAGE 20

6 Figure 1. Model of Technology Integration with Contextual Variables in Multi-Levels Note: Technology integration variables are in bold text. Uses of Technology In order to become proficient users of te chnology, both students and teachers must have opportunities to utilize computers (Dwyer et al., 1990; ISTE NETS Project, 2005b, 2007). Thus, the school must make computers and software available. Notably, another ongoing issue is how to efficiently educate students so they are both literate and proficient technol ogy users (Barrnett, 2003). This debate about how to effectively integrate technology for the support of improved achievement revolves around pedagological philosophies at the school level.

PAGE 21

7 Instruction delivery. On one side, many educators believe the power of technology should be used for delivering instruction to the student by providing efficient and patient delivery of individualized instruction aimed at mastery learning at the appropriate level (e.g., curriculum-based software such as tutorials and individually administered learning packages). We st Virginia incorporated Integrated Learning Systems (ILS) in a large state-wide reform initiative in all elementary schools. Results of a longitudinal study about this initiative indicated that students made significant gains in basic skills of reading and mathematics (Mann, Shakeshaft, Becker, & Kottkamp, 1999). Meta-analyses are research studies that statistically combine the results of individual experimental or quasi-experimental studies conducted on a particular intervention to establish an overall effect size. A meta-analysis exam ining the impact of ILS on mathematic achievement found educationally meaningful and statistically signi ficant positive effects (Kulick, 2003). In addition, control studies of tutorial programs in all subjects almost always had an edu cationally meaningful and significantly positive effect (Kulick, 2003). However, not all research supports the effectiven ess of using technology to deliver instruction. Lockee, Moore, and Burton (2004) reported in their research synthesis on programmed instruction, which is the foundation for computer-assisted, computer-based tu torials, and web-based tuto rials, that almost all research conducted was of poor quality so that the resu lts could not be generalized beyond that particular study. The results of poor quality research cannot be used to support the effectiveness of using technology. In addition even when the quality of the research wa s adequate, there were still contradictory results for how to use technology to deliver instruction (e.g., Hill, Wiley, Nelson, & Han, 2004; Mory, 2004; Park & Lee, 2004; Shapiro & Niederhauser, 2004). In a recent study about the effects of math and reading software products, no significant difference was found between the test scores of students using the products and control students who were not (Dynarski, Agodini, Heaviside, Novak, Carey, Campuzano, Means, Murphy, Penuel, Javitz, Emery, & Sussex, 2007). Product creation. On the other side of the debate, many educators believe the power of technology should be used as a creative tool by students to support their construction of concepts and knowledge, higher order thinking skills, and problem solving (e.g., word processing concept mapping, spreadsheets, and databases).

PAGE 22

8 Research conducted by the Apple Classroom of Tomorrow (ACOT) program found that students who used computers as tools in the classroom not only performed well on standardized tests, but also exhibited other skills such as collaborating with peers, presenting information in a variety of way, communicating well about complex processes, and learning independently (Apple Computer, Inc., 1995). Similar results have been found by three meta-analyses about the impacts of using word processors on student writing (Goldberg, Russell, & Cook, 2003; Kulick, 2003; Pe nuel, Kim, Michalchik, Lewis, Means, Murphy, Korbak, & Whaley, 2002). In add ition, Lowther, Ross, and Morriso n (2003) conducted a mixed method study comparing using laptops vs. computers in the classroom on middle school students’ writing achievement gains and found significant positive results when using laptop computers. Furthermore, with multi-level modeling statistical analysis, O’Dwyer, Russell, Bebell, and Tucker-Seeley (2005) found significant positive relationships between elementary school students’ writing achievement and their use of a computer in school to edit papers. Jonassen and Reeves (1996) recommend that com puters be used as cognitive tools that support learning and concept formation. For example, students can practice higher order problem solving skills by working with computers as tools for researching and for creating products. Taylor, Casto, and Walls (2007) compared the achievement results of two groups of students, one with technology, and the other without technology, engaged in the same units of study. They found that using technology had a positive impact on student achievement at both the elementary and se condary levels. Kulick (2003) reported educationally meaningful and statistically signifi cant effects when integrating the use of word processing in the instruction of writing. These and other researchers pr opose that using computers as cognitive tools supports students’ higher-level thinking skills and deeper underst anding of content. For instance, computer-mediated communication or the threaded discussion is another important tool used to support learning through writing, especially within distance-learning courses. However, Romiszowski and Mason (2004) report that research about effectiveness of computer-mediated co mmunication for supporting learning has been scarce and inconclusive. Assessment. Technology can be used for measuring learning. However, most standardized tests do not use technology as an assessment vehicle. Thus, traditional standardized tests may not sample all of the skills

PAGE 23

9 and abilities that these students have learned from the integration of technology (Russell & Higgins, 2003; Wenglinsky, 2005). Student Achievement Student achievement is impacted by many interacting factors at various school levels, in a variety of ways, and within different content areas. To i nvestigate the relationships of specific technology integration variables, all of these interacting factors must be identified and then included in each of the models that examine relationships between technology integration and each outcome. School Level Factors that Impact Achievement Teacher. The use of technology is one of many instructional strategies that must be woven together by teachers to enhance students’ learning opportunities. In tegrating technology that engages students within a nurturing and motivating learning environment requires both art and skill on the part of the teacher. Furthermore, the teachers’ attitudes and comfort level with technology as well as their technology skills impact the effectiveness of their t echnology instruction (Becker, 2001). The skill level of the teachers and quality of their interactions with students is impacted by their previous teaching experiences, their level of formal education and expertise within their field, and their ongoing participation in professional development (Marzano, 2003; National Center for Education Statistics, 2005). Classroom. Besides the teacher, other factors impact the dynami cs of the learning environment and the process of the teaching and learning. Students learn best with daily exposure to the curricula materials within a learning environment that has predictable and consistent procedures (Marzano, 2003). School-wide attendance and stability impact the consistency of the learning environment and the amount of curriculum that is covered. The number of stud ents within the classroom impacts the time the teacher has to give each student individual attention and meet each student’s learning needs. Students’ positive and negative interactions also affect the quality of the learning e nvironment. School climates that are not conducive to learning have more incidents of non-academic student behavior and student misconduct (National Center for Education Statistics, 2005). Using technology at school has been related to improved attendance and better conduct (Barron, Hogarty, Kromrey, & Lenkway, 1999). Improving attendance and increasing on-

PAGE 24

10 task academic behavior could increase achievement. This study investigated attendance and student conduct as both outcome and predictor variables. Student demographics Education is a complex phenomenon. Students at different developmental levels may respond differently to different modes of technology integration. Young elementary students learn the prerequisite skills (e.g., how to read; how to do arithmetic computations; and organizational skills) that older students use as tools for learning new information and concepts (Bruning, Schraw, Norby, & Ronning, 2004; Marzano, 2003). Hence, the method of technology integration may in teract with the predominant kinds of learning tasks that students must accomplish an d differentially impact students’ achievement. Overall school demographic factors (i.e., socio-economic level, and proportion of minorities and special populations) may moderate the dynamic learni ng process, and, in the end, the mean academic attainment of the organization (National Center for Education Statistics, 2005). Researchers have reported differential instructional methods used in schools based on high and low socio-economic status (Becker, 2001; Lubienski, 2006; Wenglinsky, 2004). Ultimately students’ achievement, attendance, an d conduct are impacted by numerous experiences in multiple classrooms with many teachers within a school. Althou gh many studies investigating technology use in elementary and secondary education have demonstrated positive gains in achievement (e.g. Kulik, 2003; Slavin, 2005; Taylor et al., 2007; Wenglinsky, 2005); other researchers have reported inconclusive results or no gains (Gredler, 2004; Hill et al., 2004; Lockee et al., 2004; McLellan, 2004; Metri Group, 2006; Mory, 2004; Park & Lee, 2004; Rieber, 2004; Romiszowski & Mason, 2004; Shapiro & Niederhauser, 2004). However, these studies may have not utilized the statistical methods that can examine the multilevel variables that are nested with in the complex educational environment (O’Dwyer, Russell, & Bebell, 2004, 2005; Wenglinsky, 2005 ). Wenglinsky (2005) provides recommendations to enhance the analysis of his previous studies with NA EP data by using multi-level modeling statistics with the nested data. Another cautionary note about the methodology of the study. The study utilizes SEM, a regression-type technique. While the technique possesses many advantages over conventional regression it does not directly take into account the multiple levels of analysis involved in the analysis of school data. The NAEP data occur at multiple levels; many of the independent variables are at the school level, whereas the dependent variable is at the student level. While the use of design effects takes the clustered nature of the sample into account in adjusting standard

PAGE 25

11 errors, it does not explicitly model this clustering. Thus, some techniques that both had the advantages of SEM and took into account the multilevel nature of the data would be preferable (Wenglinsky, 2005, p. 89). Unlike Wenglinsky’s (2005) study that used cro ss-sectional data from National Assessment of Educational Progress (NAEP), this study utilized longitudinal data that tracked technology and achievement of the same schools over four years. Examining the large longitudinal datasets of Florida with multilevel modeling statistical analysis allowed the examination of the relative proportion of the impact of technology (Luke, 2004). Furthermore, statistical adjustments were made for the differential impacts of student demographics and the attributes of the school learning environments on achievement (Raudenbush & Bryk, 2002). Research Questions The following research questions were investigated: 1. What is the relationship between indicators of technology integration and changes in mean school achievement (FCAT NRT scaled scores for reading, mathematics, and FCAT rubric score in writing) when controlling for school level (elementary, middle, and high), school socio-economic status, percent of minority students, percent of limited English proficiency students, percent of stude nts with disabilities, teacher qualifications, and learning environment quality? 2. What is the relationship between indicators of technology integration and changes in mediating outcomes (attendance rates and student conduct)? The following hypotheses were us ed to answer these questions: 1. After controlling for school level (elementar y, middle, and high), school socio-economic status, percent of minority students, percent of limited English proficiency students, percent of students with disabilities, teach er qualifications, and learning environment quality, mean school achievement (FCAT NRT scaled scores for reading and mathematics and FCAT rubric scores for writing) will have a positive relationship with indicators of technology integration. 2. After controlling for school level (elementar y, middle, and high), school socio-economic status, percent of minority students, percent of limited English proficiency students, percent of students with disabilities, teach er qualifications, and learning environment quality, mean school absence rates will have a negative relationship with indicators of technology integration and mean school level of student misconduct will have a negative relationship with indicators of technology integration. Research Plan To investigate the research questions and test thes e hypotheses, this study used repeated measures with 2-level modeling to assess the relationships betw een technology integratio n factors and changes in attendance, conduct, reading, mathematics, and writin g achievement at the school level. This study was conducted using four points of time (2003-04, 2004-05, 2005-06, and 2006-07 school years) as the repeated

PAGE 26

12 measures in the first level in the model and school-l evel variables at the second level. The following categories of predictor variables were included in the model: school level demographics; school level learning environment, school level teacher qualifications, and school le vel technology integration. Outcome variables were school level Florida Competency As sessment Test (FCAT) Norm Referenced Test (NRT) scaled scores in reading and mathematics and the FCAT rubric scores in writing. In addition, school level changes in attendance and student conduct were used in the model as both outcome variables and moderating variables. Multilevel modeling allowed the disaggregated analysis of technology integration within the nested data by statistically controlling for the effects of the other confounding variables in the multi-level models. (Luke, 2004; Raudenbush & Bryk, 2002). Table 1 delineates the predictor and outcome variables that were used in this study, their type, how they are measured, and their source. Variables Table 1. Variables by Category, Type, Measureme nt, and Source in the 2-Level Model Variable Type Measurement Source Technology Integration Predictor Variables Technology Support rank “Our school-based technical support is provided by:” + “Our school-based instructional technology specialist is:” + “How dependable is the Internet connection at your school?” + “How often do you experience delays when using the Internet at your school?” + “What is the average length of time at your school for a technical issue to be resolved?” STAR Survey Teachers regularly use for delivery of instruction continuous “Approximately what percentage of your teachers regularly uses technology in the following ways?” Percentage ranges are converted to average for the range and then all percentages for the all uses will be averaged STAR Survey Teachers regularly use for administrative purposes continuous “Approximately what percentage of your teachers regularly uses technology in the following ways?” Percentage ranges are converted to average for the range and then all percentages for the all uses will be averaged STAR Survey

PAGE 27

13 Variable Type Measurement Source Student access to content software continuous “What percentage of student computers at your school has the following software types available on them?” Percentage ranges are converted to average for the range STAR Survey Student access to office/ production software continuous “What percentage of student computers at your school has the following software types available on them?” Percentage ranges are converted to average for the range STAR Survey Student access to advanced production software continuous “What percentage of student computers at your school has the following software types available on them?” Percentage ranges are converted to average for the range STAR Survey Frequency that students use content software ordinal “How often do students at your school use the following types of software?” STAR Survey Frequency that students use toolbased software ordinal “How often do students at your school use the following types of software?” STAR Survey Technology Magnet School categorical School was designated as a magnet school or program with a specialty in technology in 200506 Master School Identification File Learning Environment Predictor Variables Learning Environment continuous Students Absent 21+ Days; Stability Rate ; proportion Suspensions and Incidents of Crime and Violence, Offenses per Number of Students Florida School Indicators Report Teacher Qualifications continuous Average Years of Experience; Master’s Degree or Higher; Classes Taught by Teachers Teaching Out of Field – for analysis proportion in field will be used Florida School Indicators Report School Level Elementary categorical binary MSID files Middle/ Junior categori cal binary MSID files High categorical binary MSID files Demographic Variables Free or Reduced Lunch Status continuous Economically Disadvantaged Students AYP Report Minority continuous FCAT Reading/ SSS Results – Number of Students White, Black, Hispanic, Asian/Pacific Islander, American Indian/Alaskan, AYP Report

PAGE 28

14 Variable Type Measurement Source Limited English Proficiency (LEP)* continuous Limited English Proficiency/ ESOL AYP Report Florida School Indicator Report* Students with Disabilities* continuous Students with Disabilities AYP Report Florida School Indicator Report* Gifted* continuous Gift ed* Elementary & Middle School AYP Report Florida School Indicator Report* Outcome Variables: Achievement Reading continuous Reading FCAT (NRT) scale score Assessment and School Performance Math continuous Mathematics FCAT (NRT) scale score Assessment and School Performance Writing continuous Writing FCAT rubric score Assessment and School Performance Outcome Variables: Mediating Variables Absence Rate* continuous change in percentage of Students Absent 21+ Days Florida School Indicators Report Florida School Indicators Report 2003-04 to 2005-06* Student Misconduct* continuous change in proportion Suspensions and Incidents of Crime and Violence, Offenses per Number of Students Florida School Indicators Report 2003-04 to 2005-06* Florida Indicators Report is only available until the 2005-06 schools year Definitions Technology for this study included computer software and associated hardware (i.e., scanners, printers, DVD players, projectors, mp3 players, personal organizers, and digital cameras). Technology integration occurs when technology is used as an integral component together with other instructional methods to support students’ learning of the designated curriculum. For this study, technology integration referred to using the com puter to support student achievement with either curriculum-based software or tool-based software.

PAGE 29

15 Socio-economic status (SES) designates the level of access family access to resources. Typical measures include family income, pare nts’ education, parents’ occupation, and educational resources in the home (Lubienski, 2006). For this study, the proxy for school SES was measured by the percent of students with free or reduced lunch status as reported by the school. Teacher qualifications was measured with three variables obtained from the on-line Florida Indicators Report: average years of ex perience, advanced degree attainme nt, and teaching in certified field. Positive student learning environment was measured by six variables obtained from the on-line Florida Indicators Report: Absent 21+ Days (Students); Stability Rate; Suspensions both in-house and outof-school; and Incidents of Crime and Violence, Offenses, Student Membership (Division of Accountability, Research and Measurement, Florida Department of Education, 2007b). Content software included specially designed software that is organized to systematically deliver instruction to the student in order to teach specific concepts, sk ills, or information. As students inte ract with this curriculum-based software, their responses are analyzed by the program to determine the specific content to be presented next. Examples of these pr ograms are Integrated Learni ng Systems, tutorials, simulations, and on-line textbooks with integrated exercises and answers. Tool-based software included production software that is used to create products that communicate or present information to others (e.g., word processors, presentation programs, or videoediting programs); software used to locate information and conduct research (e.g., browsers with search engines, electronic encyclopedias, Internet archives, electronic databa ses, and virtual libraries); and software used as a cognitive tool to organize information, support problem solving, and facilitate the deeper understanding of concepts (e.g., databases, spreadsheets, graphic organizers). Office/ production software includes the traditional programs included in an office suite (e.g., word processing software, spreadsheet software, presentation software, and graphics software). Advanced production software includes more advanced editing and authoring software used to create products (e.g., multimedia authoring software; video editing software; concept mapping software; web authoring software). Florida Comprehensive Achievement Test (FCAT) is a series of standardized tests that are used to measure student achieving and school achievement progress in Florida. All students enrolled in public

PAGE 30

16 schools are required to participate in these tests from grades 3 through 10 (Florida Department of Education, 2005a). For this study, school mean scaled scores on the FCAT norm referenced tests for reading and math were used to measure reading and math achievement. The mean school rubric score for the FCAT writing was used to measure writing achievement. Florida School Indicators Report (FSIR) is an interactive on-line database with longitudinal data about school-level factors such as aggregated stude nt demographics, attendance, student conduct, teacher variables, student membership, and staff charact eristics (Division of Accountability, Research and Measurement, Florida Department of Education, 2007b). Data for variables used in the study were available through the 2005-06 school year. For this st udy, the data used for the outcome variables for absences and misconduct was obtained from this database. In addition, all of the student demographic variables and positive learning environment va riables were obtained from this database. Average Yearly Progress Reports are available on the Florida Sch ool Grades website (Division of Accountability, Research and Measurement, Florida Department of Education, 2007c). These reports provided demographic information about the school proportions of low socio-economic status, minority, and Limited English Proficiency students, as well as proportion of students with disabilities (Florida Department of Education, 2007b). Digital divide is the gap between schools that have high levels of student access to technology and high levels of instructional methods that integrate technology and schools that have low levels of student access to technology and high levels of instru ctional methods that in tegrate technology. Delimitations This study was conducted using four points of time (2003-04, 2004-05, 2005-06, and 2006-07) as the variables in the first level in the model and school-level variables at the second level. Accordingly, all student variables were aggregated and their average scores for the school were added to the school-level of the model. By using this procedure, information was lo st, and the results of the analysis can not connect the impact of integration of technology variables to the gains in individual students’ achievement. This method of analysis was chosen because Florida does not provid e public access to student-level data due to requirements for student confidentiality. Although using longitudinal student data would have been more informative, the school-level longitudinal data connected the impact of technology indicators with the

PAGE 31

17 changes in schools as measured by changes in m ean achievement scores, mean attendance scores, and changes in numbers of reported student misconduct incidents. In addition, this study only looked at variables at two levels of the model. Additional variables that may impact the outcomes, such as leadership, funds available, and technology plans were not included in the model. All variables in the model were selected because they can directly impact the learning environment of the students. Limitations The results of this study will have to be interp reted in light of the limitations as well as the delimitations. This study was conducted using existing data that were collected by the Florida Department of Education. Technology has undergone rapid change over the last three years, thus the design of the Florida’s technology surveys has been modified slightly over time. Clarification of the items, movement of the items within the survey, and variations in specific respondents may have impacted the data. Further, the degree of accuracy of these measures may be questio nable since all of the technology indicators were reported by the principal and/or a designated technology specialist. Data were not collected directly from students or teachers within each sch ool about how they used the technolog y, so the responses used may not accurately represent their views. Other contextual variables used in the model may not adequately measure the constructs. The use of the percentage of students who have free or reduced lunch status as the only proxy for socio-economic status of schools may not accurately represent this po pulation of schools. The professional qualifications of teachers that impact the teachers’ ability to weave together the dynami c variables during the teaching and learning process may not be captured by measurin g their years of experience, advanced degrees, and teaching in their field of expertise. Certainly staff development measures would be an important variable to include with this factor. However, due to the changes in the technology survey, the amount of professional development can only be measured through the variable proportion of the technology budget devoted for technology training, which may not adequately measure this construct. Other variables may have been left out of the model that impact students’ achievement. Education is a complex phenomenon, and there are many factor s and contexts that influence student achievement.

PAGE 32

18 Moreover, it is important to emphasize that this study examined relationships among predictors and school level achievement. It cannot determine causality. Educational Significance These findings add to the data available for supporting responsible decisions by educational leaders for investing in and implementing technology initiatives in Florida schools that support equitable digital opportunities. This study also supports the research community by adding to the technology integration knowledge base from longitudinal research w ith large databases. As researchers continue to add to this knowledge base about technology integration from research conducted with other states’ longitudinal data, policy makers will be able to co mpare Florida technology integration initiatives with technology initiatives in other states. As a result, co nfirmation of best practices of technology integration will be accomplished at the national level. Another outcome from this study was recommendations for revisions and new items in the survey to better measure the integration of technology in future research. The results from this study have been shared with the Bureau of Instruction and Innovation, Florida Department of Education, so it can support the dissemination of important information needed by schools for planning technology initiatives and staff development programs that support technology integration to enhance student achievement. If this information is used for responsible technology planning and the implementation of technology initiatives by sc hools and districts, it may indirectly expand the educational opportunities of over 2.67 million students in Florida public schools (Florida Department of Education, 2007a).

PAGE 33

19 Chapter 2: Literature Review This chapter begins by examining the various factors that impact student achievement through the framework of the Carroll model of school learning and contrasts this model with other proposed conceptual frameworks. Next, school reform or organizational change is discussed within the theoretical framework of complexity theory. Then the connections between school reform and standards-based education, especially as technology is used as an agent of change, are hi ghlighted. Essential conditions necessary for technology integration, along with the confounding factors at each of the multiple -levels of district, school, classroom, and student are discussed. Last, research evidence fo r the relationships between technology integration and student outcomes of achievement, attendan ce, and student conduct are reviewed. Theoretical Frameworks The design of this study is based on two theoretical Frameworks. The first framework, the Carroll Model of School Learning explains the dynamics of student achievement within the teaching and learning environment over time. The second framework, Complexity Theory, explains how organizations adapt and change over time. Carroll Model of School Learning and Student Achievement According to the Carroll model of school learning th e degree of learning is the proportion of the amount of time spent learning to the amount time of time needed to learn, which he delineates in a mathematical equation (Carroll, 1963, 1989). needed time spent actually time f learning of Degree _ _ These two categories are measured by five variab les. The numerator, time actually spent learning is determined by the interaction of the opportunity to learn and perseverance, while the denominator, time needed to learn, is determined by the interaction of the student’s aptitude, quality of instruction, and ability to understand instruction. Carroll defines opportunity to learn as the amount of time set aside for instruction of selected curricula with specific outcome goals. Pers everance is the amount of time that the student is

PAGE 34

20 willing to engage in the learning activities about the curricula. Optimal leaning occurs when these two variables match; however if the student needs additiona l learning opportunities and is willing to engage in more on-task learning activities than are offered, then the degree of learning will be decreased. When the conditions are reversed and the student does not have the perseverance to remain on-task for all of the activities, learning also will be reduced. Aptitude is a m easure of the student’s rate of learning, which is impacted by the previous learning experiences as well as individual characteristics. The amount of time needed to learn can be decreased when the quality of instruction is not optimal or when the student does not have the verbal ability to understand the instruction. On the other hand, poor quality instruction can be mediated when the student has high ability to understand the instruction or a high aptitude. Although students may have low rates of learning, they can still ma ximize their learning if they have high degrees of perseverance and are given the opportunity to learn. The expanded equation for the degree of learning includes the interaction of these five variables. d unders to ability n instructio of quality aptitude ce perseveran learn to y opportunit f learning of Degree tan _ _ _ _ The degree that a student learns is the overlap of the time spent learning, which includes the overlap of opportunity to learn and perseverance, with the time needed to learn, which includes the overlap of aptitude, quality of instruction, and ability to understand the instruction (see Figure 2). Maximum learning occurs when the interaction of all of these variables balance so that the time needed to learn is exactly the same as the time actually engaged in active on-task activities. However, optimal learning occurs when aptitude, quality of instruction, and ability to understand, exactly match with the student’s perseverance and the pace of the opp ortunity to learn (see Figure 3).

PAGE 35

21 Figure 2. Degree of Learning Carroll Model of School Learning

PAGE 36

22 Figure 3. Optimal Learning in Carroll Model of School Learning The most important concept provided by the Carroll model of school learning is the importance of time, as measured by three of the variables in the dy namic learning process. This concept changes the focus of aptitude from a limiting innate ability of the student to an enabling ability that is under the control of the student through perseverance during the amount of time required to learn. Carroll states that increasing student motivation does not increase the learning rate or the amount of learning when time is held constant (Carroll, 1989). Consequently, this model supports the delivery of equality of opportunity, but not the equality of attainment for all students. Bloom (1968, 1976, 1984) on th e other hand, proposed that learning could accelerate after students acquired the cognitive entry behaviors and a ffective entry characteristics. Once the limiting condition of not having the prerequisite skills, knowl edge, and attitude has been corrected, students can accelerate their achievement so that 80 – 90% of students can attain what is usually realized by only 20% of students (Guskey, 2001). In order to accelerate learning, students must be actively engaged in appropriate levels of instruction with embedded formative assessm ent, feedback, corrective activities, and reassessment that is aligned with the skills taug ht. Bloom proposed that acceleration was possible because differences in entry level cognitive sk ills accounted for 50% of the variance in school achi evement (1968, 1976), while

PAGE 37

23 differences in affective characteristics accounted for 25% of the variance (1968 1976); however, because they are correlated, 65% of the variance in school ac hievement can be accounted for with both (1984). The addition of quality instruction could explain another 25% of the variance for a total of 90% of student achievement (Bloom, 1976; Guskey, 2001). By providing multiple high quality instructional methods and activities, students could attain improved achievement of two standard deviations in a group setting (Bloom, 1984). Slavin (1987, 1994) focused on aspects of l earning that the teacher controls in his Quality, Appropriateness, Incentive, time model (QAIT). When overlaying the QAIT on Carroll’s model of school learning, Slavin delineated additional tasks for the teacher that support factors Carroll attributed to the learner (see Figure 4). For example, Slavin added that the teacher provides incen tives in order to promote student motivation or perseverance, and the teacher deliv ers instruction at the appropriate level in order to assure that students have the ability to understand the instruction. For maximum learning, all factors in the model must be present, as each can be the bottle neck that limits learning. Slavin proposes that improvements in all QAIT factors will yield greater achievement than improvements in only one because each factor has a ceiling.

PAGE 38

24 Figure 4. Quality, Appropriateness, Incen tive, and Time (QAIT) Model of instructional effectiveness relating alterable elements of instruction to student achievement (Slavin, 1987). As models become more complex, they include more levels of the organization. The Carroll model of school learning delineated two levels, student and teac her. Although Slavin also suggests two levels, he puts students and teachers in the cla ss level, and then adds school as the next level. Marzano (2003) organizes the factors that impact student achievement into three levels – student, teacher, and school – and situates these levels within the school district level. According to Marzano, there are three factors at the student level: home atmosphere, learned intelligence and background knowledge, and motivation; three factors at the teacher level: instructional strategi es, classroom management, and classroom curriculum design; and five factors at the school level: guaranteed and viable curriculum, challenging goals and

PAGE 39

25 effective feedback, parent and comm unity involvement, safe and orderly environment, and collegiality and professionalism (p. 10). When examining the impacts of factors at various levels, Marzano reports that 13% of the variance in student achievement is from the teachers’ activities and 7% is from the factors at the school level. The opportunity to learn was the most important factor that impacts student achievement and time was the second most important factor. Both of th ese factors are grouped together into a guaranteed and viable curriculum school-level f actor. Taken together, sc hool and teacher effectiv eness have an immense impact on student learning. When students entered the school at the 50th percentile of achievement and have participated in two years instruction at that school, students in the least effective school with the least effective teacher had achievements at the 3rd percentile, while students in the most effective school with the most effective teacher had achievements at the 96th percentile (Marzano, 2003). Another important contribution of the Carroll model of school learning is the ability to use student perseverance to measure the student’s motivation to learn. Marzano (2003) explains motivation as the reason that students do things. This creates the link between students’ affective attributes and their activities at school. Students have direct control over their achievement by the duration of the time they spend attending to the instruction (Berliner, 1990). Wh en students find activities interesting, they are more likely to participate for longer periods of time. Thus the most motivating activities would be long term projects that students are passionate about and find personally meaningful (Marzano, 2003). Ringstaff and Kelley (2002) reported that research has found that when students use technology to learn with student centered project-based methods, their attitudes, self -confidence, attendance, an d time-on-task increased. Increased perseverance could also explain the relationship between academic performance and student conduct and attendance. As students increase their pers everance or motivation, they decrease the amount of time spent in off-task behaviors or misconducts. Increased motivation to engage in personally meaningful learning activities could also lead to improved attend ance. Perseverance requires that the student have the opportunity to learn meaningful material and the time to spend learning it. Instructional time becomes the most critical variab le that impacts student achievement (Berliner, 1990; Bloom, 1968, 1976, 1984; Caroll, 1963, 1989; Marzano, 2003; Slavin, 1987, 1994). Berliner (1990) defines the multi-faceted components of time as allo tted time, engaged time, time-on-task, and academic learning time (ALT). ALT is the amount of allocated time that the student is engaged in time-on-task with

PAGE 40

26 activities that are aligned with the curriculum and the evaluation instruments used to measure the learning. ALT is directly related to the amount of learning and achievement. ALT, and as a result achievement, is adjusted by transition time, wait time, success rate, ap titude, perseverance, and pace. Berliner used success rate as a measure for the quality of instruction and ability to understand factors in the Carroll Model of School Learning. High success rate was above 70%, wh ile low success rate was below 30%. Nevertheless, achievement may not be demonstrated if the instrument used to measure learning does not align with the activities and curriculum (Berliner, 1990; Russell & Higgins, 2003; Wenglinsky, 2005). Given that learning requires time (Berliner, 1990; Bloom, 1968, 1976, 1984; Carroll, 1963, 1989; Marzano, 2003; Slavin, 1987, 1994), multiple measurements of student achievement need to be conducted over time. School level achievement is the mean of all student achievement w ithin a school. Meaningful change in school level achievement requires time for documentation. Thus, research examining the changes in school level achievement must be longitudinal. Complexity Theory and Organizational Change Within the business sector, complexity theory has been used to explain the functioning of organizations. Utilizing complexity theory, organizations are viewed holistically as systems that have independent agents or elements orga nized in structures and nested at different levels (Caldwell, 2005; McElroy, 2000). These agents and levels interact, b ecome interdependent, and produce collective behavior, as the organization evolves and adapts to achieve its purpose (Holland, 2006; O’Day, 2002; Wilensky & Resnick, 1999). Systems have boundaries that separate them from their outside environment. Information is exchanged through feedback loops among the elements within the organization and with the environment outside the organization (Caldwell, 2005; McElroy, 2000; Morrison, 2002). This exchange of information is essential to the adaptation of the organization (O’Day, 2002; Wheatley, 1999). The degree to which information is exchanged among the elements and the outside environment delineates whether the system is open or closed (Caldwell, 2005; Wheatley, 1999). Previous researchers have noted th e prevalence of nested contextual factors in educational settings and the difficulty these cause for fi nding answers to research questions about the impact of instruction on student outcomes. Bronfenbrenner (1976) proposed that the ecological structure of the educational environment consists of many nested and interacting levels, all of which have impact on how children

PAGE 41

27 learn. Weick (1976) suggested that school districts and schools work well because they are loosely coupled organizations within these nested levels. That is, organizational change is influenced on both the micro level by students and teacher s, and on the macro level by schools an d districts. Thus, school systems are usually very stable organizations, and change occurs very slowly, if at all (Cuban, 1986, 1998, 2001; O’Day, 2002; Tyack & Cuban, 1995). When change occurs within an organization, it is not always smooth and linear (Caldwell, 2005; Jacobson & Wilensky, 2006). Time is a differentiating aspect of the change process at the macro and micro levels (Weick & Quinn, 1999). At the macro level, th e change response can be episodic and non-linear as when an organization responds to a specific event (e.g., No Child Left Behind (NCLB) laws for accountability or the acquisitio n of new technology), while at the micro level, change is usually continuous, such as the teachers’ response to the access to new technology resources. The key component included in complexity theory is the impact of the dynamic flow of information between agents and levels of the organi zation on the organization’s ability to adapt (O’Day, 2002; Wheatley, 1999). Without information, the organization stagnates and cannot change. If information flows freely, then new important information can be received and utilized to improve organizational functioning. All agents within the or ganization need access to the new info rmation, an understanding of the goals of the organization, and iterative information abou t the results of the organization’s responses for the organization as a whole to successfully adapt (Wheatley, 1999). Complexity Theory, explains how schools adapt and change in response to information obtained from outside the school as well as information obtained from inside the school. On the macro level districts and schools receive information from the state and federal government in the form of legislation, from Universities and Research Centers in the form of reports and recommendation, and from the local community in the form of resources and requests. Th is information is translated by agents in the organization into curriculum and resources supplied to teachers and students. The Carroll Model of School Learning explains how at the microlevel the dynamics of the teaching an d learning process support student achievement. At all organizational levels the examination of the continuous flow of information that triggers the responses, adaptations, and changes in the dynamic processes of the organization must be conducted over time.

PAGE 42

28 Changes in School Organization and Instruction The accountability movement has resulted in the delineation of new standards in each curriculum area. This information from outside of the organization has pressured school organizations to change. The responses by some schools systems in order to m eet the new standards and accountability requirements have resulted in school reform. The integration of technology has been an important component in this change process. Accountability and Standards The goal of the accountability movement is to pr ovide the crucial information needed for the school organization to adapt and improve student achi evement. This professional accountability is achieved not only through communication of standards from within the professional community, but also from outside the boundaries of the specific organization. Additional information is provided by the results of state assessments. Several conditions are necessary for the accountability syst ems to support improved instruction and learning: (1) principals and teachers must have access to the right amount of accurate and valid information; (2) they must have the motivation to use the information; (3) they must know how to interpret the information; (4) they must have the resources needed to implement the changes; and (5) all teachers and administrators must share information abou t instructional process and student learning as well as share responsibility for student outcomes (O’Day, 2002). Professional organizations in all content areas ha ve made recommendations about what should be taught in all subject areas as well as how the content should be taught. The first standards were for mathematics by the National Council of Teachers of Mathematics (1989) followed by other organizations (English and Language Arts Standards by the National Council of Teachers of English (NCTE) and the International Reading Association (IRA) in 1994; social studies standards by the National Council for the Social Studies (NCSS) in 1994; science standards by the National Committee on Science Education Standards, and Assessment & National Research Council (NSES) in 1996, and even the position statement on Technology and Young Children—Ages 3 through 8 by the National Association for the Education of Young Children (NAEYC) in 1998).

PAGE 43

29 Embedded in these standards for all subject ar eas are recommendations and implications for technology. For example, the following recommendations for technology have been offered by professional organizations: Students use a variety of technological and information resources (e.g., libraries, databases, computer networks, video) to gather and synthesize information and to create and communicate knowledge (NCTE & IRA, 2006). Technology used to gather data enhances accuracy and allows scientists to analyze and quantify results of investigations (NSES, 1995, p. 148). Technological advances connect students at all levels to the world beyond their personal locations (NCSS, 1994, Thematic Strands III). Appropriate technology is integrated into the re gular learning environmen t and used as one of many options to support children’s learning (NAEYC, 1998, p.2). Technology is an important, integral, and integrated component in all domains of learning. As a result, the International Society for Technology in Education (ISTE) developed the National Education Technology Standards (NETS) for students in 1998, for teachers in 2000, and for administrators in 2001 (ISTE NETS Project, 2005a, 2005b, 2007). Since then, individual states have used the recommendations from these professional organizations to write their curriculum standards in each subject. To date all states except Iowa have adopted, adapted, aligned or refere nced the NETS in the stan dards or curriculum that they have set for accountability (ISTE NETS Project 2005c). Indeed, Florida even provides on-line supports at Sunshine Connections ( http://www.sunshineconnections.org/home.htm ) for teachers to use for developing curricula that meets these standards. School Reform In order for schools to meet th ese state standards of accountability, school-wide reform programs have been initiated to raise the academic standards for a ll children, especially those who are at risk due to high levels of poverty. Successful school reform programs leave the process of school change to the schools as they adapt to meet the state standa rds (Borman, Hewes, Overman, & Brown, 2003). Philosophical models that usually include new instructional methods with specific curricular materials and ongoing professional development underpin successful reforms. Another key com ponent is reform support by both the teachers and administrators. Berends, Kirby, Naftel, and McKelvey (2001) conducted a longitudinal study with observation points in 1997 and 1998 using multi-level modeling statistical techniques to examine the effects of teacher, school, and design-team f actors on implementation of school

PAGE 44

30 reforms and student performance. They found the fo llowing factors related to successful implementation and student achievement: (1 ) supportive district environment; (2) t eacher support of clearly communicated reform design plans; (3) strong principal leadership; and (4) adequate resources. Of note, findings from the meta-analysis of comprehensive school reform and student achievement conducted by Borman et al. (2003) demonstrated that the impacts of school reform benef ited all schools regardless of their poverty levels and that the strongest effects were achieved after the fifth year of implementation. Therefore, other important factors in successful reform implementation in clude the duration and fidelity of the program. Tyack and Cuban (1995) attribute the success and failure of school reforms to the teachers’ and the public’s ideas of what constitute “real school.” Reform movements of ten fail or are short-lived because they counter beliefs about purposes of education and me thods of instruction. For reforms to be successful, schools must enlist the support and dialogue of the community. This needed exchange of information between the organization and the outside environment is explained by complexity theory (Morrison, 2002; O’Day, 2002; Wheatley, 1999). However, since schools are loosely coupled organizations, within the complex school environment, teachers are the key to the adoption of reform s (Cuban & Tyack, 1996). Teachers may not embrace changes unless they see th at the implementation will make their job of educating students more efficient and productive. Thus, adaptations of the school organization occur gradually as teachers alter and adapt reforms in orde r to improve the teaching an d learning process while maintaining the basic stru cture of “real school.” Accountability and standards influence school chan ge from outside of the organization, while teachers influence school change from within the or ganization. The pressures and information flow from without and the resistance and acce ptance from within allow the scho ol organization to adapt while maintaining its integrity and common purpose. The scho ol change process requir es time and free flowing communication of relevant information. Technology can be a key element in this process. Essential Conditions for Integration of Technology Initiatives ISTE has delineated a list of ten essential cond itions for the successful implementation of the NETS (ISTE NETS Project, 2005a). This list can be ma pped to the multi-level factors that interact in a complex school organization. At the boundary of information flow between the agents within the organization and the outside environment are Community Support and External Conditions. At the school

PAGE 45

31 or district level are Shared Visi on and Support Policies. At the teach er or classroom level are Skilled Personnel, Professional Development, Technical Assistance, Content Standards and Curriculum Resources, and Student-Centered Teaching. At the student level are Equitable Access and Assessment and Accountability. Technology impacts the successful implementation of new technology integration initiatives at each of these levels. District and Scho ol Level Factors Communication. Adequate communication is essential between the school system and the local community and other external agencies, such as the stat e and federal departments of education, as well as professional organizations. These communications provide the feedback loops that support organizational adaptation. Technology has impacted the access of information by individuals and the flow of information between the local community and outside agencies. Information about federal, state, and professional organization funding for technology initiatives is disseminated through the Internet (e.g., Grants.gov at http://www.grants.gov/index.jsp and Bureau of Grants Management at http://www.firn.edu/doe/grants/grantsdev/compgrants/cgmain.htm ). In fact, the Federal Funding Accountability and Transparency Act of 2006 (PL 109-282) established a ‘‘searchable website’’ with free access by the public that lists information about all federal financial assistance awarded that is over $25,000. The Internet has also impacted the dissemination of information, data, and resources for accountability standards (e.g., ISTE NETS Project at http://www.iste.org/inhouse/nets/cnets/index.html ). These sponsored initiatives guide the implementation and direction of school reform. In turn, the data collection methods and analyses involved in follow-up evaluation of schools’ technology integration programs have also been impacted by technology. Web surveys are conducted on-line, data is stored digitally, and the analyses of results are conducted using statistical software (e.g., STAR Survey at http://www.flinnovates.org/survey/ ). Schools communicate with parents and the community through school websites and e-mail (Bureau of Instruction and Innovation, Florida Department of Education, 2007a). Leadership and vision. Leadership at the school or district level is essential for creating a shared vision and developing support policies for technolo gy integration. To facilitate change, communication must be eased so that administrators, teachers, an d students mainta in active involvement. Technologies such as e-mail, listservs, websites, and wikis pr ovide the vehicle for disseminating timely, accurate,

PAGE 46

32 distributed, information needed for collective vision (Morrison, 2002). The integration of technology throughout a school system as a school reform must be consistent with the district’s overall educational mission, vision, and strategic plan (TSSA Collaborative, 2001). The School Technology Plan form ally articulates the district vision and blueprint for the technology integration initiative. An effective plan cons iders the needs of agents at all levels within the school and provides resources and supports to adequately meet these needs. Involving all stakeholders in creating the technology plan affords the best chance for a successful technology integration initiative (Barnett, 2001; Fulton, Glenn & Valdez, 2004). Anderson and Dexter (2001) identified six important categories of decisions that are ma de during the planning process and specifically delineated in the technology plan: strategic planning and goal setting, budgeting and spending, organization, curriculum, evaluation, and external relations. They also recomm end that for best results the school must become a learning organization with distributed leadership. In fact, teachers who are more professionally involved in sharing instructional practices are more likely to use and have students use computers (Becker, 2001). Imperative for the success of the technology integration initiative is on-going funding support for infrastructure, hardware and software upgrades, t echnology support personnel, and staff development (Anderson & Becker, 2001; Fulton et al., 2004). Indeed, one study found that schools with the highest levels of software investment ove r five years had the greatest propor tion of teachers assigning computer work in class and students using computers (Anderson & Becker, 2001), and districts in Texas that spent the most on hardware and software had the highest po sitive correlation with average student tests scores (Christensen, Griffin, & Knezek, 2001). School-level differences. The integration of technology has been implemented differently for various levels of schools. Findings from the Technology Integration in Education Initiative Statewide Survey Report (2002) indicated th at in Texas the most frequent location for students to engage in technology activities was different depending on the level of school. Most frequent locations in Texas were computer labs in middle schools (51%) and technology classes in high school s (67%), while Chicago’s public schools had more computers in the classroom at the elementary level (77%) than the high school (52%) (Hart, Allensworth, Lauen, & Gladden, 2002).

PAGE 47

33 The results from a survey of t eachers in a large school distri ct conducted by Barron, Kemker, Harmes, and Kalaydjian (2003) demonstrated that com puters are used differently in elementary, middle, and high schools. Computers in elementary schools (29%) were used significantly more often for problemsolving than in middle (23%) and high schools (20%). Computers used for a communication tool varied by school level with most use in elementary (59%), then middle (54%), and finally high schools (48%). In contrast, computers were used significantly more often as research tools in high schools (40%) than elementary (32%) or middle schools (40%). Wenglinsky (1998, 2005) also found school level differences in the way students used computers for math instruction. At the fourth grade level, students used computers more often for learning games (54.5%) than drill and practice (35.9%) or simulations and applications (7.5%), while eighth grade students used computers mo re often for drill and practice (34.3%) than learning games (29.2%) or applications (27.2%). Classroom Level Factors Teacher. At the classroom level or teacher level, the teacher has the greatest impact on the implementation of any school reform, including the integration of technology (Cuban, 1998; Tyack & Cuban, 1995). First, teachers impact the classroom learning environmen t and student achie vement through their primary responsibilities of coordinating instructional strategies, classroom management, and classroom curriculum design (Marzano, 2003). Second, while they implement the instruction of their curriculum with integrated technology, they facilitate students’ acquisition of the NETS standards (Barron et al., 2003). Knezek, Christensen, and Fluke (2003) identified two teacher factors, will and skill, along with access to technology that impact t echnology integration th rough structured equa tion modeling analysis. O’Dwyer and colleagues (2004, 2005) used multi-leve l modeling statistical analysis and found similar teacher factors that were related to the success of techno logy initiatives: teacher sk ill, comfort level, and perceived importance of technology. Teachers’ technology skills in using software and computers were positively related to the extent that they used computers professionally and the extent to which they had their students use computers for production of products and analyzing information (Becker, Ravitz, & Wong, 199 9). In fact, the teachers’ use of multi-media production software was positively related to the variety of ways that students used

PAGE 48

34 computers (Becker, 2001). Thus, in order to have the most impact on student achievement by technology, not only must teachers have skills in using technology for productivity an d integrating tech nology into the curriculum, but they must also have expertise in using instructional strategies, classroom management, and classroom curriculum design. Equally important for effective integration of technology, teachers must believe that technology is an important element in the instructional design, and they must desire to integrate it into their daily lessons. Professional development. Professional development can have a positive impact on the teacher level factors that impact student achievement: instructional strategies, classroom management, curriculum design, integration of technology, and technology skills. Due to the rapid advances in technology, professional development for technology integration needs to be ongoing for both new and experienced teachers. However, novice and experienced teach ers may have different needs for professional development regarding technology integration. Alth ough newly graduated teachers may be proficient technology users, they may not be skilled in the clas sroom management necessary for effective delivery of technology integrated lessons for students in the cl assroom environment. In addition, many new teachers have not experienced effective mode ling of technology integration in their teacher preparation (Benner, Shapley, Heikes, & Pieper, 2002). Many experienced t eachers have had no formal training in technology integration. For experienced teachers, technology skills may be either se lf-taught or acquired through staff development programs. Indeed, 93% of teachers repo rted that they learned about using technology independently (Smerdon, Cronen, Lanahan, Anderson, Iannotti, & Angeles, 2000). Teachers progress through several stages of inst ructional and technological evolution as they become expert integrators of technology (Apple Computer, Inc., 1995; Coughlin & Lemke, 1999; Dwyer et al., 1990). During the initial level of technology inte gration, as they learn technology production skills, teachers become aware of the possibilities of technol ogy for improved achievement. Given that many teachers do not feel confident in in tegrating technology into their daily instructional routines, impacting teachers’ perceptions about the usefulness and desira bility of integrating tech nology as well as their comfort level in using technology is vital during this stage (Donnelly, Dove, & Tiffany-Morales, 2002). Smerdon and colleagues (2 000) reported that only one thir d of teachers in their nationally representative study felt well-prepared to use technology for instruction. Two year s later, researchers in

PAGE 49

35 another nationally representative study found that 84% of teachers felt that they had the technical skills necessary to be at least somewhat prepared to use co mputers and the Internet fo r instruction, and nearly 92% of teachers had taught some activities that us ed technology (Adelman, Donnelly, Dove, TiffanyMorales, Wayne, & Zucker, 2002). The professional development needs of these teachers had progressed from becoming aware of the potential of technology to the next stage of integration, how to integrate technology into the curriculum. However in 2000, only 67% of the teachers report ed that they had oppor tunities for follow-up activities or advanced training (Smerdon et al., 2000) In 2002, 88% of teacher s expressed a medium or high level of need for professional development about the integration of technology into the curriculum, and 89% wanted to see demonstrations of these type s of classroom activities (Adelman et al., 2002). Indeed, O’Dwyer and colleagues (2004, 2005) found that having a variety of professional development opportunities about technology, especially when focuse d on the integration of technology, was significantly and positively related to having students use technology during class time. Teachers who participate in more professional development activ ities for longer periods of time are more likely to use technology in their instru ction (Adelman et al., 2002; Smerdon et al., 2000). Adelman et al. found that formal professional devel opment that included more key features had greater impact for increasing the extent to which teachers instruct with technology. The top ranked key features of formal training for teache rs included teaching at the appropriate skill-level of the teach er, opportunities for meaningful engagement with colleag ues and materials, and input from teachers in the district in the preparation and delivery of multiple sessions that oc cur over substantial time. Professional development that increased teachers’ instructiona l use of technology focused on inte grating technology into instruction that was directly related to the content areas taught. The key feature identified that was lacking in formal training was follow-up planning time to implement new practices. Although professional development increased teachers’ use of tec hnology, teachers also reported significant barriers to integrating technology activities into instruction for students. The greatest barriers were lack of release time to learn how to use tech nology (82%), not having enough computers in their classrooms (78%), and not having enough instructiona l time available to incorporate technology activities (80%) (Smerdon et al., 2000). Adelman et al. (2002) confirmed similar barriers of lack of time for

PAGE 50

36 practicing technology skills, developing lessons, and sche duling student activities, as well as availability of too few computers. Access to computers. Having access to enough resources is also important for the successful integration of technology. In a national study conducted in 1999, researchers found that approximately half of the teachers with Internet access used computer s for classroom instruction (Smerdon et al., 2000). O’Dwyer et al. (2004, 2005) found that the availability of technology was significantly positively related to teachers’ professional use of technology and their st udents’ use of technology during class time. Teachers reported that having too few computers was a barrier to integrating technology (Adelman et al., 2002). The availability of computers has increased over time. Adelman et al. (2002) reported that 47% of teachers had between 2 to 5 computers within their classroom. On the national level, the student/instructional computer ratio had decreased from 12:1 in 1999 to 4:1 in 2003 (Parsad & Jones, 2005). By 2006, there were 3.5 students per computer nationally and 3.8 students per computer in Florida (Technology Counts, 2006). Location is an important factor that impacts th e frequency that teachers use computers with their students for instruction. Fo r example, Becker (2001) reported that secondary teachers who had access to 5 to 8 computers within their classroo m reported that students frequently used computers during class twice as often as teachers who used computer labs. In 199 8, 62% of secondary teacher s with one computer per four students in their classroom used computers freque ntly with their students for instruction, while only 18% of teachers who used computer labs frequently used computers with their students (Becker et al., 1999). Mann et al. (1999) reported that teachers, who ha d computers in their classrooms rather than in computer labs, spent more time using computers for reading, math, and writing instruction. Indeed, the students who had access to the computers in their cla ssroom had greater achievement gains than students who had access to the computers in labs. Similar results were found in 2001, 77 % of teachers with one computer per four students in their classroom used computers frequently for instruction, while only 21% of teachers with no computers in their classroom had their students freque ntly use computers for instruction (Adelman et al., 2002). Moreover, r ecent studies about the impact of on e-to-one access to laptop computers by teachers and students have report ed significant increases in how often teachers use computers with their

PAGE 51

37 students (Bebell, 2005; Lowther et al., 2003; Shapley, Sheehan, Sturges, Caranias-Walker, Huntsberger, & Maloney, 2006; Silvernail & Lane, 2004). Access to software. Teachers can use software for variou s educational purposes in their instructional activities. To get the greatest return, technology must be a necessary component of the lesson (Donnelly et al., 2002). In 1999, the software most often used by secondary students was word processing; in contrast, elementary students most often used dr ill and practice software. Neve rtheless, word processing software was used by over 50% of students in grades 4 – 12 (Becker et al., 1999). Smerdon et al. (2000) reported that within classrooms, students used com puters for tool-based instruction with spreadsheets and word processors (61%), solving problems and analyzing data (50%), creating multimedia projects (45%); researching on the Internet (51%), practicing drills (50%), and demonstrations/ simulations (39%). A similar ranking of instructional activities was found by Adelman et al. in 2001: writing with word processors (77%), researching on the Internet (70%), learning computer skills (70%), as a reward or for free time (62%), and practicing drills (60%). Technical and technology integration support. Once teachers have access to the computers needed for instruction, they must be able to count on having the support necessary to utilize them with students during the lesson. Breakdowns can occur when the hardware and network do not function or when the teacher is not proficient with the instructional methods for integrating the softwa re into the lesson. Thus, teachers require two types of support in order to utilize technology for instruction: technical support and instructional technology support. Staff development is often entwined with support, as the technology specialist in the school often performs both roles (Donnelly et al., 2002). Ronnkvist, Dexter, and Anderson (2000) found from their 1998 nationally representative survey that 87% of schools had someone who served as a technology coordinator, but only 19% performed this function full-time in the sc hool. Almost half of these schools had technology coordinators who were also classroom teachers. By 2001, 38% of schools had a paid full-time technology coordinator (Adelman et al., 2002). Although the increase in full-time technology coordinators is important, the support that they supply to individual teachers is small. On av erage, full-time coordinators spent 22.8 minutes per teacher each week maintaining the functioni ng of hardware and softwa re and 22.1 minutes per teacher each week supporting staff de velopment, while on average part-tim e coordinators spent 8.4 minutes

PAGE 52

38 per teacher each week maintaining the infrastructure and 10.8 minutes per teach er each week supporting staff development. Of note, 41% of teachers believed that both instru ctional and technical support were available only to them sometim es (Adelman et al., 2002). By 2001, 97% of teachers reported having technical support, while 83% reported they had support for integrating technology into thei r instruction (Adelman et al., 2002). However, only 73% of the teachers reported that their technical needs were supported fairly to extremely well, and 50% had their support needs for integration of technology met fairly to extremel y well. Dexter, Anderson, & Ronnkvist (2002) further analyzed their results by includ ing indicators for the quality of th eir technology support. Access to resources and professional development that focused on the integration of technology had the greatest impact on teachers’ perceptions of the quality of their support. The frequency that teachers used technology with their students was positively impacted by the teachers’ perceptions about the availability and quality of the support. Student-Level Factors Demographic and personal characteristics of students that impact the outcomes of the individual student are student level factors. Examples of stud ent level factors that have been found to impact individual achievement are gender, ethnicity, socio-ec onomic status, disability, and English as a second language. In addition, students’ abilities, attitudes, motivation, and home environment have been found to impact achievement (Marzano, 2004). These individual variables can also be aggregated to the next level. For example, several students’ misbehavior can impact the class l earning environment and thus imp act class level achievement. The number of special education, limited English proficient students, or gifted students can make increased demands on the teacher’s time. These students may need special supports and accommodations to be successful. When the proportion of students needing acc ommodations is high, the de sign of the curriculum and depth of coverage may change. Fewer opportunities to engage deeply with the curriculum may impact the overall achievement level of a class. Socio-economic status. Aggregated student level factors can have even broader impact on both predictors and outcomes. Research studies have found that the proportion of students who are economically disadvantaged can impact school level variables. Adelman et al. (2002) found that students in low socio-

PAGE 53

39 economic schools have less access to the most modern computers and Internet in school and decreased access to computers at home. Conversely, as a result of federal funding progra ms, Anderson and Becker (2001) found little difference in the initial infrastructure of computer hardware and software between schools with high and low proportions of students elig ible for Title 1 funding. Similar findings were found by Benner et al. (2002) in a large scale study about technology integration in Texas schools. Schools that served the most economically disadvantaged students and had the greatest Technology Literacy Challenge Funding had the greatest number of classroom Internet connections. Indeed, over th e past five years, Texas schools with the greatest number of students at poverty level made the greatest gains in technology resources. Wenglinsky (1998, 2005) also found that the gap between high and low economic students for access and use of technology for math in struction had been eliminated by the time of his study that used secondary data collected by National Assessment of Educational Progress (NAEP) in mathematics in 1996. However, when recent expenditures are examined, differences arise. At the school level, the socioeconomic status of a school, as determined by the community in which the school is located, impacts the amount that the school spends on technology, especi ally for hardware and support (Anderson & Becker, 2001). Schools in economically disadvantaged areas spend less than half the amount that high income area schools spend on additional hardware and on-going support. Teachers in high income schools had more resources available to them and were more likely to attend professional development sessions on technology (Ronnkvist et al., 2 000), while teachers in low econom ic schools had the least access to technology support and training (B enner et al., 2002; Wenglinsky, 1998 2005). As a result, teachers may have different training needs based on the economic status of their school s. Teachers in low socioeconomic schools needed more training in basic technology skills, while teachers in high socio-economic schools needed training about the integration of technol ogy (Benner et al., 2002). In addition, Anderson and Becker (2001) found that the school’s degree of investment in hardware, software, and technology supports were positively related to the frequency that teacher s assigned students to com puter work. Consequently, students in low socio-economic schools had fewer technology skills when compared with those in high socio-economic schools (Adelman et al., 2002). On the other hand, low socio-eco nomic status does not always negatively impact a school. Adelman et al. (2002) foun d that teachers in high-poverty schools that participated in the Technology

PAGE 54

40 Literacy Challenge Fund program were significantly more likely to receive incentives for participating in professional development. These grants encouraged 30% of the technology fu nds to be allocated to professional development. Additionally, level of poverty was not a statistically significant predictor of the impact of school reform programs (Borman et al. 2003). Schools with the greatest levels of poverty were just as likely to have positive results from their school reforms as all other schools. Moreover, Benner et al. (2002) found that students in economically disadva ntaged schools participated in more technology activities than higher socio-economic schools. However, Becker (2001) found that the educational experiences of students in low economic schools were different from those in high economic schools. Using logistic regression, Becker found an interaction between ability, school level, and sc hool socio-economic status. He concluded that economically disadvantaged students used computers more often for learning basic skills. Adelman and colleagues (2002) confirmed this conclusion in their integrated studies of educational technology. They found two significant differences in high frequency computer use between high-poverty and low-poverty schools. Students in low-poverty schools more often used computers for drills (42% vs. 25%) and for freetime (42% vs. 26%). Wenglinsky (1998) found that poor eighth grade students were less likely to use computers for simulations or applications than non-poor students (22% vs. 33%) and were more likely to use computers for the lower-order thinking skills involv ed in drill and practice (34%). He found that this trend had not changed in 2000 (Wenglinsky, 2005). There are many factors that facilitate the integr ation of technology at both the macro and micro levels. Factors at the macro or district and school levels include communication, leadership and vision, and school level differences. Factors at the classroom level include the teacher, professional development, access to computers and software, and support for techno logy integration. At the mi cro-level or the student level, student demographics and personal characteris tics, especially socio-eco nomic status impacts the integration of technology into the daily instructional routine. Research Evidence for the Relationship of Technology and Student Outcomes Student Achievement Outcomes How technology is used has had an impact on studies measuring its effectiveness. Interestingly, even when researchers looked at how technology was utilized by the same programs, they did not always

PAGE 55

41 draw the same conclusions. Borman et al. (2003) recommended the Co-NECT program, which includes technology integration in its model, as having the greatest need for additional research because of its potential. This is the same program that Oppenheimer (2003) observed when he visited a high school in Massachusetts and concluded that the technology used by the students distr acted them from deeper understanding of subject matter, so that they performed at a similar level as middle school students. Conversely, when Weng linsky (2005) visited the same school, he found that the quality of the students’ work was similar to that of advanc ed placement students and that the students demonstrated sophisticated problem solving skills. Berends et al. (2001) reported that six out of 18 schools that implemented the CoNECT program had gains in reading test scores and ten out of 17 schools had gains in math test scores, when compared with the average of the other schools in the district. In addition, the standardized tests that were used to measure achievement did not assess the de eper level of knowledge and skills that students had gained from using the technology (Russell & Hi ggins, 2003; Wenglinsky, 2005). These various interpretations about the same program provide some insight into the disparity of the results that have been reported on the effectiveness of technology on student achievement. Moreover, even though individual researchers have used the same data and drawn the same overall conclusions, they have disagreed on methods, variables, and specific outcomes. For example, both Wenglinsky (2004, 2006) and Lubienski (2006) agreed that student achievement is impacted by the socioeconomic status of the student and school in their studies that were conducted using secondary data from NAEP in 2000 to examine the differences in math ematics achievement by ethnicity. However, they disagreed on the methods for conducting the multi-level analysis. Specifically, they differed on the level of alpha needed to control for Type I errors, how many individual variables to include, whether to use individual variables or composite variables in the multilevel models, as well as the appropriate level of the model for inclusion of these variables. As a result, Wenglinsky (2004, 2006) and Lubienski (2006) came to different conclusions about the relationships of various teaching methods with achievement of black, Hispanic, and white students. Ultimately, they both concluded that experimental research is needed to confirm the best teaching practices to use with minority students. Meta-analysis. The many problems that exist with experimental research on the impact of technology on student achievement ar e illustrated by the descriptions of the methods used in meta-analysis

PAGE 56

42 research. Meta-analysis is a research method that combines the results of many experimental research studies conducted over time to determine the relative im pact of various factors th at are associated with student outcomes. Borman et al. (2003) used a total of 232 studies that met specific exclusion criteria in their metaanalysis on the relationship of comprehensive school reform and student achievement. However, when the requirements for inclusion were more restrictive and th e studies were filtered so that they used a control group, the number of studi es decreased to 145. When the criteria al so required that the study was conducted by an independent evaluator and used a control group, the number of studies used in the analyses decreased to 109. The researchers reported that many studies did not even meet the standards for initial eligibility because they did not include inform ation needed for computing effect size, and many did not report the sample size used. The researchers lost 53% of the initial studies that they found when more stringent quality criteria were required for the analysis. Borman et al. (2003) used the initial inclusion criteria to restrict the studies selected to whole school-wide reform programs conducted by entities outside the school with at least 10 different evaluation studies. This resulted in 33 models identified. Additio nal inclusion criteria restricted the studies to those with reports on outcome measures of student achieve ment necessary for comput ation of effect sizes; experimental, quasi-experimental, or prepoststudy design; and students who were in regular education in the U.S. and were not duplicated in other studies. Reports from 232 studies, obtained from ERIC and PsychLit databases, Google searches, and requests to the developers were used in their analyses. The overall effect size of the 232 studies and 1111 indepe ndent observations had a mean effect size of 0.15 ( Z = 33.26, p < .01); however, when the requirements for inclusion specified that the study was conducted by an independent evaluator and used a control group, the number of studies used in the analyses decreased to 109 and the number of independent observations decreased to 461 with a mean effect size of 0.09 ( Z = 10.59, p < .01) a difference in effect size of 0.06 standard deviations. Results from 1,017 independent samples indicated a mean effect size of 0.13 ( Z = 10.81, p < .001) for reading achievement from comprehensive school reform, while the mean effect size of math achievement from 679 independent samples for comprehensive school reform was 0.15 ( Z = 9.86, p < .001). However, school reform was the

PAGE 57

43 focus of these meta-analyses, and technology was only one of the components of some of the school reforms. Meta-analysis has been used to look specifically at the effect of instructional technology on student outcomes. Waxman, Lin, and Michko (2003) conducted a meta-analysis on the effectiveness of teaching and learning with technolo gy on student outcomes. Out of an initial 200 articles, only 42 were included in the analyses, a loss of 79% of the research studies. They used ERIC databases, Google and Metacrawler search engines, and reference lists of artic les in specific educational journals to find articles published between 1997 and 2003. Inclusion criteria included: experimental, quasi-experimental, or preand post-test design that was published in refereed j ournals; focus on K-12 classrooms that meet over 50% of the time face-to-face; and studies wherein the control group did no t have access to computers and the statistics necessary to calculate effect sizes for all groups were reported. When there was more than one comparison in a study, each was weighted in inverse proportion to the total number of comparisons in the study. Studies with multiple outcomes were included when statistics were available to calculate effect sizes yielding a total of 282 effect sizes. Analysis of th e twenty-nine articles that reported student cognitive outcomes had a mean study-weighted effect size of .448 (p<.001) with 95% confidence intervals that did not include zero (Waxman et al., 2003). However, Waxman et al. did not disaggregate the effect size of the student cognitive outcomes by subject area. Several meta-analysis studies have been conducted to measure the effectiveness of various technology indicators and instructional methods for supporting student achievement in different content areas. For example, Kulik (2003), as part of his litera ture review on the effects of using instructional technology in elementary and sec ondary schools, specifically analyzed 61 studies conducted after 1990. However, he did not report how many studies were reviewed that did not meet inclusion criteria. For this study, he searched ERIC, Dissertation Abstracts, and NSF databases for experimental and quasiexperimental studies in the following categories: integrated learning systems (ILS), reading management systems, writing programs for reading, word proce ssing and Internet resources, microcomputer-based laboratories and science tutoring and simulation. Specifi cally, he looked for Level II interventions that had a common theoretical basis but may have had different implementations and Level III innovations that were clearly defined with specific materials, implementa tion procedures, and professional development. The

PAGE 58

44 experimental and quasi-experimental studies had to have controlled comparisons of outcome measures that could be used to measure effect sizes. However, Kulik used the median effect size of the group of studies to measure the overall effect of the technology intervention. He did not calculate a mean weighted effect size nor did he examine specific aspects of the interventio n by performing regression analysis on the study technology variables and outcomes. Instead Kulik (2 003) used a narrative approach to describe the characteristics of the intervention in each study. Kulik (2003) found a median eff ect size of 0.28 for the impact of integrated learning systems (ILS) on composites of scores from students’ reading and mathematics achievement. However, Kulik (2003) found that the median effect sizes of seven studies that investigated the impact of ILS for only mathematics instruction on student mathematics achieve ment was 0.38. The effect sizes for ILS for only mathematics instruction ranged from 0.14 for grade six to 1.05 for grade eight with an overall median effect size of 0.40. Time for implementation of these programs ranged from several months to five school years. Penuel and other researchers at SRI (2002) co nducted an evaluation synthesis to examine the impact of technology used to promote connections between home and school and improve student outcomes. They used a systematic search of the Internet, research organizations, journals and educational databases for abstracts of articles from 1995 to 2001 about experimental, quasi experimental, or prepostdesign studies that measured student learning or engagement, parent involvement, or outcomes of parentschool communications. They specifically searched for combinations of keywords: home, parent, family, school, technology, computer, laptop, and voicemail. Although they had millions of hits, examined 98 abstracts, and reviewed 28 research articles, in total th ey calculated 103 effect sizes for all sub analyses that were reported in 19 articles. Thus, out of the initial 28 research articles reviewed, only 19 met the methodological criteria to be includ ed in the analyses, a loss of 32%. Only two studies used experimental designs. Several issues may influence the interpretation of the results reported by Penuel and colleagues (2002). The first is that two of the studies used in th e calculation of effect sizes were embedded in schoolwide reform initiatives. The positive results from these programs reported for technology innovations cannot be separated from the total effect of all the co mponents of the school-wide reform. The second issue is that several of the studies included multiple levels for more than one school year, so the results from

PAGE 59

45 these same students may be included in more than one level of the results. The third issue is that several studies were sponsored and funded by the vendor and some studies did not reveal their funding. Nevertheless, similar to Kulik (2003), Penuel et al. (2002) found positive relationships between technology and mathematics achievement in their meta-analysis, although their effect sizes were not as large. The weighted effects of different technology treatments (laptop, home desktop, and discrete software) on mathematics achievement ranged from -0.01 to +0.94, with weighted m ean effect size of 0.18. Although a total of nineteen effect sizes were calculated, two of th e programs with three of th e highest effect sizes were embedded in school reform programs (Penuel et al ., 2002). These results suggest that the specific instructional methods used with the technology have a great impact on the effec tiveness of the treatments. Indeed, Kulik (2003) also investig ated the effect size of ILS on mathematics achievement in nine studies when the implementation incl uded reading instruction as well. Most studies used elementary age populations, and implementation time was longer than the studies with just mathematics, from six months to three years. The effect sizes fo r mathematics ranged in these studies from 0.04 to 0.58, with median effect of 0.17.The range of effect size for reading was from 0.00 to 0.44 with an overall mean effect for reading of 0.06. Interestingly in Kulik’s (2003) meta-ana lysis, schools that had the three highest effect sizes for mathematics had the three lowest effect sizes for reading. Likewise, the two schools with the highest effect sizes in reading, had effect sizes within the three lowest in mathematics. These results may indicate that an ILS has the greatest positive impact on student achievement when it is used for only one subject area at a time, as demonstrated by the difference in m ean effect sizes for mathem atics, 0.40 when ILS is used exclusively for mathematics vs. 0.17 when ILS is used for instruction in both mathematics and reading. There may be a threshold for the minimum amount of focused time that students need to interact with ILS to achieve results as well as maximum amount of time that students are able to attend to instruction in this format. Additional, meta-analyses have been conducted to specifically find the e ffects of instructional technology on reading ach ievement. For example, Pearson, Ferdig, Blomeyer, Jr., and Moran (2005) conducted a meta-analysis to find the effects of di gital literacy tools on reading performance of middleschool students in strategy use, metacognition, reading motivation, reading engagement, and reading comprehension. They used the following inclusion criteria for selection of research studies: reports had to

PAGE 60

46 be published between 1988 and 2005 and peer-reviewed; experimental, quasi-experimental, or preposttest design; sample included grades 6, 7 and 8 or aggregat ed results that also included either grades 5 or 9; reading skills including comprehension, metacognition, strategy use and/or motivation as the outcome; technology as the independent or moderating variable; and statistics reported that are necessary for calculating effect size. Using search engines, journal databases, inte rnational journals, and websites of professional organizations, 204 articles were located. Ultimately, 20 articles that met the inclusion criteria were selected for analysis, a loss of 90% of the st udies reviewed. The researchers used a random effects model. The effects for each study were weighted and ag gregated to find an overall effect for the study, and then these 89 effects were weighted to determine an ove rall effect of all studies. Pearson et al. (2005) used two of the same studies in their an alyses as Waxman et al. (2003). Reading comprehension was the only criterion outcome that Pearson et al. (2005) could analyze, and had a weighted mean effect size of 0.49 ( z =4.36, p <.0005). The researchers also looked at contextual variables to determine differences in effect sizes. As expected, they found that sp ecial populations such as at risk readers and students with learning disabilitie s had smaller effect sizes than the general education population ( d =0.32 vs. 0.52, Q =4.42, p <.05). Penuel et al. (2002) reported we ighted effects of different technology treatments (laptop, home desktop, and discrete software) on reading achievement that ranged from 0.07 to 1.26 with weighted mean effect size of 0.10. However, two of these programs were also embedded in a school reform initiative, so the positive results can not be attributed to the technology but to the total effect of the school reform program. In addition, some studies occurred over time for multiple grades, so some students are included in the analysis more than once. Kulik (2003) conducted a meta-analysis of 13 studies to find the effect of the Writing to Read program on reading achievement in kindergarten, first grade, and elementary grades. Effect sizes were greatest for the youngest students, kindergarten (0.63 and 1.06 with median effect of 0.84), first grade (0.18 to 0.78 with median effect of 0.40), and elementary (-0.01 to 0.70 with median effect of 0.25). Effect sizes for three controlled studies of Accelerated R eader, a reading management program, conducted by independent evaluators ranged from -0.02 to 1.12 with median effect of 0.43 (Kulik, 2003). Reading effect sizes for ILS studies ranged from 0.00 to 0.44 (Kulik, 2003).

PAGE 61

47 Results from the meta-analysis reported in three separate studies by Pearson et al. (2005), Penuel et al. (2003) and Kulick (2003) indicate that instructional technology can have a positive effect on reading achievement. Mean effect size for these studies ranged from 0.10 to 0.49. Of interest, there were no overlaps in studies used by these researchers. Further research is needed to find the specific technology integrations that produce consistent positive improvements in reading achievement. In addition, it is important to disaggregate this information by student demographics to determine the interventions that are most likely to improve readin g achievement for each group. Goldberg et al. (2003) conducted a meta-analysis on the effects of computers on student writing. They identified 99 articles through searches of ERIC Educational Abstracts, PsychLit, and Dissertation Abstracts databases; websites of government and professional organizations; Google search engine; ejournals; and contacting researchers in the field. Incl usion criteria included studies conducted between 1992 and 2002; longitudinal studies of the impact of word-processing over time or comparison of paper and pencil writing with using a computer for writing; samp le of K – 12 grades; and outcome measures that include quality, quantity, or revisions of student writin g that are not focused on spell checkers, grammar checkers or test administration. Twenty-six studies me t the inclusion criterion and provided the statistics necessary to calculate effect sizes, a loss of 74% of the reports reviewed. Pre-post test designs were analyzed using only post-test, and we ighted effect sizes were used to determine the overall effect size for each of the three outcomes. Additional tests of homogeneity and publication bias were conducted. Regression analyses were conducted with moderator variables. Goldberg et al. (2003) used fourteen studies to calculate the mean weighted effect size for quantity of writing ( d =0.50). Goldberg et al. also found a mean weight ed effect size of 0.40 higher for the quality of writing for students who wrote with a computer when compared to the quality of writing of students who wrote with pencil and paper. Through regression analysis on moderating aspects that impact using computers for writing, the researchers found that studen ts in middle school made greater gains in quantity and quality of writing than those in high school and elementary school. Confirming results were found by Kulik (2003) by his meta-analysis on the effects of technology innovations on writing achievement. He found that the effects of using word proc essing ranged from -0.42 to 0.54 with median effect size of 0.30. Penuel (2002) also found similar results. Five programs that used

PAGE 62

48 technology to improve writing had a weighted mean effect of 0.34. Weighted effects for each program ranged from -0.09 to 0.40. Nineteen effect sizes were calculated; however, the two highest effect sizes were from schools involved in school-wide reform. In ad dition, the assessment of writing was problematic since the studies measured writing skills with different methods. In fact, none of the studies used students’ writing samples obtained from their class work (Penuel, 2002). These three meta-analyses (Goldberg et al., 2 003; Kulik, 2003; Penuel, 2002) demonstrate the powerful impact that using a computer for word pr ocessing can have on students writing achievement. Mean effect sizes for these meta-analyses ranged from 0.30 to 0.50. Only one study was used in both the meta-analyses of Kulik and Goldberg. Follow-up research is needed to investigate the best instructional methods for integrating word proce ssing into the curriculum and daily activities of students that produce the greatest improvement in st udents’ writing achievement. Important information about the quality of the experimental research conducted to investigate the impact of technology on student achievement can be gleaned from the descriptions of the methods used in meta-analysis research. All meta-analytic researcher s included in the limitations of their studies the quantity of reports that lack the tech nical information that was needed to calculate effect sizes so they could not be included in their analyses (Borman et al., 2003 ; Goldberg et al., 2003; Kulik, 2003; Pearson et al., 2005; Penuel et al., 2002; Waxman et al., 2003). In addition, Borman (2003) found that having a study conducted by the developer increased the effect size by 0.16 standard deviations over evaluations conducted by third parties. Penuel et al. (2002) expressed concerns that more than half of their research reports were sponsored by vendors, which might indicate a conflict of interest, while other reports did not even designate the source of their f unding. As a result, Waxman et al. (2003) only included articles in refereed journals. Nevertheless, Waxman et al. also complained about the quality of many of the technology reports included in these journals. Even peer reviewed journals that publish educational research have not maintained the quality of reported re search by requiring authors to report all information needed to replicate the study and a ll statistical information necessary fo r calculating effect sizes. Indeed, between 32% and 90% of the reports reviewed by the me ta-analyses researchers in this literature review were not analyzed because they did not meet the mi nimum quality requirements necessary to be used in their meta-analysis. There is great need for quality experimental and quasi experimental research with

PAGE 63

49 control groups to be conducted to determine the best practices for integrating technology that will positively impact student achievement. In addition, the reports about this research should comply with the quality standards for reporting results that can be used by independent researchers in further studies and educators for making informed decisions for technology integration planning and program implementation. Table 2. Meta-analysis Research Studies about the Integration of Technology and Student Achievement Citation Purpose Time Period Studies Effect Size Borman, G. D., Hewes, G. M., Overman, L. V., & Brown, S. (2003). Comprehensive school reform and achievement: A meta-analysis. Review of Educational Research, 73 (2), 125-230. relationship of comprehensive school reform and student achievement; reading achievement and CSR; math achievement and CSR 232 studies and 1111 independent observations overall mean effect size of comprehensive school reform (CSR) and achievement 0.15 ( Z = 33.26, p < 01); CSR and reading achievement 0.13 (Z = 10.81, p < .001); CSR and math achievement 0.15 (Z = 9.86, p < .001) Waxman, H. C., Lin, M. & Michko, G. (2003). A meta-analysis of the effectiveness of teaching and learning with technology on student outcomes. Naperville, IL: Learning Point Associates. Retrieved February 17, 2008, from http://www.ncrel.org/tech/effects2/ teaching and learning with technology on student outcomes 1997 and 2003 42 studies cognitive outcomes mean study-weighted effect size of .448 (p<.001) with 95% confidence intervals that did not include zero Kulik, J. (2003). Effects of using instructional technology in elementary and secondary schools: What controlled evaluation studies say Arlington, VA: SRI International. Retrieved February 17,2008, from http://www.sri.eu/policy/csted/reports /sandt/it/Kulik_ITinK12_Main_Report.pdf 61 studies conducted after 1990 ILS effect size 0.28; mathematics overall mean effect of 0.17; reading overall mean effect size 0.06; word processing median effect size of 0.30.

PAGE 64

50 Citation Purpose Time Period Studies Effect Size Penuel, W. R., Kim, D. Y., Michalchik, V., Lewis. S., Means, B., Murphy, R., Korbak. C., & Whaley, A., (2002). Using technology to enhance connections between home and school: A research synthesis. Arlington, VA: SRI International. Retrieved February 17, 2008, from http://ctl.sri.com/publications/display Publication.jsp?ID=83 19 articles and 103 effect sizes Mathematics achievement effect size weighted mean effect size of 0.18; writing weighted mean effect of 0.34. Pearson, P. D., Ferdig, R. E., Blomeyer, Jr., R. L., & Moran, J. (2005). The effects of technology on reading performance in the middleschool grades: A meta-analysis with recommendations for policy. Naperville, IL: Learning Point Associates. Retrieved February 17, 2008, from http://www.ncrel.org/tech/reading/ind ex.html 20 articles Reading comprehension weighted mean effect size of 0.49 (z=4.36, p<.0005) Goldberg, A., Russell, M., & Cook, A. (2003). The effect of computers on student writing: A meta-analysis of studies from 1992 to 2002. Journal of Technology, Learning, and Assessment, 2 (1). Retrieved February 17, 2008, from http://escholarship.bc.edu/jtla/vol2/1/ 13 studies Writing mean effect size 0.40 Research synthesis. Not all research supports the effectiveness of using technology to deliver instruction. For example, Lockee et al. (2004) repo rted in their research synthesis on programmed instruction, which is the foundation for computer-assisted, computer-based tutorials, and web-based tutorials, that almost all research conducted was of poor quality so that the results could not be generalized beyond that particular study. As suggested by the findings from the meta-analyses of research, the results of poor quality research cannot be used to supp ort the effectiveness of using technology. Even when researchers tried to control the confounding variables in order to isolate the impact of using technology to deliver instruction, mixed results were reported. Hill and colleagues (2004) reported in their research synthesis on using the Internet to deliver instruction that results yielded both positive and negative impacts on learning. Kmitta & Davis (2004) reported that most research studies have found a low to moderate positive effect for computers on student achievement, although with a great deal of variance.

PAGE 65

51 Mixed findings also have been reported when specific aspects of the instruction have been examined. Although the organization of the instructional material is often thought to impact learning, Shapiro and Niederhauser (2004) reported from their literature review that some research studies have demonstrated that system structure for the delivery of content has positively impacted learning, while other research studies found none or even negative effects. They explain that these conflicting results are due to the interaction between the structure an d the attributes of the learners, su ch as the prior knowledge, goals, and metacognitive skills of the learner. Consequently, the ability to adapt instruction for individual learning differences would make the use of technology to deliver instruction very appealing. However, Park and Lee (2004) reported that “no convincing evidence was found to suggest that such individual differences were useful for differentiating alternative treatments” (p. 659). Another important variable that influences learning of content and that technology shows promise for manipulating, is the feedback provided to the learner. Nevertheless, Mory (2004) reported that only half of the research studies found any ef fect from task-specific feedback and even less from information-based feedback. In addition, another area of disagreement was the optimal timing of feedback that maximizes learning, whether to use immediate or delayed feedback. Mory (2004) interpreted that the differences in findings among the studies were due to the various ways that researchers defined the treatments used in the studies. Computer-mediated communication (CMC) has been used to create learning environments that students interact through the comput er interface with objects, simulate d personalities, and/or other real participants (Romiszowski & Mason, 2004). There are many forms of CM C that either occur at the same time or synchronously (e.g., written chat, audio conf erencing, and multi-user object-oriented environments) or delayed time or asynchronously (e.g., discussion boards, e-mail, and listserv). Luppicini (2007) defines CMC used for educational purposes as “the process by which people create, exchange, and perceive information using networked telecommunications systems that facilitate encoding, transmitting, and decoding messages” (p. 143). At the most basic level, threaded discussions have been used to support learning through asynchronous written discourse (R omiszowski & Mason, 2004). Within discussion forums, students write about issues and respond to their classmates over a course of time. However, Romiszowski and Mason (2004) reported that scant research has been conducted and results have been

PAGE 66

52 inconclusive from research that compares different instructional methods that would support the increased effectiveness of discussion forums on student achievement. Luppicini (2007) stated that mixed findings were a result of different tasks with differen t objectives and differ ences in participants. Mircroworlds and virtual realities are at the highest level of CMC, as they use technology to support synchronous, immersive interactivity. Simulations are a form of these virtual realities that has been used educationally both to deliver content and as a tool to support learning. Gredler (2004) makes a qualitative distinction between simulations that are “ope n-ended evolving situations with many interacting variables… [through which learners] experience the effects of their deci sions” (p. 571) and models where students solve “a well-defined problem” (p. 572) to learn and understand specific relationships among variables. Rieber (2004) explains this distinction as the difference between using technology as an object to think with in order to solve problems versus learning information from a model that was designed by someone else. Using computer simulations to learn basic content has not always been successful. For example, students using simulations were no more successful than students in control classes. When students were not taught prerequisite knowledge before engaging in discovery learning simulations, they learned inaccurate information (Gredler, 2004 ). In this case, using discovery learning to infe r the underlying scientific relationships of the mode l placed too high a cognitive load on students to be successful (Gredler, 2004; Rieber, 2004). Students needed additional supports during the simulation activity. However, even when technology was used to deliver the scaffolds or prompts to support activities, discussions, and selfmonitoring processes, results indicated that students’ learning did not improve unless the program provided instruction that matched what the student needed (Dennen, 2004). In addition, Dennen explained that not only must students know how to use the supports provided for higher-order problem-solving, but they also must have the prerequisite knowledge to know when to request the supports. Although using computers to support higher-order problem-solving in an open-ended complex and ill-defined case study or real-world virtual reality offe rs promise, little research has been conducted to evaluate the educational benefits (Gredler, 2004; McLe llan, 2004; Rieber, 2004). Virtual realities used for educational applications have been implemented mainly for professional (medical) and military training

PAGE 67

53 (McLellan, 2004). Much of the current research has been developmental or design research, which focused on the iterative refinement of the product, often without a theoretical fram ework (Rieber, 2004). Kim and Reeves (2007) proposed that using the com puter as a tool should be examined using the theoretical framework of expertise and distributed cognition. They explained that the use of cognitive tools requires extended time for students to develop the expertise needed to be able to use the tool as an extension for enhanced intellectual activity and distributed production. Initially, when a new cognitive tool is introduced during a problem-solving activity, it produces additional cognitive load while the student becomes familiar with the interface and learns how to use the tool for lear ning. The process of learning to use cognitive tools is iterative, as the learner develops expertise in both subject matter and using the tool. The relationship between the learner and the tool is dynamic and complex; they cannot be separated, and thus, learning must be assessed while the student uses the tool (Kim & Reeves, 2007). Large-scale longitudinal research. Several research studies have investigated the effects of large scale technology initiatives over time. One of the firs t large-scale state-wide educational reform initiatives based on the use of technology was implemented by West Virginia (Mann et al., 1999). The Basic Skills/ Computer Education program (BS/CE) was first implemented in all kindergarten classes in 1990-91 school year, and then over the next eight years, with each successive year, it was implemented in the next higher grade level. Each school was provided with a networked file server and enough computers and printers to equip each class in the targeted grade level with th ree or four computers and a printer as well as the decision on how and where to impl ement the program, either in each classroom or in computer labs. Counties could select integrated lear ning systems from two providers, eith er Jostens Learning or IBM that matched their pedagogical practices. Thus, all schools in a grade level were given the same software for basic reading, writing, and mathematics skills development, and all teachers participated in thorough professional development prior to implementation and on-going support during initial implementation. To investigate the effects of this program, Mann et al. (1999) examined the longitudinal gains in achievement of fifth graders in 1996-97. The study used mixed methods that included survey data from students and teachers, interviews with teachers and principals, observations, and document analysis, and gain scale scores from Stanford-9 reading, language arts, and mathematics achievement tests. The sample included a stratified sample of 18 schools based on achievement, intensity of program implementation,

PAGE 68

54 geography, vendors, and SES. The sample included 950 fifth grade students who began their participation in the program in kindergar ten in 1989-90 and all 290 third through fifth grade t eachers in the 18 schools. Researchers surveyed all students and teachers in the sa mple and interviewed all principals and fifth grade teachers as well as selected teacher s in lower grades. Addition data we re collected from documents related to technology planning and implementation at the dist rict, school, and classroom level, as well as state records. There were three components in the regression model for the BS/CE program used for determining the impact of technology on student achievement gain s: hardware and software access and use; student and teacher attitudes; and teacher traini ng and involvement. Re sults of the regression model accounted for 11% of the variance in the achievement gains of the stude nts. Moreover, the research ers found that the children without home computers made the greatest gains in total basic skills, total language, language expression, total reading, reading comprehension, and vocabulary. In addition, the placement of the computers in the classroom was important, as teachers who had computers in their classroom reported higher skill levels for planning, managing, and delivering instruction as well as using computers more often for instruction in reading, math, writing. Thus, the students in cla ssrooms with computers made the greatest gains in achievement. The variable that had the most impact on student achievement gains was time, that is, both the frequency that students participated in the BS /CE program during each y ear and their accumulated experience in the program over all of the years of the study. However, the researchers point out that all student achievement gains cannot be attributed to the BS/CE program alone because West Virginia was involved in other reforms during the same period of time (e.g., building renovations, significant increases in teacher salaries, instituting a stat ewide curriculum framework, stat e-wide standards testing, and accreditation visits) that also impacted student achievement. Wenglinsky (1998, 2005) investigated the relationships between technology used with instructional methods and math, science, and reading achievement. In 1996, 6,000 fourth grade students and 7,000 eighth grade students and in 2000 (Wenglinsky, 1998, 2005), 13,000 fourth grade students and 15,000 eighth grade students took mathematics assessments, and 13,000 fourth grade and 15,000 eighth students took science assessments in 1996 and 2000 (Wenglinsky, 2005). Although Wenglinsky does not report the number of eighth grade students who took the NAEP reading assessment in 1998, The NAEP

PAGE 69

55 1998 Reading Report Card for the Nation and the State reports that 22,000 eighth grade students took the reading test. However, the NAEP data are cross-sectio nal and different cohorts of students are tested each year, so the student data collected by NAEP can not be used for repeated measures longitudinal analysis. Wenglinsky used t-tests with two groups at a time w ith Bonferonni adjustments for multiple comparisons to statistically test for differences among different groups of students. He used structural equation modeling with multiple indicators to find models that best explained the relationships among student variables, teacher variables, technology indicat ors, and student achievement in ma thematics, science, and reading. Due to the administration method of the NAEP that uses different booklets of tests with students rather than the complete test, total scores were imputed an d design effects were used for the analysis. Wenglinsky (1998, 2005) found positive relationships between technology used with specific instructional methods that focused on higher order thinking skills and achievement in both mathematics and science for both fourth and eighth grades when examining secondary data from the NAEP in 1996 and 2000. Interesting, when all uses were included, increased computer use at school had a negative relationship with mathematics and science achievement at both grade levels. Professional development for computers was related to higher achievement in eigh th grade for math, science, and reading. Using computers to revise drafts was significantly rela ted to reading achievement. Learning games were associated with higher achievement at the fourth grade level in both mathematics and science. Noteworthy, the variable that had the greatest relationship with ma thematics, science, and reading achievement at all grade levels was socio-economic status. In a more recent mixed method study of one school district, Lowther et al. (2003) investigated the impact of using laptops in the cla ssroom on teaching strategi es and student achievement. They selected one treatment class and two control classes in the same school at the same grade level in four middle schools and one elementary school resulting in 21 classrooms (12 laptop and 9 control classes) in grades 5, 6, and 7. Control classes had access to 5-6 desktop computers within the classr oom. Previous writing and science achievement scores for some of the laptop and control students were compared before the treatment. Results indicated significant writing advantage of the control group and a significant science advantage of the laptop group. Researchers collect ed data through classroom observations, district writing assessment, problem solving task rubric, student surveys, student focus groups, teacher interviews, and district parent

PAGE 70

56 interview. School observations were conducted us ing the School Observation Measure (SOM) and the Survey of Computer Use (SCU). Observers were trained to use these instruments, and the inter-rater reliability for the five category response rubric SOM was 67% for identical responses and 95% for responses within one category difference, and the interrater reliability for identical responses to the five category rubric was 86%. Randomly selected writing samp les of 59 control and 59 laptop students in sixth and seventh grades were assessed by trained reviewers with the district four point rubric on four dimensions. Fifty-two laptop students and 59 control students were randomly selected to complete a specially designed problem solving task. Trained reviewers used a rubric with seven components to blindly assess the student responses for 3 levels of performance. Inter-rater correlations ranged from 0.73 to 0.79. Parallel forms of on-line student surveys were administered to 257 laptop students and 134 control students. Reliabilities of the scores of Likert-style items were .795 and .854 for the laptop group and .735 and .806 for the control group. Seventy-one randomly selected students participated in six focus groups, and six teachers were randomly selected from teachers in the control and laptop groups at each grade level to participate in interviews. Lowther et al. used analysis of variance (ANOVA), multivariate (MANOVA), t tests for independent samples, and chi square tests of indepe ndence statistical tests on quantitative data collected. Effect sizes were computed using Cohen’s d with pooled standard deviations. The researchers reported that laptop computers positively impacted students’ writing and problem solving skills when compared to students without access to laptops. Results of MANOVA analysis on overall writin g indicated that both sixth and seventh grade laptop students performed better than the control students. Effect sizes for the four dimensions ranged from 0.53 to 1.47. Results of MANOVA analysis on problem solving achievement indicated that sixth and seventh grade laptop students performed better than the control students. Positive effect sizes for the five of the seven dimensions ranged from 0.38 to 0.76. Multi-level model research. Several recent studies used multi-level modeling techniques for analysis. Russell, O’Dwyer, Bebell, and Tucker-Seeley (2004) examined the relationship between students’ mathematics test scores and computer use at home and school. For this study, th e researchers obtained a stratified sample of fourth grade teachers who were high, medium, and low user s of technology from an original sample that included 200 schools in 22 school districts in Massachusetts between spring 2001 and

PAGE 71

57 spring 2003. Teachers were recruited to participate in this study. The re sults from an ad ditional district were incorporated with this sample resulting in 1,213 students and 55 teachers from 25 el ementary schools in 9 school districts. After excluding special stude nt populations, students with disabilities and English Language Learners, the sample included 986 students. When compared with the averages for all students in the state, the students in the researcher’s sample had a higher ratio of students to computers, had a lower proportion of economically disadvantaged students, and had higher average academic performance scores. Students and teachers completed surveys about their technology use in Spring 2003, and achievement data from the Massachusetts Comprehensive Assessment System (MCAS) fourth grade mathematics scores and subscale scores from 2002-03 were obtained from the districts. Reliability of the total mathematics score was .86 and the subscale scores ranged from .32 to .71. Reliabilities of composite scores made to measure student technology variables ranged from .54 to .74, and reliabilities of the composite scores used to measur e teacher use of techno logy ranged from .45 to .89. Scores were standardized for the analysis (Russell et al., 2004). The multi-level models with student level and teach er level factors for the total mathematics score accounted for 16% of the total variance explained. Si gnificant student-level f actors in the full model included the student’s grade 3 reading score and the nu mber of computers at home, and significant teacherlevel factors included the teacher mean 3rd grade student reading score and a negative relationship with teacher directs students to create products using technology. However, Russell and colleagues report that the impact of the measure, the teacher directs st udents to create products using technology, became insignificant when the more parsimonious model was an alyzed using only the significant factors. Thus, results from this study seem to indicate little relationship between technology use and student achievement in mathematics. O’Dwyer et al. (2005) examined the relations hip between home and school computer use and students’ English/ Language Arts test scores. O’Dwyer and colleagues (2005) used the same sample as Russell et al. (2004) with the same technology indicators from student and teacher surveys. The fourth grade MACS English/ Language Arts total score and sub domain scores for reading, literature, and writing were used as outcome variables in the analysis.

PAGE 72

58 The full model with all student-level and teacherlevel factors explained 23% of the variance of the total scores on the fourth grade MACS English/ Language Arts test. Significant student-level factors that were negatively related to total score were ho w often students use a computer in school to create presentations and recreational home use of the computer. Significant positive student-level factors related to English/ Language Arts Achievement included the frequency that students use a computer in school to edit papers, how many books the student owns at home, how many computers at home, and the student’s 3rd grade reading score. The only signi ficant teacher-level factor relate d to achievement was the teacher mean student 3rd grade reading score. When the more parsimonious model with only significant factors was analyzed, all factors remained sign ificant and accounted for 24% of the total variance explained. The researchers followed up with analyzing the writing scores in the sub domain of the MACS English/ Language Arts test. The full model with all student-level and teacher-level factors explained 12% of the variance. Significant student-l evel factors and teacher-level factors were the same as the total MACS English/ Language Arts test, except that recreation hom e use was no longer significant. Follow-up analysis for the Reading and Literature sub domain scores ex plained 25% of the total va riance. Significant studentlevel factors and teacher-level factors were the same as the total MACS English/ Language Arts test, except that how often students use a computer in school to create presentations was not significant. The consistent positive relationship between technology and student English/ Language Arts achievement found by O’Dwyer et al. (2005) was with using a computer at school to edit papers. The negative relationships between student-level technology factors, recreational home use and using computers at school to create presentations, and student English/ Language Arts ach ievement were not significant for all three outcome measures. Shapley et al. (2006) evaluated th e first year of the Texas Techno logy Immersion Pilot, which was conducted in 22 middle schools with high proportions of economically disadvantaged students. Requirements of the program were bot h students and teachers had laptops with productivity tools, wireless access to on-line curriculum resources, and ongoing t echnical support. Data we re collected for this evaluation study through site visits, observations, preand post campus technology survey by technology coordinators, preand post teacher surveys, and preand post student su rveys in fall 2004 and spring 2005. The initial cohort used in this study included 5,564 sixth grade students and 1,304 teachers in both

PAGE 73

59 immersed and control schools. Additional demographic data were collected through the Texas Public Information Management System and the Academic Excellence Indicator System. Student level discipline data were collected from 2605 middle schools in spring 2005. Multi-level models were analyzed to determine the relationships between predictor technology immersion indicators obtained from students and t eachers responses on the su rveys with achievement outcomes measured by the Texas Assessment of Knowledge and Skills (TAKS). Internal reliability measures of the TAKS were reported to be between hi gh .80’s and low .90’s. Results found from the initial analysis during this first year indicate there was no significant relationship between technology immersion and reading or mathematics achievemen t for sixth graders. Shapley et al (2006) explain that the lack of significant results for student achie vement outcomes was because most schools had only partial immersion and teachers reported that most stude nts only participated in technology activities once or twice a month. In addition, the researchers expected that analyses of long itudinal data collected from future data were needed to reveal the impacts of immersion on student achievement. Dynarski and colleagues (2007) used 3-level multi-level modelling to determine the effectiveness of reading and mathematics software products for incr easing student achievement. This study included a total of 33 districts, 132 schools, and 439 teachers. Software products we re grouped together for first grade reading, fourth grade reading, sixt h grade math, and Algebra. Districts and teachers were recruited because they did not already use the tech nology. An experimental design was used; teachers were randomly assigned to treatment or control groups. Teachers in control groups were able to use technology products that they had available to them. Teachers in experime ntal groups were also fre e to discontinue using the products or use the products in ways that were not intended. Students’ achievement was measured by the researchers in the fall and again in the spring. Achievement data also we re collected from the districts and schools when available. In addition, the researchers observed each classr oom three times during the school year to assess product implementation. Teachers were interviewed about implementation issues, and background information was collected with a teacher survey. All teachers were trained to use the products. Additional variables included in the model were student age, gender, pre-test scores, teacher gender, teacher experience, masters degree, school race and et hnicity, percent of students in special education, percent of students eligible for free lunch, time using treatment product, time using other products,

PAGE 74

60 adequate preparation time, students having problems accessing products, computer specialist on staff, and student, classroom, and school random effects (Dynarski et al., 2007). Results from the first year of the two year study indicated that the differences in test scores between the control and treatment groups for first gr ade reading were not significantly different. All products used for the first grade treatment had a tutorial-practice-assessment modular structure. The only moderating effect that was statisti cally significant was the teacher-stude nt ratio. Three of four products used for fourth grade reading were tutorials with practice and assessments for specific reading skills. The fourth product provided access for teachers to digital re sources that they could choose to use to supplement the reading curriculum. The differences in test scores between the control groups and treatment groups were not statistically significant; however, there was a moderating effect for the duration that the product was used. The three 6th grade math products were tutorials with practice and assessments. Results indicated no statistical difference between the test scores of the treatment and control groups, and there were no moderating effects from student, classroom, or school variables. The Algebra products covered the conventional curriculum. One product was a full curric ulum with most activities carried out during class periods “off-line”. The other two products were supplements to the regular curriculum. Results again indicated that there was no statisti cal difference between the test scores of control and treatment groups, and there were no statistically significant moderating effects from student, classroom, or school variables (Dynarski et al., 2007). Researchers at SRI International (2007) conducted an experimental study using 2-level multi-level modelling statistical analysis to determine if the integrated curriculum “SimCalc Mathworlds” could enhance the understanding of seventh grade students about rate and proportionality. Participants were selected from volunteers who attended a summer workshop and had complete data. For a two to three-week period the treatment group (48 teacher s ) used the SimCalc unit, while th e control group (47 teachers) used the existing textbook. All teachers r eceived three days of training. Th e researchers developed a student assessment using psychometrically recommended procedures to measure student learning (Roschelle, Tatar, Shechtman, Hegedus, Hopkins, Knudsen, & Stroter, 2007). Results indicated a significant ov erall treatment effect (0.84, t (93)=9.1, p <0.0001). Most of the difference between groups occurred on the co mplex skills assessment portion (effect = 1.22, t (93)=10.0,

PAGE 75

61 p <0.0001. The researchers attribute th e significant differences to the incr eased focus of the instruction that students received in the treatment group on cognitive complexity. The cognitive complexity of teaching was measured by the daily report of teachers using a 4-point Likert scale to designate the degree of lowerorder and higher-order goals (Roschelle et al., 2007). These positive results need to be contrasted with the insignificant results reported by Dynarski et al. (2007). The duration of the study conducted by Roschelle et al. was three weeks, while the duration of the study conducted by Dynarski et al. was an acade mic year. It is possible that the newness of the treatment impacted students’ motivation to learn during the one instructional unit. However, another difference may have been the essential integration of the technology into the highly interactive curriculum to stimulate cognitive complexity. The instructional implementation of the math products evaluated by Dynarski et al. may not have been delivered with as high a focus on cognitive complexity. Summary of research on student achievement. Research synthesis literature reviews yielded inconclusive results about the impact of technology integration on student achievement (e.g., Gredler, 2004; Hill et al., 2004; Lockee et al., 2004; Luppicini, 2007; McLellan, 2004; Metri Group, 2006; Mory, 2004; Park & Lee, 2004; Rieber, 2004; Romiszowski & Mason, 2004; Shapiro & Niederhauser, 2004). Results from meta-analysis suggest that computer use has positive effects on student reading (Kulik, 2003; Pearson et al., 2005; Penuel et al., 2002), writing (Goldberg et al., 2003; Kulik, 2003; Penuel et al., 2002), and mathematics achievement (Kulik, 2 003; Penuel et al., 2002). These e ffects may be enhanced when they are embedded in school-wide reform programs (Borman et al., 2003; Kulik, 2003; Penuel et al., 2002). Indeed, Mann et al. (1999) found positive impacts of technology when embedded in a long-term state-wide reform initiative. Structural Equation Modelling with large scale NAEP assessments found that technology use was positively related to student science, math, and reading achievement when used to enhance higher order thinking skills (Wen glinsky, 2005). Mixed method research also found positive impacts from using computers on writing and problem solving achievement (L owther et al., 2003). However, research using multi-level modelling statistical techniques with large scale data found no significant relationships between computer use and student achievement in math (Dynarski et al., 2007; Russell et al., 2004) and reading (Dynarski et al., 2007). One study found a positive re lationship between using a computer to edit papers with reading and writing achievement (O ’Dwyer et al., 2005), and another for using integrated technology

PAGE 76

62 for a mathematics unit (Roschelle et al., 2007). However, the time span for all of these multi-level modelling research studies was only two years or less. Time may be the critical component. Results from the comprehensive school reform and student achievement meta-analysis found that the years of implementation had a large impact (Borman, 2003). Indeed, schools that had implemented their comprehensive school reform model for five years had almost twice the effect size as all schools in general, and after seven years of implementation, schools had effect sizes of 0.50. Given that multi-level models have f ound significant positive relationships between having access to computers and the frequency with which student s and teachers use them (O ’Dwyer et al., 2004, 2005; Shapley, 2006), there may be mediating variables that must be impacted first before student achievement is effected. Table 3. Large Scale Research Studies about the Relationship of Technology Integratio n with Student Achievement Citation Purpose Time Period Studies Effect Size Mann, D., Shakeshaft, C., Becker, J., & Kottkamp, R. (1999). West Virginia story: Achievement gains from a statewide comprehensive instructional technology program Retrieved February 17, 2008, from http://www.mff.org/publications/pu blications.taf?page=155 stratified sample 18 schools; 950 students 8 years mixed methods positive impact on reading, writing, and math basic skills Wenglinsky, H. (2005). Using technology wisely: The keys to success in schools New York: Teachers College Press. Math, reading and science 13,000 students in 1996 and 28,000 in 2000 1996 and 2000; one year Structural Equation Modeling positive relationship of technology used with higher order thinking skills and problem solving with achievement in math and science Lowther, D. L., Ross, S. M., & Morrison, G. M. (2003). When each one has one: The influences on teaching strategies and student achievement of using laptops in the classroom. Educational Technology Research and Development 51 (3) 23-44. 5 schools and 12 classes 2001-02 mixed methods positive relationship with overall writing

PAGE 77

63 Citation Purpose Time Period Studies Effect Size Russell, M., O’Dwyer, L., Bebell, D., & Tucker-Seeley, K. (2004). Examining the relationship between students’ mathematics test scores and computer use at home and at school Boston, MA: Technology and Assessment Study Collaborative, Boston College. Retrieved February, 17, 2008, from http://escholarship.bc.edu/intasc /28/ 9 school districts, 25 schools, 55 teachers, and 986 students 20012003 multilevel models little relationship between technology use and student achievement in mathematics O’Dwyer, L. M., Russell, M. Bebell, D. J., & Tucker-Seeley, K. L. (2005). Examining the relationship between home and school computer use and students’ English/ language arts test scores. The Journal of Technology, Learning, and Assessment, 3 (3) Retrieved February 17, 2008, from http://escholarship.bc.edu/jtla/ 9 school districts, 25 schools, 55 teachers, and 986 students 20012003 multilevel models consistent positive relationships between students using a computer at school to edit papers and writing achievement Shapley, K., Sheehan, D., Caranikas-Walker, F., Huntsberger, B., & Maloney, C. (2006). Evaluation of the Texas technology immersion pilot: First-year results. Austin, TX: Texas Center for Educational Research. Retrieved February 17, 2008, from http://www.tcer.org/research/etxtip /index.aspx 22 middle schools, 5,564 sixth grade students, and 1,304 teachers 20042005 one school year multilevel modeling no significant relationship between technology immersion and reading or mathematics achievement Student Behavioral Outcomes The lack of significant results found by current research for the relationships of technology integration with student achievement may be a result of mediating variables such as students’ motivation to learn, their conduct, and their attendance. If technolo gy integration positively impacts these factors, it may result in a positive impact on student achievement. Ringstaff & Kelley (2002) reported that technology has had a positive impact on student self-confidence, responsibility, and attitudes toward learning. These impacts also lead to improved student attendance rates and decreased dropout rates (Ringstaff & Kelley, 2002). Kmitta and Davis (2004) reported that research studies have demonstrated that computers supported students improved motivation to learn and their behaviour at school. Barron et al. (1999) investigated the relationships between student conduct and the number of computers per student in Florida schools. Barron et al. collected school level data for the number of computers available from the State of Florida Computer

PAGE 78

64 Use Surveys for 1993-94 (N = 2250), 1994-95 (N = 2200), and 1995-96 (N = 2350) school years. Schools were divided into two groups by the direction in the trend of students using computers at school. Thus, schools with decreasing trends of students using technology included 74 elementary schools (5%), 27 middle schools (6%), and 17 high schools (5%) with increasing trends of students using technology included 786 elementary schools (56%), 231 middle schools (52%), and 148 high schools (47%). Schools that were stable or had inconsistent trends were purposively excluded from the analyses. In addition, schools that were not elementary, middle or high schools were purposively removed from the sample. School demographic information and discipline and attendance information were obtained for the 1995-96 school year from the State of Florida Department of Education Annual School Reports. Differences between schools in the proportions of students on Free or Reduced Lunch status and proportions of minority students were controlled for by the statistical analysis. Effect sizes were used to compare the two groups Findings indicated that schools with increasing trends for students using computers had better student conduct measured by mean differences in total conduct violations and better attendance rates measured by mean differences in rates of attendance (elementary schools total conduct violations d =-0.14 and attendance rate d =0.25; middle schools total conduct violations d =-0.35; and high schools total conduct violations d =-0.23 and attendance rate d =0.09). However, middle schools did not experience increases in attendance rates; indeed, the trend was decreased rates ( d =-.09). The researchers pointed out that this inconsistent result may have been because there were a large number of factor s that can impact student outcomes such as student socio-economic status that were not controlled in the study. Moreover, they also pointed out that the unit of analysis in this study was the school, not the individual student. Barron et al. suggested that future studies look at how computers are used not just the ratio of computers to students. Although Waxman et al. (2003) found no significant effect between technology and students’ behavioral outcomes using meta-analysis, in 2006, Barron et al.’s results were supported by a new study that included student level data. Shapley et al. (2006) reported a positive relationship between technology immersion and decreased number of students referrals ( d =0.16) and suspensions ( d =0.06) during the initial year of a large scale middle school laptop immers ion initiative. Although they found a significant difference between treatment and control groups for improved school attendance rate ( d =-0.08), the

PAGE 79

65 researchers explained that this significant difference was the same before treatment as well and concluded that there were no apparent gains in student attendance from the treatment. Moreover, Muir-Herzig (2004) examined the relationship between at-risk student attendance and teacher use of technology, student use of technology, and overall use of technology. Results indicat ed no significant relationships between any of the technology uses and attendance. Mu ir-Herzig explained that the overa ll use of technology by all teachers and students in the study was very low, and that w ithout appropriate use by students, their attendance was not impacted. All studies call for further investigation with longitudinal data. Summary of research on student behavioral outcomes. Ringstaff and Kelley (2002) and Kmitta and Davis (2004) reported that research has demonstrated positive relationships between technology integration and positive student behaviours. In addition, positive relationships were found between technology integration and improved student conduct (Barron et al., 1999; Shapley, 2006). However, Waxman et al. (2003) reported no significant effect between technology and students’ behavioral outcomes using meta-analysis. Furthermore, other studies reported no significant differences in the relationship of technology integration with student attendance (Bar ron et al., 1999; Muir-Herzig, 2004). These mixed results may be due to the duration of the studies and the measurement of the variables. It is important to examine the relationship between technology integration and student behavioral outcomes over extended period of time. Table 4. Research Studies about the Relationship between Technology Integration and Student Behavioral Outcomes Citation Purpose Time Period Studies Effect Size Barron, A.E., Hogarty, K.Y., Kromrey, J.D., & Lenkway, P. (1999). An examination of the relationships between student conduct and the number of computers per student in Florida schools. Journal of Research on Computing in Education, 32 (1), 98-107. top 5% increasing trends for technology and bottom 5% decreasing trends for technology resulting in 850 elementary schools, 258 middle schools, and 165 high schools 1993-94 to 199596 comparison of effect size schools with increasing trends for students using computers had better student conduct and attendance rates

PAGE 80

66 Citation Purpose Time Period Studies Effect Size Waxman, H. C., Lin, M. & Michko, G. (2003). A metaanalysis of the effectiveness of teaching and learning with technology on student outcomes. Naperville, IL: Learning Point Associates. Retrieved February 17, 2008, from http://www.ncrel.org/tech/effects 2/ 3 studies 1997 and 2003 metaanalysis mean studyweighted effect size for behavioral outcomes was slightly negative .091, (p > .05) Shapley, K., Sheehan, D., Caranikas-Walker, F., Huntsberger, B., & Maloney, C. (2006). Evaluation of the Texas technology immersion pilot: First-year results Austin, TX: Texas Center for Educational Research. Retrieved February 17, 2008, from http://www.tcer.org/research/etxt ip/index.aspx 22 middle schools, 5,564 sixth grade students, and 1,304 teachers; discipline records from 2605 middle schools 20042005 one school year multi-level modeling positive relationship between technology immersion and decreased number of students referrals Summary Complexity theory provides the framework that explains how schools as institutions respond to school reform efforts and academic standards for incr easing student achievement by adapting instructional methods. Technology facilitates this change process. Both the teaching and learning process and the interactions of students evolve as teachers integrate technology into the curriculum. Some research has revealed positive relationships between technology use and student achievement, while others have reported no significant relationships. However, there are many factors that infl uence student achievement that need to be controlled or included in research investigations. Moreover, few large scale, longitudinal studies have investigated the relationship between technology integration and student achievement using multi-level modelling that takes into account the nested nature of educational data and moderating factors. This research study built on previous studies by us ing multi-level modeling with state-wide, school-level data about their technology integration and their aver age school achievement scores collected over three years.

PAGE 81

Chapter 3: Methods While traditional statistical techniques cannot handle violations of the basic assumption of independence, multi-level modeling is a flexible approach that allows the analysis of data that have a nested structure. This is especially important with educatio nal data analyses, since the variables that compose the educational environment and instructional dynamics are nested by nature. Multi-level modeling also allows for analyses to be conducted on longitudinal data when there are missing observation points, without losing participants or information. Therefor e, when examining changes, this t echnique is preferable to repeated measures Multivariate Analysis of Variance (MAN OVA), because MANOVA cannot handle longitudinal data that is “messy” or unbalanced with uneven time points and missing data (Luke, 2004). Also, in this study, multi-level modeling allows the variance to be decomposed into between school and within school components (Luke, 2004; Raudenbush & Bryk, 2002). This study used mu lti-level modeling analyses with longitudinal data over a four year period. Data sources The Division of Accountability, Research, and Measurement of the Florida Department of Education provides educational data in order to promote longitudinal research that will improve the outcomes of students in Florida schools Division of Accountability, Research and Measurement, Florida Department of Education, 2007a). Many aggregated vari ables at the school level are available in a variety of publicly available on-line databases. For this study data school-level data were downloaded from four of these on-line databases. Downloaded data for schools from the different datasets were connected using the school identification number. Master School Identification (MSID) files The identification number and school level designation were obtained from the Master School Identification (MSID) files in Florida. The file for the current y ear (2006-07) was available to the public at the FLDOE website (Division of Accountability, Res earch and Measurement, Florida Department of Education, 2007a). Files for the other years included in the study (2003-04 to 2005-06) were obtained by e-

PAGE 82

68 mail request to the person designated on the FLDOE website. The categorical designation as magnet school or magnet program with technology specialty was obta ined from the MSID file for 2005-06 (Bureau of Education Information and Accountability Services, Florida Department of Education, 2007). Instrument: Florida Comprehensive Assessment Test The FCAT Reading, Math, and Writing were chos en to measure achievement because they are required to be administered to all students enrolled in public schools as part of the state of Florida student assessment and school accountability program developed by the Department of Education. The FCAT in Reading and Mathematics has been administered every y ear to all students in public schools in grades 3 through 10 since 1998. These assessments were developed using psychometric procedures to assure their validity and reliability. The scores of the 2003 FCAT for grade levels 3 to 10 were reported to have internal consistency reliability measured with Cronbach’s alpha and KR-20 for Math between .87 and .93 and for Reading between .88 and .91(Human Resources Research Organization & Harcourt Educational Measurement, 2003). The reading an d mathematics assessments include multiple choice, short response, and extended response items. The mathematics FCAT also includes guided-response items. In 2005, the test was changed from the version that was based on the Stanford Achievement Test Series, Ninth Edition (SAT9) to the version based on the Stanford Achi evement Test Series, Tent h Edition (SAT10) (Florida Department of Education, 2005b). The Total Reading wa s normed in the spring of 2002 with a stratified national sample by geographic area, socio-economic status, urbanicity, and ethnicity (Harcourt Assessment, Inc., 2002, 2004). Scaled scores have approximately equal units on a continuous scale and allow scores within a domain to be compared ac ross levels (Harcourt Assessment, Inc., 2002, 2004). Norm referenced scaled scores for Total Reading and Total Mathematic s ranged from 400 to 800 (Harcourt Assessment, Inc., 2002, 2004). The mean scaled scores for the spring norms of Total Reading for grades 3 to 10 ranged between 621 and 702, with standard deviations between 36.7 and 39.2 (Harcourt Assessment, Inc., 2002, 2004). The mean scaled scores for the spring norms of Total Mathematics for grades 3 to 10 ranged between 606 and 700, with standard deviations between 35.9 and 40.8 (Harcourt Assessment, Inc., 2002, 2004). The writing FCAT is a 45-minute essay that is scored based on focus, organization, support, and conventions (Florida Department of Education, 2005a). In 2006, multiple-choice items were added to the FCAT Writing Test and the score was changed to a scale score that range from 100 – 500; however the

PAGE 83

69 rubric-based sub-score was also reported (Florida Department of Education, 2006). These rubric scores range from unscorable to 6. Each school’s mean FCAT NRT score in Reading and Mathematics, for grad es 3 to 10 and mean rubric-based score in Writing for grades 5, 8, and 10 for 2004, 2005, 2006, and 2007 were obtained from the Florida Comprehensive Assessment Test, Assessmen t, and School Performance System (Division of Accountability, Research and Measurement, Florida Department of Education, 2007b. Obtaining these datasets involved downloading separate MS Ex cel files for each FCAT given for each grade (3rd 10th) for each school year (2003, 2004, 2005 and 2006) from the Florida Depart ment of Education Assessment and School Performance website. Instrument: Florida School Indicators Report (FSIR) Other school factors for the school years between 2003-04 and 2005-06 were obtained from the interactive on-line Florida School Indicators Re port (Division of Accountability, Research and Measurement, Florida Department of Education, 2007c). Indicators for the outcome variables of attendance and student conduct for the school years between 2003-04 and 2005-06 were downloaded for each school in the state. In addition, other indicators used as pred ictor variables for learning environment variables that included both student variables and teacher professional variables were downloaded for all the schools for the school years between 2003-04 and 2005-06. Data for the school year 2006-07 was unavailable. These data files were all downloaded as comma delimited data files. Instrument: Average Yearly Progress Reports Demographic variables were obtained from the Average Yearly Progress Reports for school years 2003-04, 2004-05, 2005-06, and 2006-07 on the Florida School Grades website (Division of Accountability, Research and Measurement, Florida Department of Education, 2007d). These indicators provided demographic information about the school proportions of low socio-economic status, minority, and Limited English Proficiency students, as well as proportion of students with disabilities (Florida Department of Education, 2007b). These file s were downloaded in comma delimited format. Instrument: System for Technology Accountability and Rigor (STAR) Surveys Each year, the Florida Departme nt of Education surveys every school in Florida about how technology is used by teachers an d students within their schools (Bur eau of Instruction and Innovation,

PAGE 84

70 Florida Department of Education, 2007a). The instrument that has been used since 2002 was called the System for Technology Accountability and Rigor (STAR) Survey. In the 2006-07 school year, the survey name was changed to Florida Innovates Survey (STAR), and was administered in sp ring 2007. This survey contains 78 items and was organized into five sec tions: digital learning environment, instructional leadership, Florida di gital educators, access to technology, and infrastructure and support. The response rate on the survey has been very high – 97% in 2003-04 (N = 2514); 96% in 2004-05 (N = 2553); 97% in 2005-06 (N = 2658); and 97% in 2006-07 (N = 2700). Survey items included radio buttons with 2 to 5 options; check boxes that allowed the selection of all that applied; and open-ended responses that usually involved reporting a numeric quantity. Predictor Variables Predictor variables were organized by category and by how they were added to the multi-level model. The first category was school level. Then demographics and learning environment variables were added to the model. Finally, technology integration predictor variables were added to the model. School Level Predictors at the School Level All classifications of schools were obtained from the Master School Identification (MSID) files (Division of Accountability, Research and Measurement, Florida Department of Education, 2007a). The definitions of these variables were obtained from the Technical assistance paper: Master School Identification File – 2006-07 (Bureau of Education Information and Accountability Services, Florida Department of Education, 2007). Elementary school level was a categorical predictor that was designated in the MSID file. Middle school level was a categorical predictor that was designated in the MSID file. High school level was a categorical predictor that was designated in the MSID file. Technology Magnet was a categorical predictor that was designated in the MSID file. These schools could also be an elementary, middle, or high school. Demographic Predictors at the School Level Two factors were determined through exploratory factor analysis of the demographic variables. The first factor loaded free or reduced lunch stat us students, minority students, and limited English proficiency/ESOL students on one factor. The second fact or included special popula tions of students: gifted

PAGE 85

71 and students with disabilities. It was decided to add each of these variables separately in the multi-level model, because it was expected that gifted and students with disabilities would have opposite relationships with student achievement. Combining these groups for the analysis would negate any relationship with student achievement that existed. In addition, limite d English proficiency students had been expected to group with students with disabilities as a factor for sp ecial populations of students that needed specialized and individualized support in order to make adequa te gains in achievement. Rather than an academic grouping, the factor that limited English proficiency st udents loaded appeared to represent students with limited resources and social power. The focus of this study was academics; therefore, the relationships of the demographic variables with student outcomes were examined separately. Free or reduced lunch status students was measured by the indicator Economically Disadvantaged Students in the Measuring Adequate Yearly Progre ss Files (Division of Accountability, Research and Measurement, Florida Department of Education, 200 7d). This indicator measured the percentage of students who were eligible for free or reduced-price lunch and students enrolled in a USDA-approved Provision 2 school (Florida Department of Education, 2007d). Minority was measured by the indicat ors available in the Measuring Adequate Yearly Progress Files – Number of Students White, Black, Hi spanic, Asian/Pacific Islander, and American Indian/Alaskan. The classification Multiracial was not included in the count (Florida Department of Education, 2007d). The proportion of minority students was calculated by adding the numbers of students classified as Hispanic, Asian/Pacific Islander, and Am erican Indian/Alaskan or by subtracting the number of White students from the total number of students, and then dividing by this total minority of students by the total number of students. Limited English Proficiency (LEP) was measured by the number of students in the school who were currently being served in an English for Speakers of Other Languages (ESOL) program, as well as students who had attained English proficiency for up to two years after exiting the ESOL program (Florida Department of Education, 2007b). Student with disabilities was a measure that included the total number of Primary and Other Exceptionality fields with disabilities, other than gifted students (Florida Department of Education, 2007b).

PAGE 86

72 Gifted is only reported for elementary and middle school; therefore, this variable was included in a separate analysis for these levels (Florida Department of Education, 2007b). Another analysis was conducted with all school levels in cluding high school leve l without the gifted variable. Because the data were not available for gifted students in 2006-07, the data from 2005-06 were used to impute values. Learning Environment Predictors at the School Level Predictor variables used to measure the learning environment were obtained from the on-line Florida Indicators Report (Division of Accountability, Research and Measurement, Florida Department of Education, 2007c). The data for only three years (2003-04, 2004-05, and 2005-06) were available. The values from 2005-06 were used to impute values for 2006-07. Exploratory factor analysis was conducted with all of the variables used to measure the positive student learning environment and all of the variables used to measure teacher qualifications. Initially, Scho ol Staff and Student Membership were going to be used to create a variable to measure the ratio of students per instructional staff and included in the composite variable used to measure positive student learning environment. However, the low factor loadings (less than .3) obtained through exploratory factor analysis revealed that this was not a good measure to use for the learning environment. All othe r variables were used to create the composite score used to measure positive learning environment and teacher qualifications. Cronbach’s alpha used to measure internal consistency reliability of the scores fo r these composite variables are depicted in Table 5. Positive student learning environment was measured by six variables (Absent 21+ Days (Students); Stability Rate; Suspensions both in-house and out-of-school; and Incidents of Crime and Violence, Offenses, Student Membership) obtained from the on-line Florida Indicators Report (Division of Accountability, Research and Measurement, Florida Department of Education, 2007c). This variable served as a proxy for positive impacts to the school learning environment. Absence was measured by the indicator Students Absent 21+ Days. This was obtained from the percentage of students from the year’s total enrollment who were absent 21 or more days during the 180day school year (Florida Department of Education, 2007c). This variable was subtracted from 100% to produce a rate of absence for students who were absent less than 21 days. Stability rate was measured by the percentage of students in the October membership count who were still present in the February membership (Florida Department of Education, 2007c).

PAGE 87

73 Student conduct was measured by the indicators Suspensions and Incidents of Crime and Violence, Offenses (Florida Department of Education, 2007c). Suspensions were measured by the mean of the percentage of students who served in-school suspensions and the percentage of students who served out-of-school suspensions. To change this to a positiv e rate for the percentage of students who were not suspended the percentages of suspensions were subtracted from 100. Incidents of Crime and Violence, Offenses was obtained from the total number of reported incidents occurring on school grounds, on school transportation, or at school-sponsored events. Categories of offenses included in this report were: violent acts against persons; alcohol, tobacco, or other drugs; property offenses; fighting and harassment; weapons possession; other nonviolent incidents and disorderly conduct. The total number of incidents was divided by the Student Membership to get a ratio of the number of incidents per student and then multiplied by 100. The result was then multiplied by negative one to make the score be a penalty for positive learning environment. The sum of each of these scores was used for the composite variable that measured positive school learning environment. That is absence (percentage of students not absent over 21 days), stability, suspensions (the percentage of students who were not suspended in house and the percentage of students who were not suspended out-of-school) were added together. Then the score for Crime and Violence, Offenses per student multiplied by 100 was subtracted from the total. Teacher qualifications were measured by three variables (Average Years of Experience, advanced degree attainment, and teaching in ce rtified field) obtained from the On-line Florida Indicators Report (Division of Accountability, Research, and Measurement, Florida Department of Education, 2007c). Average Years of Experience was measured by the indicator Teachers: Average Years of Experience (Florida Department of Education, 2007c). It consists of the average number of years of in-state and out-ofstate teaching experience for teachers in the school To change this to a positive measure for Teacher Experience the mean Average Years of Experience for the school was divided by the Average Years of Experience for all of Florida’s schools. Advanced Degree was measured by the indicator Teachers: Master’s Degree or Higher (Division of Accountability Research and Measurement, Florida Department of Education, 2007c). This indicator was the percentage of instructional staff in the school with an advanced degree (master’s degree, a doctorate, or a specialist’s de gree). Teaching in certificate field was measured by

PAGE 88

74 the indicator Classes Taught by Teac hers Teaching out of Fi eld (Division of Accoun tability, Research and Measurement, Florida Department of Education, 2007c). This indicator was a measure of the percentage of classes being taught by classroom teachers teaching out-offield for core academic courses (English, reading, language arts, mathematics, science, fore ign languages, civics, government, economics, arts, history, and geography). It was subtracted from 100% to yield the positive variable Teaching In-field. Teacher skill was measured by summing three variables, that is, the valu e of (Average Years of Experience + Advanced Degree + Teaching In-field). Technology Integration Measures The response data from the STAR surveys were downloaded from the Florida Innovates website as spreadsheets in either text format or comma deli mited format. The survey re sponses for each year were in multiple files. Only responses to survey items that were included on the survey for all four years were used in this data analysis. A side-by-side table comp aring the exact survey items from each year is included in the Appendix B. Factor analysis was conducted to validate the grouping of the items used in the analysis. Internal Consistency Reliability of the scores for each of the groups of variables from survey items used in this study is reported in Table 5. Student access to software. To determine the school level variables for access to software for student use in the school, the item “What percentage of student computers at your school have the following software types available on them?” was used. Factor analysis was used to separate the types of software into three categories: content software, office/ production software, and advanced production software. Then an overall level of availability was computed by calculating the mean degree of availability of all of the programs in each of th ese categories. When no response op tion was designated for a particular software program, it was assumed that the software wa s not available and zero was used in calculating the mean. The Cronbach’s alpha for the scores for stud ent access to software is depicted in Table 5. Percent of teachers who regularly use computer technology. The responses to the item “Approximately what percentage of your teachers regularly uses technology in the following ways?” were used to measure the degree of teacher computer technology use at the sc hool level. Factor analysis was used to separate the types of software into two categories: delivery of instruction and administrative purposes. The composite score for each of these factor s was computed by calculating the mean percentage

PAGE 89

75 of teachers who regularly used each of the software programs in the fact or that were common in the items across all four years. When no response option was designated for a particular software program, it was assumed that the software was not used and zero was us ed in calculating the mean. The internal consistency of the responses is depicted in Table 5. Frequency that students use software. The item “How often do students at your school use the following types of software?” was used to measure the frequency of student t echnology usage. This item was worded slightly differently in the 2003 survey and had to be reverse coded in order to compare the results (see Appendix B). Factor analysis was used to separate the types of soft ware into two categories: content delivery software and tool-based software. The internal consistency reliability of these responses is depicted Table 5). Support for technology. A composite variable was created by summing the scores of the following items: “Our school-based technical support is provided by:”; “Our school-based instructional technology specialist is:”; “How dependable is the Internet connection at your school?”; “How often do you experience delays when using the Internet at your school?”; and “What is the average length of time at your school for a technical issue to be resolved?” (See Appendix B). F actor analysis was used to separate the types of support into two categories: human/ time and hardware/ Internet. Response options to the first two items included level of support from none to full-time Items that have no responses were recoded as no support. Variables were recoded so that higher scores designated more support. Responses to “How dependable is the Internet connection at your school?” were recoded so that the option very dependable had the highest value. The item “How often do you experience delays when using the Internet at your school?” was reworded and ordered in 2005-06. The responses to items for 2003-04 and 2004-05 were recoded so that no delay has the highest score. The responses to the item “Wha t is the average length of time at your school for a technical issue to be resolved?” were recoded so that the shortest length of time had the highest value. One other item was considered for inclusion for this category, “What percentage of the money spent on technology for your school is devoted to professional development in technology-related training?” Low factor loadings (less than .3) during the factor analysis demonstrated th at this item was not contributing to the measurement of support. It was not used for crea ting the composite variable. The internal consistency reliability measures of these responses for sup port for technology are depicted Table 5.

PAGE 90

76 Table 5. Internal Consistency Reliability for Predicto rs by Year Measured with Cronbach’s Alpha Cronbach Coefficient Alpha Variables 2003 2004 2005 2006 Learning environment predictors Positive learning environment .68 .66 .65 Teacher Qualificati ons .30 .37 .33 Student access to all soft ware .79 .79 .77 .75 Student access to content so ftware .80 .81 .69 .67 Student access to office/ produc tion software .73 .70 .73 .66 Student access to advanced produc tion software .6 0 .60 .66 .68 Teachers regularly use so ftware .78 .77 .78 .75 Teachers regularly use for delivery of instruction .70 .68 .72 .69 Teachers regularly use for administrative purposes .67 .64 .64 .56 Frequency students use all software .37 .57 .76 .78 Frequency Students Use Content Delivery Software .27 .50 .52 .55 Frequency Students Use Production Tool Software .54 .63 .82 .83 Tech Support .54 .56 .54 .51 Tech Support Human/ Time .63 .68 .62 .62 Tech Support Internet/ Hardware .62 .57 .54 .53 Outcome Measures There were two categories of outcomes examined : student achievement and student behavioral outcomes that were mediating outcomes. Student Achievement Reading achievement. School level reading achievement was measured by the mean FCAT Reading norm referenced scores fo r each school for each year (Divisio n of Accountability, Research and Measurement, Florida Department of Education, 2007b) The mean scores from all grade levels in the school were averaged to obtain a school mean score for each year. Mathematics achievement. School level mathematics achievement was measured by the mean FCAT Mathematics norm referenced scores for each school for each year (Division of Accountability, Research and Measurement, Florida Department of Education, 2007b). The mean scores from all grade levels in the school were averaged to obtain a school mean score for each year.

PAGE 91

77 Writing achievement. School level writing achievement was measured by the mean FCAT Writing rubric scores for each school (Divis ion of Accountab ility, Research and Measurement, Florida Department of Education, 2007b). The mean scores from all grade levels in the school was averaged to obtain a school mean score for each year. The mean ru bric-based score reported for 2006 was used to match the data from the preceding years. Mediating Behavioral Outcomes Absence was measured by the indicator Absent 21+ Days (Students). This was obtained from the percentage of students from the year’s total enrollment who were absent 21 or more days during the 180day school year (Division of Accoun tability, Research and Measurement, Florida Department of Education, 2007c). Data were not available for the 2006-07 school year. Student misconduct was measured by the indicators Suspensions and Incidents of Crime and Violence, Offenses (Florida Department of Educa tion, 2007c). Suspensions were measured by the percentage of students who served in-school suspensions and the percentage of students who served out-ofschool suspensions. Incidents of Crime and Violence Offenses were obtained from the total number of reported incidents occurring on school grounds, on school transportation, or at school-sponsored events. Categories of offenses included in this report were: violent acts against persons ; alcohol, tobacco, or other drugs; property offenses; fighting and harassment; weapons possession; other nonviolent incidents and disorderly conduct. The total number of incidents was divided by the Student Membership to get a ratio of the number of incidents per student. The composite measure for school level of student misconduct was obtained from the sum of the percentage of students serving in-school suspensions, the percentage of students serving out-of-school suspensions, and the number of crime incidents per student for the school. Data were not available for the school year 2006-07. Data Preparation Procedures Merging Data Files First, the Master School Identification Files (MSI D) for each academic year were entered into the dataset. They were downloaded from the FLDOE website in MS Excel format or text format. Magnet school information was added to this file. Then for each year, the additional files were brought into the dataset. The next set of files brought into the data set were the school level mean FCAT NRT scores for

PAGE 92

78 reading, mathematics, and writing. These scores were used to measure the outcome variables in the study. The files were obtained for each of the four school years from the FLDOE at the Assessment and School Performance: Florida Comprehensive Assessment Test website and then merged with the MSID files to determine the number of schools at each school level with FCAT sc ores. Next, the Florida School Indicators Reports (FSIR) files for each school for each year that were obtained on-line from the FLDOE were merged. The FSIR did not include any information about minority status or information at the high school level about the proportion of students on Free or Reduced Price Lunch Programs. Therefore, data also were obtained from the AYP Reports on the FLDOE Evaluation and Reporting website for each school for each year. These files we re merged so that missing demographic information in the FSIR was filled in with data from the AYP. Last, the technology indicator variables from the responses to the STAR surveys were brought into the data set. As the files for each year were merged, missing data were analyzed. Complete procedures are delineated in Appendix C. These cleaned files were saved in comma delimited text format so they could be imported into SAS 9.1. Exploratory Factor Analysis Exploratory factor analysis was conducted with the data that was going to be grouped to make composite variables. Because the variables in each cat egory were expected to be correlated, the prior communality was inspected to determine if the items we re correlated. Because ther e was a high degree of correlation among the items, exploratory factor analysis was conducted using Principal Axis Factoring and oblique Promax rotation with listwise deletion of missing data. The number of factors was determined through inspection of the scree plot, proportion of the variance accounted for by each factor, parallel analysis, and consistency of interpretability over th e four years. Because the data were not normally distributed, exploratory factor analysis was conducted with the original data and with the natural log transformation of these data. Results from both of thes e analyses grouped the items in the same factors, so the original data were used for all the rest of the an alyses. The results from the exploratory factor analysis for each year for each composite vari able are delineated in Appendix C. Sample This study spans the four school years from 2003-04 to 2006-07. The sample was filtered to include only public elementary, middle, and high schools in Florida that participated in the System for

PAGE 93

79 Technology Accountability and Rigor (STAR) survey for all four years so that the relationships between technology indicators and school level achievement could be compared across school levels (see Table 6). Table 6. Number of Schools Used in the Analysis by Outcome School Level Elementary Middle High All School Levels Reading FCAT 1496 444 346 2286 Math FCAT 1511 444 346 2301 Writing FCAT 1480 437 347 2264 Student Conduct 1517 446 349 2312 Attendance 1517 446 349 2312 Descriptive Statistics for Outcome Variables Table 7 lists the descriptive statistics for each of the outcome variables by school level for each year, including range, mean, standard deviation, skew and kurtosis. To prevent collinearity new composite variables were created for positive l earning environment to be used in the analysis for student conduct and attendance. The student misconduct composite was left out of the composite variable measuring positive learning environment for the student conduct outcome, and the percent of students with more than 21 days absent was left out of the composite variable meas uring positive learning environment for the student attendance outcome. Table 7. Descriptive statistics for outcome variab les by school level and school year Variable School Level N School Year Mean STD Min Max Skew Kurt FCAT Reading All Schools 2286 2003-04 664.65 21.71 613.0 754.0 0.27 -0.51 2004-05 657.95 25.14 606.5 768.0 0.71 -0.06 2005-06 668.85 21.88 622.0 767.0 0.58 0.01 2006-07 667.69 22.35 618.5 762.5 0.62 -0.03 Elementary 1496 2003-04 653.00 14.62 613.0 700.7 0.04 -0.46 2004-05 643.61 13.27 606.5 693.3 0.18 -0.25 2005-06 657.37 13.75 622.0 704.3 0.17 -0.36 2006-07 655.58 13.30 618.5 704.7 0.19 -0.27 High 346 2003-04 693.18 13.09 649.5 754.0 0.25 1.58 2004-05 699.84 14.67 658.0 768.0 0.42 1.60 2005-06 703.92 13.84 667.5 767.0 0.48 1.56 2006-07 703.48 14.12 669.0 762.5 0.42 1.04

PAGE 94

80 Variable School Level N School Year Mean STD Min Max Skew Kurt Middle 444 2003-04 681.66 13.83 643.0 716.0 -0.04 -0.16 2004-05 673.62 13.72 640.3 710.7 0.10 -0.11 2005-06 680.22 12.11 650.0 712.3 0.05 -0.22 2006-07 680.60 12.70 649.7 715.7 0.12 -0.34 FCAT Math All Schools 2301 2003-04 658.47 30.33 595.0 791.5 0.75 -0.23 2004-05 655.59 32.47 592.0 784.5 0.56 -0.62 2005-06 664.07 32.42 594.0 793.5 0.60 -0.47 2006-07 667.42 29.96 604.0 780.5 0.50 -0.48 Elementary 1511 2003-04 640.22 13.94 595.0 689.0 0.10 -0.28 2004-05 635.87 16.40 592.0 696.5 0.22 -0.17 2005-06 644.78 16.71 594.0 704.5 0.21 -0.27 2006-07 650.14 17.06 604.0 712.0 0.16 -0.22 High 346 2003-04 710.58 15.05 672.4 791.5 0.61 2.37 2004-05 708.86 13.27 675.5 784.5 0.82 3.23 2005-06 717.56 14.81 675.0 793.5 0.71 2.27 2006-07 716.15 13.81 677.0 780.5 0.64 1.64 Middle 444 2003-04 679.96 14.82 641.0 719.0 0.11 -0.21 2004-05 681.22 14.85 647.3 723.3 0.30 -0.18 2005-06 688.01 16.02 651.7 735.7 0.30 -0.24 2006-07 688.28 14.55 657.0 731.0 0.33 -0.36 FCAT Writing All Schools 2264 2003-04 3.70 0.31 2.7 5.3 0.20 0.53 2264 2004-05 3.75 0.30 2.8 5.4 0.27 0.57 2264 2005-06 3.88 0.31 2.9 5.4 0.22 0.43 2264 2006-07 3.91 0.32 2.9 5.3 0.29 0.62 Elementary 1480 2003-04 3.64 0.30 2.7 4.6 0.09 0.20 1480 2004-05 3.70 0.29 2.8 4.7 0.13 0.14 1480 2005-06 3.84 0.31 2.9 4.9 0.12 -0.07 1480 2006-07 3.84 0.29 2.9 5.0 0.01 0.20 High 347 2003-04 3.83 0.26 3.1 5.3 0.74 3.22 347 2004-05 3.86 0.28 3.0 5.4 0.57 2.47 347 2005-06 3.91 0.30 3.2 5.4 0.80 2.25 347 2006-07 3.96 0.29 3.4 5.3 0.76 1.71 Middle 437 2003-04 3.79 0.33 2.9 4.8 0.35 -0.02 437 2004-05 3.82 0.30 3.2 4.9 0.55 0.18 437 2005-06 3.98 0.27 3.3 4.8 0.59 0.43 437 2006-07 4.13 0.31 3.5 5.0 0.55 -0.12 Percent of Students with Over 21 Days Absences All Schools 2312 2003-04 8.32 5.51 0.0 38.9 1.57 3.43 2312 2004-05 9.30 6.41 0.0 47.5 1.53 3.32 2312 2005-06 9.37 6.57 0.0 57.6 1.58 3.75 Elementary 1517 2003-04 6.29 3.14 0.0 24.5 0.82 1.32 1517 2004-05 7.59 4.76 0.2 47.5 1.92 6.78 1517 2005-06 7.52 5.11 0.1 57.6 2.36 10.27 High 349 2003-04 13.89 7.68 0.0 35.6 0.46 -0.19 349 2004-05 14.11 8.41 0.0 38.4 0.52 -0.14 349 2005-06 15.22 8.48 0.0 48.7 0.59 0.69

PAGE 95

81 Variable School Level N School Year Mean STD Min Max Skew Kurt Middle 446 2003-04 10.86 5.74 0.0 38.9 0.92 2.09 446 2004-05 11.34 6.96 0.0 44.4 1.11 2.81 446 2005-06 11.12 6.13 0.0 35.7 0.67 0.60 Student Misconduct All Schools 2312 2003-04 21.51 23.56 0.0 216.8 1.68 3.97 2312 2004-05 19.64 21.56 0.0 150.0 1.53 2.35 2312 2005-06 18.84 20.79 0.0 131.3 1.58 2.59 Elementary 1517 2003-04 10.17 14.09 0.0 132.2 3.88 21.40 1517 2004-05 8.94 11.55 0.0 104.2 3.37 16.58 1517 2005-06 9.16 12.07 0.0 121.2 3.40 16.52 High 349 2003-04 42.59 21.49 0.4 122.0 0.66 0.58 349 2004-05 39.25 20.76 1.2 108.6 0.70 0.38 349 2005-06 36.25 18.70 0.0 93.9 0.37 -0.09 Middle 446 2003-04 43.56 23.91 0.8 216.8 1.55 6.48 446 2004-05 40.67 22.08 0.3 150.0 0.95 1.66 446 2005-06 38.12 23.33 0.6 131.3 0.94 1.21 Descriptive Statistics for Predictor Variables The descriptive statistics for the predictor variab les for each of the outcomes by school level and school year are listed in Appendix C. Correlations of Technology Indicators with Predictor Variables Because technology is used with in the classroom, the variables used to measure the learning environment and the technology indicators were expected to be correlated. The correlations of composite variables used to measure the learning environment and composite technology indicators were analyzed for each outcome. The absolute value of correlations for predictor variable s ranged between 0.01 and 0.56 for all outcomes. The correlations are delineated in Appendix C. Data Analysis Plan Multi-level Models Multi-level models were used to estimate the relati onships of the predictors at each level with each outcome variable and to find the best model fit for the data using maximum likelihood estimation. In order to make meaningful comparisons, all predictors were standardized. Because the educational data for this study were nested at levels of time and school, twolevel models were progressively developed to describe the relationship between technology integration and school achievement. Due to the fact that technology is used within the classroom, the variables used to m easure the learning envir onment and the technology

PAGE 96

82 indicators were expected to be co rrelated. Therefore, because the fo cus of this study was on finding significant relationships between technology indicators and the outcome measures, variables were sequentially added to the model only to control their effect. SAS (version 9.13) statistical package (SAS Institute, Inc.) was used to analyze the data. The followi ng steps were taken to analyze the data and to test the following hypotheses for the research questions. Research Question 1 What is the relationship between indicators of technology integration and changes in mean student achievement when controlling for school level, school socio-economic status, minority, limited English proficiency, students with disabilities, gifted, teacher qualification, and learning environment quality? Hypothesis 1 After controlling for school level (elementary, middle, and high), school socio-economic status, minority, limited English proficiency, students with disabilities, gifted, teacher qualification, and learning environment quality, mean school achiev ement (FCAT NRT scaled scores for reading and mathematics and FCAT rubric scores for writing) will have a positive relationship with indicators of technology integration. In order to answer the first research question and te st the first hypothesis a series of models were built using FCAT reading as the first outcome, and then the same steps were followed to examine FCAT math, and FCAT writing. The steps were as follows: (1) The Unconditional Model with no predictors of FCAT reading achievement in the equation. (2) The Unconditional Growth Model with time added to the equation. (3) Since the traject ory may not have been linear, time2 and time3 were progressively added to form the polynomial equation. (4) School level was added to the equation of the Unconditional Growth Model with time, time2, and time3. (5) Demographic variables were added to the model separately in two steps, because students are not designated as gifted at the high school level in Florida. As a result, first the model was run using all three school levels, but without gifted in the equation. Then the model was estimated using only elementary and middle school levels with gifted as variable. Next the (6) Learning Environment variables were added to the equation: teacher qualification and student learning environment composite variables. (7) The Technology Indicators as composite variables were added to the equation. Last, (8) Technology Magnet School was added to the equation to determine if schools that had a high degree of technology infrastructure and professional development would have a positive relationship with achievement. At each step si gnificant predictors were retained. Alpha was set to .05 to designate significant parameters. The significance of the parameter estimates, the deviance statistic for the model fit, and the

PAGE 97

83 amount of variance explained by the model determined the most parsimonious and thus best model fit. The level 1 and level 2 residuals were examined for independence and normal distribution in the final model. Research Question 2 What is the relationship between indicators of technology integration and changes in mediating outcomes of attendance rate and student conduct, when controlling for school level, school socioeconomic status, minority, limited E nglish proficiency, students with disabilities, gifted, teacher qualification, and learning environment quality? Hypothesis 2: After controlling for school level (elementary, middle, and high), school socio-economic status, minority, limited English proficiency, students with disabilities, gifted, teacher qualification, and learning environment quality, percent of students absent more than 21 days will have a negative relationship with indicators of technology integration and mean school level of student misconduct will have a negative relationship with in dicators of technology integration. To answer the second research question and test the second hypothesis a series of models were built using percent of students absent more than 21 da ys as the first outcome, and then the steps were repeated using student misconduct as the second outc ome. Because the Florida Indicators Report for 200607 had not been released, these outcome variables we re only available for three years, and the models examined only 2003-04, 2004-05, and 2005-06. The steps were as follows: (1) The Unconditional Model added Attendance Rate to the equation. (2) The Unc onditional Growth Model added time to the equation. (3) Since the trajectory may not have been linear, time2 was added to form the quadratic equation. Because there were only three years of outcome variables to analyze for the trends time3 was not needed. (4) School level was added to the equation of the Unconditional Growth Model with time and time2. (5) Demographic variables were added to the model sepa rately in two steps, b ecause students are not designated as gifted at the high school level in Florida. As a result, firs t the model was run using all three school levels, but without gifted in the equation. Then the model was estimated using only elementary and middle school levels with gifted as variable. Next the (6) Learning Environment variables were added to the equation: teacher qualification and student l earning environment com posite variables. (7) The Technology Indicators as composite variables were added to the equation. Last, (8) Technology Magnet School was added to the equation to determine if schools that had a high degree of technology infrastructure and professional development had a positive relationship with achieve ment. At each step signi ficant predictors were retained. Alpha was set to .05 to designate significant parameters. The significance of the parameter

PAGE 98

84 estimates, the deviance statistic for the model fit, and the amount of variance explained by the model determined the most parsimonious and thus best model fit. The level-1 and level-2 residuals were examined for independence and normal distribution in the final model.

PAGE 99

85 Chapter 4: Results This chapter reviews the results that were obtaine d for the two main research questions separately. The steps that were used to analyze the data are delin eated. The results are reported for each h ypothesis for each research question. Research Question 1 What is the relationship between indicators of technology integration and changes in mean student achievement when controlling for school level, school socio-economic status, minority, limited English proficiency, students with disabilities, gifted, teacher qualification, and learning environment quality? The first research question was answered by conducting multi-level models with the FCAT achievement data for reading, mathematics, and writing. Hypothesis 1 The first analysis conducted to answer the first research question used the FCAT Reading outcome data to test the following hypothesis: H1: After controlling for school level (elementary, middle, and high), school socio-economic status, minority, limited English proficiency, students with disabilities, gifted, teacher qualification, and learning environment quality, mean school reading achievement (FCAT NRT scaled scores for reading) will have a positive relationship with indicators of technology integration. The first step was to build the unconditional model. The unconditional model predicted the schools’ FCAT reading from the average of FCAT reading for all schools. There were no other predictors. The average FCAT for all schools was 664.79 points ( t (2285) = 1411.95, p <.0001). Model 1: Unconditional Model Level 1: FCAT Reading = 0 + r Level 2: 0 = 00 + u0 Mixed-Effects Model: FCAT Reading = 00 + u0 + r The intraclass correlation coeffici ent (ICC) was computed to determine the proportion of variance in the FCAT Reading variable that is accounted for by the schools. The ICC was .92, which is high and

PAGE 100

86 supports using multi-level modeling for the analysis. The model fit statistics from this model were used as the baseline for model comparisons. The next step added time to the predictor equation (see Model 2a). The variance components from this analysis showed how much of the variance in the model was accoun ted for by time. There was almost no variance in the slopes between schools. Therefor e, time was set as a fixed effect, and the model with time as a fixed effect was estimated. Model 2a: Unconditional Growth Model Level 1: FCAT Reading = 0 + 1*Time + r Level 2: 0 = 00 + u0 1 = 10 + u1 Mixed-Effects Model: FCAT Reading = 00 + 10*Time + u0 + u1*Time + r Both the intercept ( t (2285) = 1384.17, p <.0001) and time ( t (6857) = 36.13, p <.0001) were significant parameters. Although there was no ad ditional explained variance between schools, time accounted for 16% of the variance within schools (see Model 2b). Model 2b: Unconditional Growth Model with Time Fixed Level 1: FCAT Reading = 0 + 1*Time + r Level 2: 0 = 00 1 = 10 Mixed-Effects Model: FCAT Reading = 00 + 10*Time + u0 + r To determine if the equation was not linear but curvilinear, time2 was added to the equation so the variance could be compared. Results indicated that time2 was significant ( t (6856) = 23.22, p <.0001) and increased the variance explained by an ad ditional 6% (see Model 2c). When time3 was added to the equation with time2, time3 also was significant ( t (4570) = -80.28, p <.0001), and all model fit indices improved. Although adding time3 increased the amount of variance between schools, it increased the variance explained by an additional 41%. Consequently, both time2 and time3 were retained in the polynomial growth model equation (see Model 2d). Model 2c: Quadratic Growth Model Level 1: FCAT Reading = 0 + 1*Time + 2* Time2 + r Level 2: 0 = 00 + u0 1 = 10 2 = 20 Mixed-Effects Model: FCAT Reading = 00 + 10*Time + 20* Time2 + u0 + r

PAGE 101

87 Model 2d: Polynomial Growth Model Level 1: FCAT Reading = 0 + 1*Time + 2* Time2 + 3* Time3 + r Level 2: 0 = 00 + u0 1 = 10 2 = 20 2 = 30 Mixed-Effects Model: FCAT Reading = 00 + 10*Time + 20* Time2 + 30* Time3 + u0 + r Next, school level was added to the Polynomial Growth Model to predict reading (See Model 3). The significance of the parameter estimates determined if school level was significantly related to the FCAT Reading and if there was an interaction with time. This model adjusted the mean school FCAT Reading and the slope of FCAT Reading growth for schoo l level. The parameter estimates of school level, time, time2, and time3 were all significant. The interactions between time and both the school levels, time2 and both the school levels, and time3 and both the school levels relative to middle school were also significant. All model fit indices indicated improved fit with this model (Table 8). This model accounted for 65% of the between school variance and an add itional 11% of the within school variance from the Polynomial Growth Model. Model 3: school level as Predictor Level 1: FCAT Reading = 0 + 1*Time + 2*Time2 + 3* Time3 + r Level 2: 0 = 00 + 01*School Level + u0 1 = 10 + 11*School Level 2 = 20 + 21*School Level 3 = 30 + 31*School Level Mixed-Effects Model: FCAT Reading = 00 + 01*School Level + 10*Time + 11*School Level*Time + 20*Time2 + 21*School Level*Time2 + 30*Time3 + 31*School Level*Time3 + u0 + r Table 8. Model 3: Time, Time2, Time3, and School Level as Predictors of Reading Effect School Level Estimate SE df t p Intercept 681.66 0.6476 2283 1052.6 <.0001 ** Time -22.3275 0.6021 6849 -37.08 <.0001 ** Time2 17.7633 0.5322 6849 33.38 <.0001 ** Time3 -3.4795 0.117 6849 -29.75 <.0001 ** School Level Elementary -28.6569 0.7375 6849 -38.86 <.0001 ** School Level High 11.5234 0.9786 6849 11.78 <.0001 ** School Level Middle 0 . . Time*School Level Elementary -11.5396 0.6857 6849 -16.83 <.0001 ** Time*School Level High 29.6256 0.9098 6849 32.56 <.0001 **

PAGE 102

88 Effect School Level Estimate SE df t p Time*School Level Middle 0 . . Time2*School Level Elementary 13.1594 0.606 6849 21.72 <.0001 ** Time2*School Level High -18.08 0.8041 6849 -22.48 <.0001 ** Time2*School Level Middle 0 . . Time3*School Level Elementary -2.9695 0.1332 6849 -22.29 <.0001 ** Time3*School Level High 3.1555 0.1767 6849 17.85 <.0001 ** Time3*School Level Middle 0 . . Covariance Parameter Estimate SE z p 175.28 5.2657 33.29 <.0001 ** Residual 10.9335 0.1867 58.56 <.0001 ** Note: p < .05; ** p < .01 The next model added student demographic variables to the School Level Model. This model was estimated twice. The first time, the model was run w ith high school as a school level and all of the demographic variables except gifted, because gifted is not a designation at the high school level (see Model 4a). The second time, the data were filtered to exclude high school as a school level and kept the gifted variable with middle and elementary schools (see Mo del 4b). The model fit statistics of the demographic model with all three school levels was compared with the School Level as Predictor Model to determine if there was a better fit (see Table 17). The significan ce of the parameter estimates determined which of the demographic variables remained in the predictor equa tion (see Table 9). The variance estimates showed the amount of the total variance that was accounted for by each model. When all of the demographics variables except gifted were added to the model (see Model 4a), the intercept was significant and the average middle school started with FCAT reading score of 679.36 ( t (2234) = 1945.24, p <.0001). The parameter estimates for school level, time, time2, time3, free or reduced lunch status, minority, limited English proficiency (LEP), and students with disabilities we re significant. Interactions with time were all significant except for those with minority. Interactions with time2 were all significant except for minority. Interactions with time3 were all significant except for minority and students w ith disabilities. All model fit indices indicated better fit with the addition of these demographics variables. Adding the demographics variables with school level explained 92% of the between school variance and 76% of the within school variance for a total of 91% of all variance explained. Model 4a: Demographics by School Level (including High School and no Gifted) Level 1: FCAT Reading = 0 + 1*Time + 2*Time2 + 3*Time3 + r

PAGE 103

89 Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* LEP + 35* SWD Mixed-Effects Model: FCAT Reading = 00 + 01*School Level + 02*SES + 03*Minority + 04*SWD + 05*LEP + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14*SWD*Time + 15*LEP*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24*SWD*Time2 + 25*LEP*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34*SWD*Time3 + 35*LEP*Time3 + u0 + r Table 9. Model 4a: Reading predicted by Time, School Level, and Demographics Variables No Gifted Effect School Level Estimate SE df t p Intercept 679.36 0.3492 2234 1945.2 <.0001 ** Time -20.953 0.6035 6443 -34.72 <.0001 ** Time2 17.2553 0.5312 6443 32.48 <.0001 ** Time3 -3.4122 0.1166 6443 -29.27 <.0001 ** School Level Elementary -25.5824 0.4006 6443 -63.85 <.0001 ** School Level High 6.7244 0.5268 6443 12.76 <.0001 ** School Level Middle 0 . . Free Reduced Lunch -5.6423 0.1728 6443 -32.65 <.0001 ** Minority -6.3466 0.2066 6443 -30.71 <.0001 ** LEP -0.5969 0.1789 6443 -3.34 0.0009 ** Students with Disabilities -1.9872 0.1271 6443 -15.63 <.0001 ** Time*School Level Elementary -13.6237 0.6986 6443 -19.5 <.0001 ** Time*School Level High 30.3797 0.9153 6443 33.19 <.0001 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch 0.9142 0.4112 6443 2.22 0.0262 Time*Minority 0.7421 0.4227 6443 1.76 0.0792 Time*LEP 1.1256 0.3518 6443 3.2 0.0014 ** Time*Students with Disabilities -0.7753 0.2973 6443 -2.61 0.0091 ** Time2*School Level Elementary 14.8613 0.6157 6443 24.14 <.0001 ** Time2*School Level High -18.5166 0.806 6443 -22.97 <.0001 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -1.0632 0.3622 6443 -2.94 0.0033 ** Time2*Minority 0.446 0.3713 6443 1.2 0.2298 Time2*LEP -0.9216 0.3068 6443 -3 0.0027 ** Time2*Students with Disabilities 0.5495 0.2604 6443 2.11 0.0349 Time3*School Level Elementary -3.315 0.1352 6443 -24.52 <.0001 ** Time3*School Level High 3.2411 0.1769 6443 18.33 <.0001 ** Time3*School Level Middle 0 . . Time3*Free Reduced Lunch 0.2308 0.07944 6443 2.91 0.0037 ** Time3*Minority -0.1406 0.08144 6443 -1.73 0.0844 Time3*LEP 0.1905 0.06699 6443 2.84 0.0045 **

PAGE 104

90 Effect School Level Estimate SE df t p Time3*Students with Disabilities -0.09712 0.057 6443 -1.7 0.0885 Covariance Parameter Estimate SE z p 40.5487 1.4746 27.5 <.0001 ** Residual 10.1746 0.1876 54.24 <.0001 ** Note: p < .05; ** p < .01 The results from the analysis in Model 4b indicated that the intercept, school level, time, time2, time3, free or reduced lunch status, minority, LEP, students with disabilities, and gifted were all significant (see Table 10). Interactions between time and elementary school level, free or reduced lunch status, LEP, and gifted were signifi cant. Interactions between time2 and free or reduced lunch status, LEP, and gifted were significant. Interactions between time3 and free or reduced lunch st atus, LEP, and gifted were significant. Because the parameter for gifted was significant in this model, an unconditional model using the same population with high schools filtered out, pr edicting FCAT reading with average FCAT reading was estimated in order to compare the fit of this mode l. All of the model fit statistics indicated better model fit. When examining the variance of FCAT reading in elementary and middle schools, adding demographics variables to the equation explained 91% of the between school variance and 78% more of the within school variance. Two sets of analyses were condu cted on the rest of the models in order to examine the relationship of gifted with technology integratio n as one of the predictors of school achievement. Model 4b: Demographics by School Level (Elementary and Middle School only) Level 1: FCAT Reading = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16*Gifted 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26*Gifted 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* LEP + 35* SWD + 36*Gifted Mixed-Effects Model: FCAT Reading = 00 + 01*School Level + 02*SES + 03*Minority + 04*SWD + 05*LEP + 06*Gifted + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP*Time + 15* SWD*Time + 16*Gifted*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Gifted*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34* LEP*Time3 + 35* SWD*Time3 + 36*Gifted*Time3 + u0 + r

PAGE 105

91 Table 10. Model 4b: Reading predicted by Time, School Level and Demographics Variables for Elementary and Middle Schools with Gifted Effect School Level Estimate SE df t p Intercept 678.32 0.3003 1805 2258.5 <.0001 ** Time -20.5295 0.6066 4908 -33.84 <.0001 ** Time2 16.8488 0.535 4908 31.49 <.0001 ** Time3 -3.324 0.1175 4908 -28.29 <.0001 ** School Level Elementary -23.972 0.3481 4908 -68.87 <.0001 ** School Level Middle 0 . . Free Reduced Lunch -5.6423 0.1845 4908 -30.59 <.0001 ** Minority -5.9592 0.211 4908 -28.24 <.0001 ** LEP -0.3349 0.1693 4908 -1.98 0.048 Students with Disabilities -1.159 0.1291 4908 -8.98 <.0001 ** Gifted 3.6388 0.1468 4908 24.79 <.0001 ** Time*School Level Elementary -13.7743 0.7103 4908 -19.39 <.0001 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch 1.2973 0.4815 4908 2.69 0.0071 ** Time*Minority 0.5315 0.4908 4908 1.08 0.2789 Time*LEP 1.6122 0.3801 4908 4.24 <.0001 ** Time*Students with Disabilities -0.1868 0.3268 4908 -0.57 0.5676 Time*Gifted 0.9406 0.3407 4908 2.76 0.0058 ** Time2*School Level Elementary 15.0662 0.6269 4908 24.03 <.0001 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -1.1625 0.4241 4908 -2.74 0.0061 ** Time2*Minority 0.4354 0.4304 4908 1.01 0.3118 Time2*LEP -1.3998 0.331 4908 -4.23 <.0001 ** Time2*Students with Disabilities 0.006271 0.2855 4908 0.02 0.9825 Time2*Gifted -0.9038 0.2977 4908 -3.04 0.0024 ** Time3*School Level Elementary -3.3724 0.1377 4908 -24.49 <.0001 ** Time3*School Level Middle 0 . . Time3*Free Reduced Lunch 0.2331 0.09296 4908 2.51 0.0122 Time3*Minority -0.1281 0.09424 4908 -1.36 0.1741 Time3*LEP 0.2944 0.07208 4908 4.08 <.0001 ** Time3*Students with Disabilities 0.01513 0.06229 4908 0.24 0.808 Time3*Gifted 0.1827 0.065 4908 2.81 0.005 ** Covariance Parameter Estimate SE z p 26.3269 1.0559 24.93 <.0001 ** Residual 9.5552 0.1984 48.15 <.0001 ** Note: p < .05; ** p < .01 The next model added the variable that measures the School Learning Environment factors to the Demographics Model by School Level Model. These in cluded teacher qualifica tions and positive learning

PAGE 106

92 environment. This model was estimated twice, first without gifted population but all school levels (see model 5a) and then with elementary and middle school levels and gifted population (see model 5b). When school learning environment factors were added with the demographic and school level variables for all school levels, the parameter estimates for the intercept, time, time2, time3, elementary and high school relative to middle school, free or reduced lunch stat us, minority, LEP, students with disabilities, teacher qualifications, and positive learning environment were significant (see Table 11). Significant interactions with time, time2, and time3 included elementary and high school re lative to middle school, free or reduced lunch status, LEP, teacher qualifications, and positive learning environment. Significant interactions with time2 were elementary and high scho ol, free or reduced lunch status, and positive learning environment. Adding the student learning environment variables explained an additional 2% of the between school variance and explained 1% less of the within school variance for a total of 92% of all of the variance explained. All of the model fit indices indicated that this model fit of the data better (see Table 17). Model 5a: School Learning Environment with Demographics by School Level (all school levels without gifted and LEP) Level 1: FCAT Reading = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04*SWD + 05* Teacher Qualifications + 06*Positive Learning Environment + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14*SWD + 15* Teacher Qualifications + 16*Positive Learning Environment 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* SWD + 25* Teacher Qualifications + 26*Positive Learning Environment 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34*SWD + 35* Teacher Qualifications + 36*Positive Learning Environment Mixed-Effects Model: FCAT Reading = 00 + 01*School Level + 02*SES + 03*Minority + 04* SWD + 05*Teacher Qualifications + 06* Positive Learning Environment + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14*SWD *Time + 15* Teacher Qualifications*Time + 16* Positive Learning Environment*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* SWD*Time2 + 25* Teacher Qualifications*Time2 + 26* Positive Learning Environment*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34*SWD*Time3 + 35* Teacher Qualifications*Time3 + 36* Positive Learning Environment*Time3 + u0 + r Table 11. Model 5a: Reading Predicted by Demographics and Stud ent Learning Environment by School Level (All School Levels without Gifted) Effect School Level Estimate SE df t p Intercept 680.14 0.4403 2298 1544.6 <.0001 ** Time -5.504 0.8634 6867 -6.37 <.0001 **

PAGE 107

93 Effect School Level Estimate SE df t p Time2 9.2097 0.7489 6867 12.3 <.0001 ** Time3 -2.1354 0.1624 6867 -13.15 <.0001 ** School Level Elementary -39.8795 0.5171 6867 -77.12 <.0001 ** School Level High 26.3373 0.6155 6867 42.79 <.0001 ** School Level Middle 0 . . Free Reduced Lunch -4.1447 0.1922 6867 -21.56 <.0001 ** Minority -6.1259 0.2129 6867 -28.77 <.0001 ** Students with Disabilities -2.1066 0.1437 6867 -14.66 <.0001 ** Positive Learning Environment 1.6474 0.1513 6867 10.89 <.0001 ** Positive Teacher Qualifications 1.003 0.1138 6867 8.81 <.0001 ** Time*School Level Elementary -11.0835 1.064 6867 -10.42 <.0001 ** Time*School Level High -9.3217 1.0637 6867 -8.76 <.0001 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch -1.411 0.489 6867 -2.89 0.0039 ** Time*Minority -0.4513 0.442 6867 -1.02 0.3074 Time*Students with Disabilities -0.4879 0.3396 6867 -1.44 0.1509 Time*Positive Learning Environment -0.7349 0.4571 6867 -1.61 0.1079 Time*Positive Teacher Qualifications 0.8114 0.3249 6867 2.5 0.0125 Time2*School Level Elementary 6.3613 0.9335 6867 6.81 <.0001 ** Time2*School Level High 7.3264 0.9384 6867 7.81 <.0001 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.2134 0.4348 6867 -0.49 0.6236 Time2*Minority 1.0957 0.3913 6867 2.8 0.0051 ** Time2*Students with Disabilities 0.4279 0.2986 6867 1.43 0.1518 Time2*Positive Learning Environment 0.9942 0.409 6867 2.43 0.0151 Time2*Positive Teacher Qualifications -0.49 0.2848 6867 -1.72 0.0854 Time3*School Level Elementary -0.8004 0.204 6867 -3.92 <.0001 ** Time3*School Level High -1.5226 0.206 6867 -7.39 <.0001 ** Time3*School Level Middle 0 . . Time3*Free Reduced Lunch 0.13 0.09581 6867 1.36 0.1749 Time3*Minority -0.2709 0.08609 6867 -3.15 0.0017 ** Time3*Students with Disabilities -0.07848 0.06536 6867 -1.2 0.2299 Time3*Positive Learning Environment -0.242 0.09033 6867 -2.68 0.0074 ** Time3*Positive Teacher Qualifications 0.07683 0.06214 6867 1.24 0.2163

PAGE 108

94 Covariance Parameter Estimate SE z p 57.0301 2.1007 27.15 <.0001 ** Residual 13.8759 0.2505 55.39 <.0001 ** Note: p < .05; ** p < .01 When the data were filtered to include only elementary and middle schools and gifted was also added to the equation, all intercept parameter estimates were significant (i.e., elementary school, time, time2, time3, free or reduced lunch status, mi nority, students with disabilitie s, teacher qualifications, and positive learning environment except for LEP). Significan t interactions with time included elementary, free or reduced lunch status, minority, LEP, gifted, and teacher qualificatio ns. Significant interactions with time2 included elementary, minority, LEP, gifted, and positive learning environment. Significant interactions with time3 included elementary, minority, LEP, gifted, and positive learning environment (see Table 12). This model demonstrated better fit than the previous model by all model fit indices (see Table 18). It explained 1% more of the between school variance and the same amount of the within school variance as the previous model and explained 91% of all the variance. Model 5b: School Learning Environment with Demographics by School Level (Elementary and Middle Schools with Gifted) Level 1: FCAT Reading = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16*Gifted + 17* Teacher Qualifications + 18* Positive Learning Environment 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26*Gifted + 27* Teacher Qualifications + 28* Positive Learning Environment 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* LEP + 35* SWD + 36*Gifted + 37* Teacher Qualifications + 38* Positive Learning Environment Mixed-Effects Model: FCAT Reading = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Gifted*Time + 17* Teacher Qualifications*Time + 18* Positive Learning Environment*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Gifted*Time2 + 27* Teacher Qualifications*Time2 + 28* Positive Learning Environment*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34* LEP*Time3 + 35* SWD*Time3 + 36*Gifted*Time3 + 37* Teacher Qualifications*Time3 + 38* Positive Learning Environment*Time3 + u0 + r

PAGE 109

95 Table 12. Model 5b: Reading Predicted by Demographics and St udent Learning Environment by School Level for Elementary and Middle School with Gifted Effect School Level Estimate SE df t p Intercept 680.36 0.3405 1805 1998.4 <.0001 ** Time -20.7804 0.8278 4900 -25.1 <.0001 ** Time2 16.61 0.7109 4900 23.37 <.0001 ** Time3 -3.2426 0.1532 4900 -21.16 <.0001 ** School Level Elementary -26.7137 0.4149 4900 -64.38 <.0001 ** School Level Middle 0 . . Free Reduced Lunch -5.5867 0.183 4900 -30.53 <.0001 ** Minority -5.368 0.2094 4900 -25.63 <.0001 ** LEP -0.5767 0.1649 4900 -3.5 0.0005 ** Students with Disabilities -0.999 0.1273 4900 -7.85 <.0001 ** Gifted 3.2493 0.144 4900 22.56 <.0001 ** Positive Learning Environment 1.6988 0.162 4900 10.48 <.0001 ** Positive Teacher Qualifications 0.6293 0.1072 4900 5.87 <.0001 ** Time*School Level -13.1048 1.0511 4900 -12.47 <.0001 ** Time*School Level 0 . . Time*Free Reduced Lunch 1.5475 0.518 4900 2.99 0.0028 ** Time*Minority 0.2554 0.5008 4900 0.51 0.61 Time*LEP 1.5728 0.3925 4900 4.01 <.0001 ** Time*Students with Disabilities -0.1581 0.3326 4900 -0.48 0.6345 Time*Gifted 0.8477 0.3482 4900 2.43 0.015 Time*Positive Learning Environment -0.5364 0.5209 4900 -1.03 0.3031 Time*Positive Teacher Qualifications 0.8785 0.3134 4900 2.8 0.0051 ** Time2*School Level Elementary 14.9872 0.9131 4900 16.41 <.0001 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -1.3266 0.4599 4900 -2.88 0.0039 ** Time2*Minority 0.5603 0.4378 4900 1.28 0.2008 Time2*LEP -1.3883 0.3428 4900 -4.05 <.0001 ** Time2*Students with Disabilities -0.04735 0.2904 4900 -0.16 0.8705 Time2*Gifted -0.8037 0.304 4900 -2.64 0.0082 ** Time2*Positive Learning Environment 0.2667 0.4604 4900 0.58 0.5624 Time2*Positive Teacher Qualifications -0.7095 0.2752 4900 -2.58 0.01 Time3*School Level Elementary -3.3955 0.1983 4900 -17.12 <.0001 ** Time3*School Level Middle 0 . . Time3*Free Reduced Lunch 0.2666 0.1011 4900 2.64 0.0084 ** Time3*Minority -0.1465 0.09575 4900 -1.53 0.126 Time3*LEP 0.292 0.07472 4900 3.91 <.0001 ** Time3*Students with 0.02797 0.06332 4900 0.44 0.6586

PAGE 110

96 Effect School Level Estimate SE df t p Disabilities Time3*Gifted 0.1601 0.06636 4900 2.41 0.0159 Time3*Positive Learning Environment -0.03561 0.1008 4900 -0.35 0.7239 Time3*Positive Teacher Qualifications 0.1438 0.0602 4900 2.39 0.0169 Covariance Parameter Estimate SE z p 22.7383 0.9189 24.74 <.0001 ** Residual 9.5272 0.1973 48.3 <.0001 ** Note: p < .05; ** p < .01 The next model added technology integration variables with the demographics, learning environment, and school level variables. These included student access to various types of software, teachers regularly using various types of software, frequency that students use various types of software, and technology support. This model was estimated twi ce, first without gifted population but all school levels (see model 6a) and then with elementary and middle school levels and gifted population (see model 6b). When the model was estimated with all school levels without gifted, the only significant technology parameter estimates were frequency that students use tool-based software and the interaction of time, time2, and time3 with teacher’s use of technology for administrative purposes and frequency that students use content software (see Table 13). Other significan t parameter estimates included the intercept, time, time2, time3, high school and elementary school relative to middle school, free or reduced lunch status, minority, LEP, students with disab ilities, positive learning environment, and positive teacher qualifications. Significant interactions with time, time2, and time3 included elementary and high school relative to middle school, free or reduced lunch status, LEP, positiv e learning environment, and positive teacher qualifications. Only one model fit index indicated that this model had better fit (see Table 17). No additional variance was explained with this model. Thre e technology integration indicators were retained in the final model for all school levels without gifted frequency that students use tool-based software, frequency that students use content software, and percen t of teachers who use technology for administrative purposes. Model 6a: Technology Integration with Demographics and Student Learning Environment by School Level (All School Levels without Gifted) Level 1: FCAT Reading = 0 + 1*Time + 2*Time2 + 3*Time3 + r

PAGE 111

97 Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06* Teacher Qualifications + 07* Positive Learning Environment + 08*Access Content SW + 09*Access Office SW + 010*Access Ad Prod SW + 011*Teachers Use Deliver Instruction + 012*Teachers use Admin + 013*Frequency Students Use Content + 014*Frequency Students Use Tool + 015*Technical Support Human + 016*Technical Support Hardware + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16* Teacher Qualifications + 17* Positive Learning Environment + 18*Access Content SW + 19*Access Office SW + 110*Access Ad Prod SW + 111*Teachers Use Deliver Instruction + 112*Teachers use Admin + 113*Frequency Students Use Content + 114*Frequency Students Use Tool + 115*Technical Support Human + 116*Technical Support Hardware 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26* Teacher Qualifications + 27* Positive Learning Environment + 28*Access Content SW + 29*Access Office SW + 210*Access Ad Prod SW + 211*Teachers Use Deliver Instruction + 212*Teachers use Admin + 213*Frequency Students Use Content + 214*Frequency Students Use Tool + 215*Technical Support Human + 216*Technical Support Hardware 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* LEP + 35* SWD + 36* Teacher Qualifications + 37* Positive Learning Environment + 38*Access Content SW + 39*Access Office SW + 310*Access Ad Prod SW + 311*Teachers Use Deliver Instruction + 312*Teachers use Admin + 313*Frequency Students Use Content + 314*Frequency Students Use Tool + 315*Technical Support Human + 316*Technical Support Hardware Mixed-Effects Model: FCAT Reading = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06* Teacher Qualifications + 07* Positive Learning Environment + 08*Access Content SW + 09*Access Office SW + 010*Access Ad Prod SW + 011*Teachers Use Deliver Instruction + 012*Teachers use Admin + 013*Frequency Students Use Content + 014*Frequency Students Use Tool + 015*Technical Support Human + 016*Technical Support Hardware + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Teacher Qualifications*Time + 17* Positive Learning Environment*Time + 18*Access Content SW*Time + 19*Access Office SW*Time + 110*Access Ad Prod SW*Time + 111*Teachers Use Deliver Instruction*Time + 112*Teachers use Admin*Time + 113*Frequency Students Use Content*Time + 114*Frequency Students Use Tool*Time + 115*Technical Support Human*Time + 116*Technical Support Hardware*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Teacher Qualifications*Time2 + 27* Positive Learning Environment*Time2 + 28*Access Content SW*Time2 + 29*Access Office SW*Time2 + 210*Access Ad Prod SW*Time2 + 211*Teachers Use Deliver Instruction*Time2 + 212*Teachers use Admin*Time2 + 213*Frequency Students Use Content*Time2 + 214*Frequency Students Use Tool*Time2 + 215*Technical Support Human*Time2 + 216*Technical Support Hardware*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34* LEP*Time3 + 35* SWD*Time3 + 36*Teacher Qualifications*Time3 + 37* Positive Learning Environment*Time3 + 38*Access Content SW*Time3 + 39*Access Office SW*Time3 + 310*Access Ad Prod SW*Time3 + 311*Teachers Use Deliver Instruction*Time3 + 312*Teachers use Admin*Time3 + 313*Frequency Students Use Content*Time3 + 314*Frequency Students Use Tool*Time3 + 315*Technical Support Human*Time3 + 316*Technical Support Hardware*Time3 + u0 + r

PAGE 112

98 Table 13. Model 6a: Technology Integration w ith Demographics and Student Learni ng Environment by School Level (All School Levels without Gifted) Effect School Level Estimate SE df t p Intercept 681.26 0.3596 2234 1894.5 <.0001 ** Time -20.2469 0.8024 6399 -25.23 <.0001 ** Time2 16.2751 0.6942 6399 23.44 <.0001 ** Time3 -3.1892 0.1502 6399 -21.23 <.0001 ** School Level Elementary -28.0814 0.4283 6399 -65.56 <.0001 ** School Level High 6.6111 0.4874 6399 13.57 <.0001 ** School Level Middle 0 . . Free Reduced Lunch -5.8016 0.174 6399 -33.35 <.0001 ** Minority -5.5842 0.1997 6399 -27.96 <.0001 ** LEP -0.7877 0.1703 6399 -4.63 <.0001 ** Students with Disabilities -1.7237 0.1251 6399 -13.78 <.0001 ** Positive Learning Environment 1.6229 0.1352 6399 12 <.0001 ** Positive Teacher Qualifications 0.8536 0.1002 6399 8.52 <.0001 ** Access Content Software 0.1047 0.09503 6399 1.1 0.2708 Access Office Software -0.1464 0.09431 6399 -1.55 0.1205 Access Advanced Production Software -0.08175 0.09657 6399 -0.85 0.3973 Teachers Use To Deliver Instruction -0.02342 0.1054 6399 -0.22 0.8242 Teachers Use For Administrative Purposes -0.02003 0.1089 6399 -0.18 0.8541 Frequency that Students Use Content Software 0.02092 0.08741 6399 0.24 0.8108 Frequency Students Use Tool-Based Software 0.2425 0.0939 6399 2.58 0.0098 ** Technical Support Human -0.02141 0.08646 6399 -0.25 0.8044 Technical Support Hardware -0.148 0.08354 6399 -1.77 0.0766 Time*School Level Elementary -14.2787 1.0076 6399 -14.17 <.0001 ** Time*School Level High 30.3597 0.9526 6399 31.87 <.0001 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch 1.9416 0.4586 6399 4.23 <.0001 ** Time*Minority 0.8543 0.4402 6399 1.94 0.0523 Time*LEP 0.8362 0.3626 6399 2.31 0.0211 Time*Students with Disabilities -0.4985 0.3066 6399 -1.63 0.104 Time*Positive Learning Environment 0.8869 0.4224 6399 2.1 0.0358 Time*Positive Teacher Qualifications 1.0263 0.2912 6399 3.52 0.0004 ** Time*Access Content Software -0.4158 0.3339 6399 -1.25 0.213

PAGE 113

99 Effect School Level Estimate SE df t p Time*Access Office Software 0.1448 0.3312 6399 0.44 0.6621 Time*Access Advanced Production Software 0.0794 0.3323 6399 0.24 0.8112 Time*Teachers Use To Deliver Instruction -0.1032 0.3698 6399 -0.28 0.7802 Time*Teachers Use For Administrative Purposes 0.7513 0.3767 6399 1.99 0.0461 Time*Frequency that Students Use Content Software -0.6827 0.3238 6399 -2.11 0.035 Time*Frequency Students Use Tool-Based Software 0.2084 0.34 6399 0.61 0.5398 Time*Technical Support Human 0.2347 0.2913 6399 0.81 0.4204 Time*Technical Support Hardware 0.4103 0.2943 6399 1.39 0.1633 Time2*School Level Elementary 15.8282 0.8842 6399 17.9 <.0001 ** Time2*School Level High -18.4859 0.8389 6399 -22.04 <.0001 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -1.8766 0.4082 6399 -4.6 <.0001 ** Time2*Minority 0.2788 0.3856 6399 0.72 0.4696 Time2*LEP -0.728 0.3172 6399 -2.3 0.0217 Time2*Students with Disabilities 0.2934 0.2681 6399 1.09 0.274 Time2*Positive Learning Environment -0.806 0.377 6399 -2.14 0.0326 Time2*Positive Teacher Qualifications -0.7878 0.2558 6399 -3.08 0.0021 ** Time2*Access Content Software 0.2281 0.2931 6399 0.78 0.4365 Time2*Access Office Software -0.02218 0.2899 6399 -0.08 0.939 Time2*Access Advanced Production Software -0.03315 0.2912 6399 -0.11 0.9094 Time2*Teach Use Deliver Instruction 0.1713 0.3258 6399 0.53 0.5992 Time2*Teach Use Administrative Purposes -0.6823 0.3282 6399 -2.08 0.0377 Time2*Frequency Student Use Content Software 0.6071 0.2854 6399 2.13 0.0334 Time2*Frequency Students Use Tool-Based Software -0.2636 0.304 6399 -0.87 0.3859 Time2*Technical Support Human -0.1704 0.2565 6399 -0.66 0.5064 Time2*Technical Support Hardware -0.2141 0.2579 6399 -0.83 0.4066 Time3*School Level Elementary -3.5408 0.193 6399 -18.34 <.0001 ** Time3*School Level High 3.2418 0.184 6399 17.62 <.0001 ** Time3*School Level Middle 0 . . Time3*Free Reduced Lunch 0.3964 0.08998 6399 4.41 <.0001 ** Time3*Minority -0.1027 0.08443 6399 -1.22 0.2237

PAGE 114

100 Effect School Level Estimate SE df t p Time3*LEP 0.1538 0.06936 6399 2.22 0.0266 Time3*Students with Disabilities -0.04336 0.05861 6399 -0.74 0.4594 Time3*Positive Learning Environment 0.1667 0.08311 6399 2.01 0.0449 Time3*Positive Teacher Qualifications 0.1556 0.05592 6399 2.78 0.0054 ** Time3*Access Content Software -0.03266 0.06428 6399 -0.51 0.6115 Time3*Access Office Software -0.00119 0.06339 6399 -0.02 0.985 Time3*Access Advanced Production Software 0.005037 0.06368 6399 0.08 0.937 Time3*Teach Use Deliver Instruction -0.0395 0.07155 6399 -0.55 0.5809 Time3*Teach Use Administrative Purposes 0.1543 0.07158 6399 2.16 0.0311 Time3*Frequency Student Use Content Software -0.1292 0.06252 6399 -2.07 0.0388 Time3*Frequency Students Use Tool-Based Software 0.05051 0.06741 6399 0.75 0.4537 Time3*Technical Support Human 0.02933 0.05634 6399 0.52 0.6026 Time3*Technical Support Hardware 0.02591 0.05654 6399 0.46 0.6468 Covariance Parameter Estimate SE z p 32.0042 1.181 27.1 <.0001 ** Residual 10.1853 0.1875 54.31 <.0001 ** Note: p < .05; ** p < .01 Similar results were found with the elementary and middle school data with gifted. The only significant technology parameter estimates were technical support for hardware and its interaction with time, time2, and time3, and the interaction of time, time2, and time3 with frequency that students use content software and teachers use technology for administ rative purposes (see Table 14). Other significant parameter estimates included the intercept, time, time2, time3, elementary school, free or reduced lunch status, minority, LEP, students with disabilities, gift ed, positive learning enviro nment, and positive teacher qualifications. Significant interactions with time, time2, and time3 included free or reduced lunch status, LEP, gifted, and positive learning environment. Only the -2 Log Likelihood index indicated that this model had better fit (see Table 17). Moreover, adding the technology integration indicators to the model did not explain any additional variance. Technology support for hardware, frequency that students use content software, and teachers’ use of technology for administ rative purposes were the onl y technology integration

PAGE 115

101 indicators retained in the final model for the data with elementary and middle schools and gifted in order to determine if the model fit improved without the noise from the technology integration variables that were not significant. Model 6b: Technology Integration with Demographics and Student Learning Environment by School Level (Elementary and Middle Schools with Gifted) Level 1: FCAT Reading = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 09*Access Content SW + 010*Access Office SW + 011*Access Ad Prod SW + 012*Teachers Use Deliver Instruction + 013*Teachers use Admin + 014*Frequency Students Use Content + 015*Frequency Students Use Tool + 016*Technical Support Human + 017*Technical Support Hardware + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16*Gifted + 17* Teacher Qualifications + 18* Positive Learning Environment + 19*Access Content SW + 110*Access Office SW + 111*Access Ad Prod SW + 112*Teachers Use Deliver Instruction + 113*Teachers use Admin + 114*Frequency Students Use Content + 115*Frequency Students Use Tool + 116*Technical Support Human + 117*Technical Support Hardware 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26*Gifted + 27* Teacher Qualifications + 28* Positive Learning Environment + 29*Access Content SW + 210*Access Office SW + 211*Access Ad Prod SW + 212*Teachers Use Deliver Instruction + 213*Teachers use Admin + 214*Frequency Students Use Content + 215*Frequency Students Use Tool + 216*Technical Support Human + 217*Technical Support Hardware 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* LEP + 35* SWD + 36*Gifted + 37* Teacher Qualifications + 38* Positive Learning Environment + 39*Access Content SW + 310*Access Office SW + 311*Access Ad Prod SW + 312*Teachers Use Deliver Instruction + 313*Teachers use Admin + 314*Frequency Students Use Content + 315*Frequency Students Use Tool + 316*Technical Support Human + 317*Technical Support Hardware Mixed-Effects Model: FCAT Reading = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 09*Access Content SW + 010*Access Office SW + 011*Access Ad Prod SW + 012*Teachers Use Deliver Instruction + 013*Teachers use Admin + 014*Frequency Students Use Content + 015*Frequency Students Use Tool + 016*Technical Support Human + 017*Technical Support Hardware + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Gifted*Time + 17* Teacher Qualifications*Time + 18* Positive Learning Environment*Time + 19*Access Content SW*Time + 110*Access Office SW*Time + 111*Access Ad Prod SW*Time + 112*Teachers Use Deliver Instruction*Time + 113*Teachers use Admin*Time + 114*Frequency Students Use Content*Time + 115*Frequency Students Use Tool*Time + 116*Technical Support Human*Time + 117*Technical Support Hardware*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Gifted*Time2 + 27* Teacher Qualifications*Time2 + 28* Positive Learning Environment*Time2 + 29*Access Content SW*Time2 + 210*Access Office SW*Time2 + 211*Access Ad Prod SW*Time2 + 212*Teachers Use Deliver Instruction*Time2 + 213*Teachers use Admin*Time2 + 214*Frequency Students Use Content*Time2 + 215*Frequency Students Use Tool*Time2 + 216*Technical Support Human*Time2 + 217*Technical Support Hardware*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34* LEP*Time3 + 35* SWD*Time3 + 36*Gifted*Time3 + 37* Teacher Qualifications*Time3 + 38* Positive Learning Environment*Time3 + 39*Access Content SW*Time3 + 310*Access Office SW*Time3 + 311*Access Ad Prod SW*Time3 + 312*Teachers Use Deliver Instruction*Time3 + 313*Teachers use Admin*Time3 + 314*Frequency Students Use Content*Time3 + 315*Frequency Students Use Tool*Time3 + 316*Technical Support Human*Time3 + 317*Technical Support Hardware*Time3 + u0 + r

PAGE 116

102 Table 14. Model 6b: Technology Integration w ith Demographics and Student Learni ng Environment by School Level for Elementary and Middle Schools with Gifted Effect School Level Estimate SE df t p Intercept 680.41 0.3455 1805 1969.4 <.0001 ** Time -21.5066 0.8706 4864 -24.7 <.0001 ** Time2 17.1501 0.7484 4864 22.92 <.0001 ** Time3 -3.3436 0.1612 4864 -20.74 <.0001 ** School Level Elementary -26.7788 0.4228 4864 -63.34 <.0001 ** School Level Middle 0 . . Free Reduced Lunch -5.5813 0.1866 4864 -29.92 <.0001 ** Minority -5.3867 0.2097 4864 -25.68 <.0001 ** LEP -0.593 0.1651 4864 -3.59 0.0003 ** Students with Disabilities -0.9745 0.1272 4864 -7.66 <.0001 ** Gifted 3.2711 0.1447 4864 22.61 <.0001 ** Positive Learning Environment 1.6996 0.1626 4864 10.46 <.0001 ** Positive Teacher Qualifications 0.6139 0.1071 4864 5.73 <.0001 ** Access Content Software 0.01653 0.1046 4864 0.16 0.8744 Access Office Software -0.07994 0.09993 4864 -0.8 0.4238 Access Advanced Production Software -0.1018 0.1046 4864 -0.97 0.3304 Teachers Use To Deliver Instruction -0.00444 0.1152 4864 -0.04 0.9693 Teachers Use For Administrative Purposes -0.09236 0.1209 4864 -0.76 0.4449 Frequency that Students Use Content Software 0.1639 0.09642 4864 1.7 0.0892 Frequency Students Use ToolBased Software 0.1914 0.1011 4864 1.89 0.0583 Technical Support Human -0.04484 0.09504 4864 -0.47 0.6371 Technical Support Hardware -0.2621 0.0894 4864 -2.93 0.0034 ** Time*School Level Elementary -12.1692 1.1153 4864 -10.91 <.0001 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch 1.5767 0.5292 4864 2.98 0.0029 ** Time*Minority 0.4593 0.5057 4864 0.91 0.3638 Time*LEP 1.5244 0.3937 4864 3.87 0.0001 ** Time*Students with Disabilities -0.104 0.3333 4864 -0.31 0.7551 Time*Gifted 0.8081 0.3522 4864 2.29 0.0218 Time*Positive Learning Environment -0.6142 0.5243 4864 -1.17 0.2415 Time*Positive Teacher Qualifications 0.8703 0.3173 4864 2.74 0.0061 ** Time*Access Content Software -0.2275 0.3654 4864 -0.62 0.5335 Time*Access Office Software -0.1091 0.3536 4864 -0.31 0.7577 Time*Access Advanced Production Software 0.2427 0.3576 4864 0.68 0.4974 Time*Teachers Use To Deliver Instruction -0.09376 0.4085 4864 -0.23 0.8185 Time*Teachers Use For 0.9038 0.421 4864 2.15 0.0319

PAGE 117

103 Effect School Level Estimate SE df t p Administrative Purposes Time*Frequency that Students Use Content Software -0.7795 0.3647 4864 -2.14 0.0326 Time*Frequency Students Use Tool-Based Software -0.08583 0.3719 4864 -0.23 0.8175 Time*Technical Support Human 0.565 0.3237 4864 1.75 0.081 Time*Technical Support Hardware 0.6596 0.3189 4864 2.07 0.0387 Time2*School Level Elementary 14.3156 0.9724 4864 14.72 <.0001 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -1.3937 0.4705 4864 -2.96 0.0031 ** Time2*Minority 0.4285 0.4423 4864 0.97 0.3327 Time2*LEP -1.3274 0.3441 4864 -3.86 0.0001 ** Time2*Students with Disabilities -0.1044 0.291 4864 -0.36 0.7197 Time2*Gifted -0.7675 0.3071 4864 -2.5 0.0125 Time2*Positive Learning Environment 0.3074 0.4632 4864 0.66 0.507 Time2*Positive Teacher Qualifications -0.6932 0.2802 4864 -2.47 0.0134 Time2*Access Content Software 0.09879 0.321 4864 0.31 0.7583 Time2*Access Office Software 0.2076 0.3092 4864 0.67 0.5019 Time2*Access Advanced Production Software -0.1561 0.314 4864 -0.5 0.619 Time2*Teach Use Deliver Instruction 0.1045 0.3596 4864 0.29 0.7713 Time2*Teach Use Administrative Purposes -0.749 0.3647 4864 -2.05 0.0401 Time2*Frequency Student Use Content Software 0.6575 0.3219 4864 2.04 0.0411 Time2*Frequency Students Use Tool-Based Software -0.00448 0.3324 4864 -0.01 0.9892 Time2*Technical Support Human -0.4997 0.284 4864 -1.76 0.0786 Time2*Technical Support Hardware -0.3582 0.279 4864 -1.28 0.1993 Time3*School Level Elementary -3.274 0.2114 4864 -15.49 <.0001 ** Time3*School Level Middle 0 . . Time3*Free Reduced Lunch 0.2851 0.1035 4864 2.75 0.0059 ** Time3*Minority -0.1216 0.09674 4864 -1.26 0.2089 Time3*LEP 0.2764 0.07505 4864 3.68 0.0002 ** Time3*Students with Disabilities 0.04149 0.06343 4864 0.65 0.5131 Time3*Gifted 0.1508 0.06696 4864 2.25 0.0243 Time3*Positive Learning Environment -0.04003 0.1014 4864 -0.39 0.6929 Time3*Positive Teacher Qualifications 0.1411 0.06147 4864 2.29 0.0218 Time3*Access Content Software -0.00749 0.07047 4864 -0.11 0.9154 Time3*Access Office Software -0.05101 0.0675 4864 -0.76 0.4499 Time3*Access Advanced Production Software 0.02369 0.06879 4864 0.34 0.7306 Time3*Teach Use Deliver Instruction -0.02464 0.07886 4864 -0.31 0.7547

PAGE 118

104 Effect School Level Estimate SE df t p Time3*Teach Use Administrative Purposes 0.1622 0.07922 4864 2.05 0.0407 Time3*Frequency Student Use Content Software -0.1381 0.07048 4864 -1.96 0.0502 Time3*Frequency Students Use Tool-Based Software -0.00025 0.07361 4864 0 0.9973 Time3*Technical Support Human 0.1054 0.06221 4864 1.69 0.0902 Time3*Technical Support Hardware 0.05083 0.06104 4864 0.83 0.4051 Covariance Parameter Estimate SE z p 22.5852 0.9142 24.7 <.0001 ** Residual 9.4602 0.196 48.26 <.0001 ** Note: p < .05; ** p < .01 The last models that were estimated in order to answer the first hypothesis included all school levels, demographic, learning environment quality, and significant technology integration variables. These models were different because the model fit to the data for all schools levels without gifted included three technology integration variables frequency that students use tool-based software, frequency that students use content software, and percent of teachers who regularly use technology for administrative purposes (see model 7a); while the model fitted to the data with elementary and middle school levels and gifted included three technology integration variables – frequency that students use content software, level of technical support – hardware, and percent of teachers who regularly use technology for administrative purposes (see model 7b). For the model with all schools levels and no gifted, the same parameter estimates and interactions identified in the previous models as sign ificant were significant again (see Table 15). Although there was no difference in the percentage of variance explained in this model than was in the Demographic Model with Student Learning Environment by school level or the Technology Integration with Demographic and Student Learning Environment Model by school level, the AIC, AICC, and BIC indices all indicated better model fit (see Table 17). The level-1 residuals for the final model for predicting FCAT Reading using all school levels without gifted ranged between -15.18 and 17.40, with a standard deviation of 2.79. Although there were outliers, skewness was 0. 10 and kurtosis was 1.86, which would indicate that the residuals were evenly distributed with most ar ound the mean. Distributi on of the empirical bayes

PAGE 119

105 intercepts ranged between -19.35 and 29.70 with standard deviation of 5.43. Skewness was 0.55, and kurtosis was 1.25, which indicated that the residuals at level-2 were within acceptable range. Final Model 7a: Significant Technology Integration Indicators with Demographics and Student Learning Environment by School Level (All School Levels without Gifted) Level 1: FCAT Reading = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Teacher Qualifications + 07* Positive Learning Environment + 08*Frequency Students Use Tool + 09*Frequency Students Use Content + 010*Teachers use Admin + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16* Teacher Qualifications + 17* Positive Learning Environment + 18*Frequency Students Use Tool + 19*Frequency Students Use Content + 110*Teachers use Admin 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26* Teacher Qualifications + 27* Positive Learning Environment + 28*Frequency Students Use Tool + 29*Frequency Students Use Content + 210*Teachers use Admin 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* LEP + 35* SWD + 36* Teacher Qualifications + 37* Positive Learning Environment + 38*Frequency Students Use Tool + 39*Frequency Students Use Content + 310*Teachers use Admin Mixed-Effects Model: FCAT Reading = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06* Teacher Qualifications + 07* Positive Learning Environment + 08*Frequency Students Use Tool + 09*Frequency Students Use Content + 010*Teachers use Admin + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Teacher Qualifications*Time + 17* Positive Learning Environment*Time 18*Frequency Students Use Tool*Time + 19*Frequency Students Use Content*Time + 110*Teachers use Admin *Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Teacher Qualifications*Time2 + 27* Positive Learning Environment*Time2 + 28*Frequency Students Use Tool*Time2 + 29*Frequency Students Use Content*Time2 + 210*Teachers use Admin*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34* LEP*Time3 + 35* SWD*Time3 + 36*Teacher Qualifications*Time3 + 37* Positive Learning Environment*Time3 + 38*Frequency Students Use Tool*Time3 + 39*Frequency Students Use Content*Time3 + 310*Teachers use Admin*Time3 + u0 + r Table 15. Final Model 7a: Significant Technology Integration with Demographics and Student Learning Environment by School Level (A ll School Levels without Gifted) Effect School Level Estimate SE df t p Intercept 681.21 0.3577 2234 1904.5 <.0001 ** Time -20.094 0.7922 6423 -25.37 <.0001 ** Time2 16.2106 0.6848 6423 23.67 <.0001 ** Time3 -3.1829 0.1481 6423 -21.49 <.0001 ** School Level Elementary -28.0224 0.4253 6423 -65.89 <.0001 ** School Level High 6.5409 0.4864 6423 13.45 <.0001 ** School Level Middle 0 . . Free Reduced Lunch -5.8135 0.1729 6423 -33.63 <.0001 ** Minority -5.563 0.1995 6423 -27.89 <.0001 **

PAGE 120

106 Effect School Level Estimate SE df t p LEP -0.7833 0.1698 6423 -4.61 <.0001 ** Students with Disabilities -1.7282 0.125 6423 -13.83 <.0001 ** Positive Learning Environment 1.6252 0.135 6423 12.04 <.0001 ** Positive Teacher Qualifications 0.865 0.1 6423 8.65 <.0001 ** Teachers Use For Administrative Purposes -0.06951 0.09241 6423 -0.75 0.452 Frequency that Students Use Content Software 0.04739 0.08422 6423 0.56 0.5737 Frequency Students Use ToolBased Software 0.219 0.09025 6423 2.43 0.0153 Time*School Level Elementary -14.4561 0.9915 6423 -14.58 <.0001 ** Time*School Level High 30.6031 0.947 6423 32.32 <.0001 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch 2.0013 0.4562 6423 4.39 <.0001 ** Time*Minority 0.7906 0.4385 6423 1.8 0.0714 Time*LEP 0.8464 0.361 6423 2.34 0.0191 Time*Students with Disabilities -0.4998 0.3062 6423 -1.63 0.1027 Time*Positive Learning Environment 0.8842 0.4209 6423 2.1 0.0357 Time*Positive Teacher Qualifications 1.0424 0.2879 6423 3.62 0.0003 ** Time*Teachers Use For Administrative Purposes 0.7114 0.3177 6423 2.24 0.0252 Time*Frequency that Students Use Content Software -0.7688 0.3115 6423 -2.47 0.0136 Time*Frequency Students Use Tool-Based Software 0.1663 0.323 6423 0.51 0.6068 Time2*School Level Elementary 15.8783 0.868 6423 18.29 <.0001 ** Time2*School Level High -18.6382 0.834 6423 -22.35 <.0001 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -1.912 0.4063 6423 -4.71 <.0001 ** Time2*Minority 0.3215 0.3842 6423 0.84 0.4027 Time2*LEP -0.74 0.3156 6423 -2.34 0.0191 Time2*Students with Disabilities 0.2905 0.2678 6423 1.08 0.278 Time2*Positive Learning Environment -0.7948 0.3757 6423 -2.12 0.0344 Time2*Positive Teacher Qualifications -0.8076 0.2518 6423 -3.21 0.0013 ** Time2*Teachers Use For Administrative Purposes -0.5955 0.2784 6423 -2.14 0.0325 Time2*Frequency that Students Use Content Software 0.6547 0.2757 6423 2.37 0.0176 Time2*Frequency Students Use Tool-Based Software -0.201 0.2889 6423 -0.7 0.4866 Time3*School Level Elementary -3.5395 0.1892 6423 -18.7 <.0001 ** Time3*School Level High 3.2667 0.183 6423 17.85 <.0001 ** Time3*School Level Middle 0 . .

PAGE 121

107 Effect School Level Estimate SE df t p Time3*Free Reduced Lunch 0.4009 0.08959 6423 4.47 <.0001 ** Time3*Minority -0.1104 0.08411 6423 -1.31 0.1892 Time3*LEP 0.1565 0.06899 6423 2.27 0.0233 Time3*Students with Disabilities -0.04226 0.05853 6423 -0.72 0.4704 Time3*Positive Learning Environment 0.1635 0.08283 6423 1.97 0.0484 Time3*Positive Teacher Qualifications 0.1596 0.05491 6423 2.91 0.0037 ** Time3*Teachers Use For Administrative Purposes 0.1337 0.061 6423 2.19 0.0284 Time3*Frequency that Students Use Content Software -0.1364 0.06052 6423 -2.25 0.0242 Time3*Frequency Students Use Tool-Based Software 0.03707 0.0641 6423 0.58 0.5631 Covariance Parameter Estimate SE z p 31.9887 1.1769 27.18 <.0001 ** Residual 10.2201 0.1879 54.38 <.0001 ** Note: p < .05; ** p < .01 For the model with elementary and middle school levels and gifted, the same significant parameter estimates were identified as in the previous models; however, the significant parameters for technology indicators changed (see Table 16). Both frequency that students use content software and technical support for hardware were significant. Inte ractions between time and teachers’ us e of technology for administrative purposes, frequency that students use content software and technical support for hardware were significant. Interactions between time2 and teachers’ use of technology for ad ministrative purposes and frequency that students use content software were sign ificant. The interaction between time3 and teachers use technology for administrative purposes was significant. Although th ere was no difference in the percentage of variance explained in this model than was in the Demographic Model with Student Learning Environment by school level or the Technology Integration with Demographic and Student Learning Environment Model by school level, the AIC, AICC, and BIC indices all indicated better model fit (see Table 18). The level-1 residuals for the final model for predicting FCAT Reading using elementary and middle schools with gifted ranged between -13.39 and 16.99 with a standard deviation of 2.69. Although there were outliers, skewness was 0.08 and kurtosis was 1.76, which would indicate that the residuals were evenly distributed with most around the mean. Distribution of the empirical bayes intercepts ranged between -13.91 and 20.73 with

PAGE 122

108 standard deviation of 4.50. Skewness was 0.32, and kurtosis was 0.31, which indicated that the residuals at level-2 were also normally distributed. Final Model 7b: Technology Integration with Demographics and Student Learning Environment by School Level (Elementary and Middle Schools with Gifted) Level 1: FCAT Reading = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 09*Technical Support Hardware + 010*Frequency Students Use Content + 011*Teachers use Admin + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16*Gifted + 17* Teacher Qualifications + 18* Positive Learning Environment + 19*Technical Support Hardware + 110*Frequency Students Use Content + 111*Teachers use Admin 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26*Gifted + 27* Teacher Qualifications + 28* Positive Learning Environment + 29*Technical Support Hardware + 210*Frequency Students Use Content + 211*Teachers use Admin 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* LEP + 35* SWD + 36*Gifted + 37* Teacher Qualifications + 38* Positive Learning Environment + 39*Technical Support Hardware + 310*Frequency Students Use Content + 311*Teachers use Admin Mixed-Effects Model: FCAT Reading = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 09*Technical Support Hardware + 010*Frequency Students Use Content + 011*Teachers use Admin + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Gifted*Time + 17* Teacher Qualifications*Time + 18* Positive Learning Environment*Time + 19*Technical Support Hardware*Time + 110*Frequency Students Use Content*Time + 111*Teachers use Admin*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Gifted*Time2 + 27* Teacher Qualifications*Time2 + 28* Positive Learning Environment*Time2 + 29*Technical Support Hardware*Time2 + 210*Frequency Students Use Content*Time2 + 211 Teachers use Admin*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34* LEP*Time3 + 35* SWD*Time3 + 36*Gifted*Time3 + 37* Teacher Qualifications*Time3 + 38* Positive Learning Environment*Time3 + 39*Technical Support Hardware*Time3 + 310*Frequency Students Use Content*Time3 + 311 Teachers use Admin*Time3 + u0 + r Table 16. Final Model 7b: Significant Technology Integration with Demographics and Student Learning Environment by School Level for Elemen tary and Middle Schools with Gifted Effect School Level Estimate SE df t p Intercept 680.44 0.3429 1805 1984.5 <.0001 ** Time -21.3167 0.8502 4888 -25.07 <.0001 ** Time2 17.0195 0.7289 4888 23.35 <.0001 ** Time3 -3.3238 0.1568 4888 -21.19 <.0001 ** School Level Elementary -26.8121 0.4187 4888 -64.04 <.0001 ** School Level Middle 0 . . Free Reduced Lunch -5.5684 0.1861 4888 -29.92 <.0001 ** Minority -5.399 0.2096 4888 -25.75 <.0001 ** LEP -0.5986 0.1647 4888 -3.63 0.0003 ** Students with Disabilities -0.9883 0.1271 4888 -7.78 <.0001 **

PAGE 123

109 Effect School Level Estimate SE df t p Gifted 3.2803 0.1444 4888 22.71 <.0001 ** Positive Learning Environment 1.7124 0.1621 4888 10.56 <.0001 ** Positive Teacher Qualifications 0.6145 0.1071 4888 5.74 <.0001 ** Teachers Use For Administrative Purposes -0.09228 0.09802 4888 -0.94 0.3465 Frequency that students use content software 0.1809 0.09183 4888 1.97 0.0489 Technical Support Hardware -0.2759 0.0889 4888 -3.1 0.0019 ** Time*School Level Elementary -12.3938 1.0834 4888 -11.44 <.0001 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch 1.6201 0.5246 4888 3.09 0.002 ** Time*Minority 0.4683 0.5034 4888 0.93 0.3522 Time*LEP 1.568 0.3919 4888 4 <.0001 ** Time*Students with Disabilities -0.1125 0.3323 4888 -0.34 0.735 Time*Gifted 0.8005 0.3504 4888 2.28 0.0224 Time*Positive Learning Environment -0.604 0.5226 4888 -1.16 0.2479 Time*Positive Teacher Qualifications 0.916 0.314 4888 2.92 0.0035 ** Time*Teachers Use For Administrative Purposes 0.8386 0.3462 4888 2.42 0.0155 Time*Frequency that students use content software -0.7416 0.3412 4888 -2.17 0.0298 Time*Technical Support Hardware 0.7107 0.3178 4888 2.24 0.0254 Time2*School Level Elementary 14.4502 0.9392 4888 15.39 <.0001 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -1.4392 0.4659 4888 -3.09 0.002 ** Time2*Minority 0.4136 0.4397 4888 0.94 0.3469 Time2*LEP -1.3702 0.3424 4888 -4 <.0001 ** Time2*Students with Disabilities -0.09047 0.2901 4888 -0.31 0.7552 Time2*Gifted -0.7713 0.3056 4888 -2.52 0.0116 Time2*Positive Learning Environment 0.3029 0.4616 4888 0.66 0.5116 Time2*Positive Teacher Qualifications -0.747 0.2762 4888 -2.7 0.0069 ** Time2*Teachers Use For Administrative Purposes -0.6818 0.3031 4888 -2.25 0.0245 Time2*Frequency that students use content software 0.6093 0.3003 4888 2.03 0.0425 Time2*Technical Support Hardware -0.3993 0.2781 4888 -1.44 0.1511 Time3*School Level Elementary -3.2895 0.2036 4888 -16.16 <.0001 ** Time3*School Level Middle 0 . . Time3*Free Reduced Lunch 0.297 0.1025 4888 2.9 0.0038 ** Time3*Minority -0.1191 0.09609 4888 -1.24 0.2154 Time3*LEP 0.2867 0.07464 4888 3.84 0.0001 ** Time3*Students with 0.03779 0.06325 4888 0.6 0.5503

PAGE 124

110 Effect School Level Estimate SE df t p Disabilities Time3*Gifted 0.153 0.06665 4888 2.3 0.0217 Time3*Positive Learning Environment -0.04094 0.101 4888 -0.41 0.6852 Time3*Positive Teacher Qualifications 0.1531 0.06046 4888 2.53 0.0114 Time3*Teachers Use For Administrative Purposes 0.146 0.0662 4888 2.2 0.0275 Time3*Frequency that students use content software -0.1278 0.06566 4888 -1.95 0.0517 Time3*Technical Support Hardware 0.05961 0.06082 4888 0.98 0.327 Covariance Parameter Estimate SE z p 22.6794 0.9161 24.76 <.0001 ** Residual 9.4785 0.1962 48.3 <.0001 ** Note: p < .05; ** p < .01 The last step was to add in USDOE funded Magnet Schools and USDOE Technology Magnet Schools as variables in the model. The USDOE funded Magnet schools were a collection of schools that were suggested to have high levels of technology infrastructure and high levels of staff development that included integrating technology into instruction. These schools were to be used as a proxy for schools that had the highest levels of technology integration over the longest period of time. Results of this model indicated that neither magnet school status nor technology magnet school status was a significant predictor of FCAT Reading. Table 17. Model Fit Indices for Models Predicting FCAT Readi ng Scores for All School Levels (without Gifted) Model -2 Log Likelihood AIC (smaller is better) AICC (smaller is better) BIC (smaller is better) Model 1: Reading Predicted by Average Reading of All Schools in Florida 68953.7 68959.7 68959.7 68976.9 Model 2a: Time as a Predictor of Reading 67756.5 67766.5 67766.5 67795.1 Model 2b: Time as a Predictor of Reading Time Fixed 67758.7 67766.7 67766.7 67789.6 Quadratic Model 2c: Time2 as a Predictor of Reading 67239.8 67249.8 67249.9 67278.5 Polynomial Model 2d: Time3 as a Predictor of Reading 63024.7 63040.7 63040.7 63086.5 Model 3: Time, Time2, Time3, and School Level as Predictors of Reading 57367.6 57395.6 57395.6 57475.8

PAGE 125

111 Model -2 Log Likelihood AIC (smaller is better) AICC (smaller is better) BIC (smaller is better) Model 4a: Reading predicted by Time, School Level, and Demographics Variables 51149.2 51209.2 51209.4 51380.5 Model 5a: Demographics and Student Learning Environment by School Level 50716.9 50792.9 50793.2 51010 Model 6a: Technology Integration with Demographics and Student Learning Environment by School Level 50663.7 50811.7 50812.9 51234.4 Final Model 7a: Significant Technology Integration with Demographics and Student Learning Environment by School Level 50685.3 50785.3 50785.9 51070.9 Table 18. Model Fit Indices for Models Predicting FCAT Reading Scores for Elementary and Middle School Levels (with Gifted) Model -2 Log Likelihood AIC (smaller is better) AICC (smaller is better) BIC (smaller is better) Model 1: Reading Predicted by Average Reading of All Elementary and Middle Schools in Florida 57659.6 57665.6 57665.6 57682.3 Model 4b: Reading predicted by Time, School Level, and Demographics Variables No High School includes gifted 38677.4 38737.4 38737.7 38902.4 Model 5b: Demographics and Teacher Qualifications by School Level 38424.4 38500.4 38500.9 38709.4 Model 6b: Technology Integration with Demographics and Teacher Qualifications by School Level 38377.3 38525.3 38527 38932.3 Final Model 7b: Significant Technology Integration with Demographics and Student Learning Environment by School Level 38393.9 38493.9 38494.7 38768.9 The result of the analysis for all the models indicated that Hypothesis 1 was partially correct. When the sample included schools at all three school levels, there was a significant positive relationship between the frequency that students use tool-based software and school level FCAT reading achievement when all other school level, demographic, and school learning environment factors were controlled. Also, there were significant interactions between technology integration variables and time, time2, and time3 with FCAT reading achievement. There was a significant positive interaction between time and the percent of

PAGE 126

112 teachers who regularly use technology for administrativ e purposes with FCAT r eading achievement, and a significant negative interaction between time and the frequency that students use content software with FCAT reading achievement. Time2 and the frequency that students use content software had a significant positive interaction with FCAT reading achievement, and time2 and the percent of teachers who regularly use technology for administrative purposes had a significant negative interaction with FCAT reading achievement. Time3 and percent of teachers who regularly use technology for administrative purposes had a significant positive interaction with FCAT reading achievement, and time3 and the frequency that students use content software had a significant negative in teraction with FCAT reading achievement. These interactions resulted in a curvilinear trend. After controlling so that all other variables were held at the mean, the trend for each school level could be examined separately, by comparing schools with different levels that students use tool-based software. Figure 5 illustrates the relationship between the average school frequency that students use toolbased software and average school FCAT Reading score for high schools. Frequencies that their students use tool-based software were compared at one and tw o standard deviations below the mean, the mean, and one and two standard deviations above the mean. This allowed the extreme cases of schools that had their students use tool-based software the most often, +2 standard deviations above the mean, and schools that had their students use tool-based software the least often, -2 standard deviations below the mean, to be compared. Schools that had students use software the most often started the study in 2003-04 with the highest FCAT Reading scores (688) and schools that ha d their students use the software the least often had started with the lowest FCAT Readin g scores (687). This difference of one point was significant because there were so many schools in the sample; however, the practical importance was modest. The interaction between the frequency that students use tool-based software and time, time2, and time3 with FCAT Reading scores was not significant, so the slopes of the trends at each level of use were the same. By 2005-06, all high schools gained in their average FCAT reading scores, and all high schools had the same FCAT Reading scores, no matter how frequently their students used the software.

PAGE 127

113 Relationship between Frequency Students Use Tool-base Software and FCAT Reading in High Schools686 688 690 692 694 696 698 700 702 2003200420052006 YearFCAT Reading High Schools + 2SD High Schools + 1SD High Schools mean High Schools 1SD High Schools 2SD Figure 5. Relationship between Frequency Students Use Tool-base Software and FCAT Reading in High Schools. Middle schools had a similar beginning pattern to high school; that is, after controlling for all other factors, schools that were two standard deviations above the mean in the frequency that their students used tool-based software had the highest FCAT Reading scores in 2003-04 (682), while those with two standard deviations below the mean had the lowest scores (681). Although this difference of one point was significant due to the large sample size, the prac tical importance is modest. Because there were no significant interactions between time, time2, time3 and the frequency that middle schools have their students use tool-based software with FCAT Reading scores, th e trends were similar for all middle schools. Between 2003-04 and 2004-05 all middle schools had a decrease in their FCAT Reading scores (+2SD = 675 and +1SD, Mean, -1SD, and -2SD = 674). After this, all sc hools increased their FCAT Reading score each year to 680 in 2005-06 and 681 in 2006-07. Again, similar to high school, middle schools, at all five levels of frequency that students use tool-based software, ended with the same score.

PAGE 128

114 Relationship between Frequency Students Use Tool-base Software and FCAT Reading in Middle Schools673 674 675 676 677 678 679 680 681 682 683 2003200420052006 YearFCAT Reading Middle Schools + 2SD Middle Schools + 1SD Middle Schools mean Middle Schools 1SD Middle Schools 2SD Figure 6. Relationship between Frequency Students Use Tool-base Software and FCAT Reading in Middle Schools. Elementary schools experienced a similar pattern to middle schools. Schools with the highest frequency of students using tool-based software began the study with the highest FCAT Reading scores (654), while all other standard deviations or levels had the same score (653). Although this difference of one point was statistically significant, it has no practical importance. Between 2003-04 and 2004-05, all elementary schools experienced a decline in FCAT Reading scores (644), and then the trend reversed to 659 for all levels of frequency of use of tool-based software in 2005-06, followed by a slight decline (657) in 2006-07. In 2004-05, 2005-06, and 2006-07, all elementary schools at all levels of frequency that students use tool-based software had the same average FCAT Reading score.

PAGE 129

115 Relationship between Frequency Students Use Tool-base Software and FCAT Reading in Elementary Schools642 644 646 648 650 652 654 656 658 660 2003200420052006 YearFCAT Reading Elementary + 2SD Elementary + 1SD Elementary mean Elementary 1SD Elementary 2SD Figure 7. Relationship between Frequency Students Use Tool-base Software and FCAT Reading in Elementary Schools. When the sample was restricted to just elementary and middle schools and percent of gifted students was included in the equation, there was a main effect with gifted, but no interactions of percent of gifted students in the school with time, time2, or time3. Thus, when all other factors were held equal, schools with highest percentages of gifted students began the study with the highest FCAT Reading scores, and this trend did not change over time (see Figure 8).

PAGE 130

116 Relationship between Percent of Gifted Students on FCAT Reading by School Level (Gifted Included)635 640 645 650 655 660 665 670 675 680 685 2003200420052006 YearFCAT Reading Elementary + 2SD Elementary + 1SD Elementary mean Elementary 1SD Elementary 2SD Middle Schools + 2SD Middle Schools + 1SD Middle Schools mean Middle Schools 1SD Middle Schools 2SD Figure 8. Relationship between Percent of Gifted Studen ts on FCAT Reading by School Level (Gifted Included). When examining the parameter estimates of the technology integration indicators within these data, there were significant main effects for relati onships between three different variables and FCAT Reading scores: the percent of teachers who regular ly use technology for administrative purposes, frequency that students use content software, and the level of technical support for hardware. In addition, these three technology integration indicators and tim e had significant interactions with FCAT reading scores. The interactions between time2 and the percentage of teachers who regularly use technology for administrative purposes and the level of technical s upport for hardware with FCAT Reading scores were significant. There were no significant interactions between any technology integration variables and time3. In order to visualize the significant relationships of each of these technology integration variables with FCAT reading, the trends are depicted in separa te charts after controlling for all other factors. There was a significant interaction with time and the percentage of teachers who regularly use technology for administrative purposes and FCAT reading scores. Each school level was examined

PAGE 131

117 separately. One and two standard deviations above the mean, the mean, and one and two standard deviations below the mean of levels of percenta ges of teachers who regularly use technology for administrative purposes were compared after controlling for all other factors. In 2003-04, all middle schools started with the same average FCAT Reading sc ore (675) (see Figure 9). The scores for all middle schools decreased in 2004-05, with middle schools that were two standard deviations above the mean in percentages of teachers who regularly used technology for administrative purposes having the least decline (668), while all other levels had the same score (667). Although this one point difference was significant because there were so many schools in the sample, it did not have practical importance. The trend for all schools was up (674) in 2005-06. In 2006-07, schools at one and two standard deviations above the mean increased to (675), while schools at the mean and one and two standard deviations below the mean for percentages of teachers who use technology for administ rative purposes remained at the same score (674). Again, this significant difference of one point did not have practical importance. Relationship between Percent of Teachers Use Technology for Administrative Purposes and FCAT Reading in Middle Schools (Gifted Included)666 667 668 669 670 671 672 673 674 675 676 2003200420052006 YearFCAT Reading Middle Schools + 2SD Middle Schools + 1SD Middle Schools mean Middle Schools 1SD Middle Schools 2SD Figure 9. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and FCAT Reading in Middle Schools (Gifted Included).

PAGE 132

118 The trends for elementary schools followed a very similar pattern to that of middle schools (see Figure 10). In 2003-04, elementary schools with two standard deviations above th e mean of percentage of teachers who regularly use technology for administrativ e purposes were predicted to have one point lower (653) than elementary schools at all other standard deviations (654). Although this difference was significant, it did not have practical importance. In 2004-05 the average school FCAT reading score declined; however, schools with two standard deviatio ns below the mean for percentage of teachers who regularly use technology for administrative purposes were predicted to decline the most (644), while schools with all other standard deviations were predicted to decline the least (645). The trend for elementary schools with all levels of percenta ge of teachers who regularly use technology for administrative purposes was up to 659 in 2005-06. In 2006-07, there was a decline to 658 for schools with two standard deviations above the mean of percen tage of teachers who regul arly use technology for administrative purposes, while all other levels declined to 657. Relationship between Percent of Teachers Use Technology for Administrative Purposes and FCAT Reading in Elementary Schools (Gifted Included)642 644 646 648 650 652 654 656 658 660 2003200420052006 YearFCAT Reading Elementary + 2SD Elementary + 1SD Elementary mean Elementary 1SD Elementary 2SD Figure 10. Relationship between Percent of Teachers Wh o Regularly Use Technol ogy for Administrative Purposes and FCAT Reading in Elementary Schools (Gifted Included).

PAGE 133

119 The interaction between time and the frequency th at students use content software with FCAT Reading score for elementary schools and middle schools with gifted was significant. The interactions with time2 and time3 were not significant. Both levels of school ha ve similar trends. Charts were made for each level of school to visualize the relationship between the frequency that students use content software and FCAT Reading achievement at one and two standard deviations above the mean, the mean, and one and two standard deviations below the mean. The trends for middle school level at two standard deviations above the mean, the mean, and two standard deviations below the mean of frequency that students use content software were examined (see Figure 11). When controlling for all other variables, mi ddle schools at all standard deviations of frequency that students use content software had the same FCAT Reading scores at each po int of time. In 2003-04, middle schools had FCAT scores at 674. FCAT Reading scores declined to 667 in 2004-05, and then rebounded to 674 in both 2005-06 and 2006-07. Relationship between Frequency Students Use Content Software and FCAT Reading in Middle Schools (Gifted Included)666 667 668 669 670 671 672 673 674 675 676 2003200420052006 YearFCAT Reading Middle Schools + 2SD Middle Schools + 1SD Middle Schools mean Middle Schools 1SD Middle Schools 2SD Figure 11. Relationship between Frequency that Students Use Content Software and FCAT Reading in Middle Schools (Gifted Included).

PAGE 134

120 When controlling for all other variables, elementary schools at one and two standard deviations above the mean and at the mean for frequency that students use content software started with FCAT Reading scores at 654, while schools at one and two standard deviations below the mean started with scores at 653 in 2003-04. At all other points in time, all levels of frequency that students use content software had the same FCAT Reading scores. Scores declined between 2003-04 and 2004-05 (645), then rebounded in 2005-06 (667), and declined slightly in 2006-07 (657). Relationship between Frequency Students Use Content Software and FCAT Reading in Elementary Schools (gifted included)642 644 646 648 650 652 654 656 658 660 662 2003200420052006 YearFCAT Reading Elementary + 2SD Elementary + 1SD Elementary mean Elementary 1SD Elementary 2SD Figure 12. Relationship between Frequency that Students Use Content Software and FCAT Reading in Elementary Schools (Gifted Included). The third significant relationship between techno logy integration variables and school FCAT Reading score was with the level of technical support for hardware. The interactions between both time and time2 and the level of technical support for hardware and FCAT Reading scores were also significant. The interaction with time3 was not significant. Elementary and middle schools have very similar trends. When examining middles schools with one and two standard deviations above the mean, the mean, and one and two standard deviations below the mean in level of technology support for hardware after controlling for all

PAGE 135

121 other variables, the FCAT Reading scores at each level at each point of time were the same (675 in 200304, 667 in 2004-05, and 674 in 2005-06 and 2005-07) (see Figure 13). After controlling for all other variables, elementary schools with one and two standa rd deviations above the mean and at the mean in level of technology support for hardware had a beginning FCAT Reading score in 2003-04 of 654, while elementary schools with one and two standard deviations below the mean had a beginning FCAT Reading score of 653 (see Figure 14). All FCAT Reading scores for all levels of technology support for hardware declined to 645 in 2004-05, and then rebounded to 659 in 2005-06 with a decline to 657 in 2006-07. There are no practical differences in FCAT Reading scores related to the leve l of technical support for hardware at either the middle or elementary school levels. Relationship between Technical Support Hardware and FCAT Reading in Middle Schools (Gifted Included)666 667 668 669 670 671 672 673 674 675 676 2003200420052006 YearFCAT Reading Middle Schools + 2SD Middle Schools + 1SD Middle Schools mean Middle Schools 1SD Middle Schools 2SD Figure 13. Relationship between Technical Support for Hardware and FCAT Reading in Middle Schools (Gifted Included).

PAGE 136

122 Relationship between Technical Support Hardware and FCAT Reading in Elementary Schools (Gifted Included)642 644 646 648 650 652 654 656 658 660 662 2003200420052006 YearFCAT Reading Elementary + 2SD Elementary + 1SD Elementary mean Elementary 1SD Elementary 2SD Figure 14. Relationship between Technical Support for Hardware and FCAT Reading in Elementary Schools (Gifted Included). Hypothesis 2 The second analysis conducted to answer the fi rst research question used the FCAT (NRT) Math outcome data to test the following hypothesis: H2: After controlling for school level (elementary, middle, and high), school socio-economic status, minority, limited English proficiency, students with disabilities, gifted, teacher qualification, and learning environment quality, mean school math achievement (FCAT NRT scaled scores for math) will have a positive relationship with indicators of technology integration. The first step was to build the unconditional model. The unconditional model predicted the schools’ FCAT Math from the average of FCAT Math for all schools. There were no other predictors. The average FCAT for all schools was 661.39 points ( t (2300) = 1019.7, p <.0001). Model 1: Unconditional Model Level 1: FCAT Math = 0 + r Level 2: 0 = 00 + u0 Mixed-Effects Model: FCAT Math = 00 + u0 + r

PAGE 137

123 The intraclass correlation coeffici ent (ICC) was computed to determine the proportion of variance in the FCAT Math variable that is accounted for by the schools. The ICC was .95, which is high and supports using multi-level modeling for the analysis. The model fit statistics from this model were used as the baseline for model comparisons (see Table 28). Next, time was added as a predictor to the equation to make the unconditional growth model (see Model 2a). The variance components from this analys is showed how much of the variance in the model was accounted for by time. Time was not significant in this equation ( z = 1.55, p = 0.0611), which indicated that there was very little variance in the slopes between schools. Therefore, time was set as a fixed effect, and the model with time as a fixed effect was estimated. Model 2a: Unconditional Growth Model Level 1: FCAT Math = 0 + 1*Time + r Level 2: 0 = 00 + u0 1 = 10 + u1 Mixed-Effects Model: FCAT Math = 00 + 10*Time + u0 + u1*Time + r Both the intercept ( t (2300) = 1005.83, p <.0001) and time ( t (6902) = 76.69, p <.0001) were significant parameters. Although variance between sc hools increased slightly (1%), time accounted for 46% of the variance within schools (see Model 2b). Model 2b: Unconditional Growth Model with Time Fixed Level 1: FCAT Math = 0 + 1*Time + r Level 2: 0 = 00 + u0 1 = 10 Mixed-Effects Model: FCAT Math = 00 + 10*Time + u0 + r To determine if the equation was not linear but curvilinear, time2 was added to the equation so the variance could be compared. Results indicated that time2 was significant ( t (6901) = 32.47, p <.0001) and increased the variance explained by an additional 7% (see Model 2c). When time3 was added to the equation with time2, time3 also was significant ( t (6900) = -43.26, p <.0001), and all model fit indices improved. Although adding time3 increased the amount of variance between schools, it increased the variance explained by an additional 10%. Consequently, both time2 and time3 were retained in the polynomial growth model equation (see Model 2d).

PAGE 138

124 Model 2c: Quadratic Growth Model Level 1: FCAT Math = 0 + 1*Time + 2* Time2 + r Level 2: 0 = 00 + u0 1 = 10 2 = 20 Mixed-Effects Model: FCAT Math = 00 + 10*Time + 20* Time2 + u0 + r Model 2d: Polynomial Growth Model Level 1: FCAT Math = 0 + 1*Time + 2* Time2 + 3* Time3 + r Level 2: 0 = 00 + u0 1 = 10 2 = 20 2 = 30 Mixed-Effects Model: FCAT Math = 00 + 10*Time + 20* Time2 + 30* Time3 + u0 + r Next, school level was added to the Polynomial Growth Model to predict Math (See Model 3). The significance of the parameter estimates determined if school level was significantly related to the FCAT Math and if there was an interaction with ti me. This model adjusted the mean school FCAT Math and the slope of FCAT Math growth for school leve l. The parameter estimates of school level, time, time2, and time3 were all significant. The interactions between time and both the school levels, time2 and both the school levels, and time3 and both the school levels relative to middle school were also significant. All model fit indices indicated improved fit with this m odel (Table 19). This model accounted for 76% of the between school variance and an additional 6% of the within school variance from the Polynomial Growth Model. Model 3: school level as Predictor Level 1: FCAT Math = 0 + 1*Time + 2*Time2 + 3* Time3 + r Level 2: 0 = 00 + 01*School Level + u0 1 = 10 + 11*School Level 2 = 20 + 21*School Level 3 = 30 + 31*School Level Mixed-Effects Model: FCAT Math = 00 + 01*School Level + 10*Time + 11*School Level*Time + 20*Time2 + 21*School Level*Time2 + 30*Time3 + 31*School Level*Time3 + u0 + r

PAGE 139

125 Table 19. Model 3: Time, Time2, Time3, and School Level as Predictors of Math Effect School Level Estimate SE df t p Intercept 679.96 0.7409 2298 917.71 <.0001 ** Time -5.525 0.6864 6894 -8.05 <.0001 ** Time2 8.7969 0.6067 6894 14.5 <.0001 ** Time3 -2.0103 0.1333 6894 -15.08 <.0001 ** School Level Elementary -39.7339 0.8428 6894 -47.15 <.0001 ** School Level High 30.6202 1.1196 6894 27.35 <.0001 ** School Level Middle 0 . . ** Time*School Level Elementary -11.0758 0.7808 6894 -14.19 <.0001 ** Time*School Level High -8.2523 1.0372 6894 -7.96 <.0001 ** Time*School Level Middle 0 . . ** Time2*School Level Elementary 6.2525 0.6901 6894 9.06 <.0001 ** Time2*School Level High 6.6827 0.9167 6894 7.29 <.0001 ** Time2*School Level Middle 0 . . ** Time3*School Level Elementary -0.7945 0.1517 6894 -5.24 <.0001 ** Time3*School Level High -1.4124 0.2015 6894 -7.01 <.0001 ** Time3*School Level Middle 0 . . Covariance Parameter Estimate SE z p 229.53 6.8721 33.4 <.0001 ** Residual 14.2093 0.2419 58.75 <.0001 ** Note: p < .05; ** p < .01 The next model added student demographic variables to the School Level Model. This model was estimated twice. The first time, the model was run w ith high school as a school level and all of the demographic variables except gifted (see Model 4a). The second time, the data were filtered to exclude high school as a school level and keep the gifted variable with middle and elementary schools (see Model 4b). The model fit statistics of the demographic model with all three school levels was compared with the School Level as Predictor Model to determine if there was a better fit (see Table 29). The significance of the parameter estimates determined which of the demographic variables remained in the predictor equation (see Table 20). The variance estimates showed the amount of the total variance that was accounted for by each model. When all of the demographics variables ex cept gifted were added to the model, the intercept was significant and the average middle school st arted with FCAT Math score of 678.04 ( t (2249) 1551.3, p <.0001). The parameter estimates for school level, time, time2, time3, free or reduced lunch status, minority, and students with disabilities were significant. Only Limited English Proficiency (LEP) was not significant

PAGE 140

126 at the intercept, and its interactions with time, time2, or time3 were also not significant; therefore, it was dropped from the equation for the FCAT Math outcome models for all school levels without gifted. After LEP was dropped from the equation, all intercept pa rameters were significant (see Model 4a Part 2). Interactions with time were significant for school level and free or reduced lunch status. Interactions with minority and students with disabilities were not significant. Interactions with time2 and time3 were significant for school level and minority. Interactions with time3 also were significant for free or reduced lunch status. All model fit indices indicated better fit with the addition of these demographics variables (see Table 29). Adding the demographics without LEP variables with school level explained 93% of the between school variance and 70% of the within school va riance for a total of 92% of all variance explained. Model 4a: Demographics by School Level (including High School and no Gifted) Level 1: FCAT Math = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* LEP + 35* SWD Mixed-Effects Model: FCAT Math = 00 + 01*School Level + 02*SES + 03*Minority + 04*SWD + 05*LEP + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14*SWD*Time + 15*LEP*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24*SWD*Time2 + 25*LEP*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34*SWD*Time3 + 35*LEP*Time3 + u0 + r Model 4a part 2: Demographics by School Level (including High School and no Gifted) Level 1: FCAT Math = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* SWD + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14*SWD 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* SWD 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34*SWD Mixed-Effects Model: FCAT Math = 00 + 01*School Level + 02*SES + 03*Minority + 04*SWD + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14*SWD*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24*SWD*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34*SWD*Time3 + u0 + r Table 20. Model 4a: Math predicted by Time, School L evel, and Demographics Va riables without Gifted Effect School Level Estimate SE df t p Intercept 678.04 0.4371 2249 1551.3 <.0001 ** Time -4.875 0.6887 6485 -7.08 <.0001 ** Time2 8.7255 0.6062 6485 14.39 <.0001 **

PAGE 141

127 Effect School Level Estimate SE df t p Time3 -2.0186 0.133 6485 -15.17 <.0001 ** School Level Elementary -37.1061 0.5004 6485 -74.15 <.0001 ** School Level High 26.4232 0.6592 6485 40.08 <.0001 ** School Level Middle 0 . . Free Reduced Lunch -4.3681 0.2024 6485 -21.58 <.0001 ** Minority -6.6521 0.2491 6485 -26.7 <.0001 ** LEP -0.09655 0.2155 6485 -0.45 0.6541 Students with Disabilities -2.3523 0.1496 6485 -15.72 <.0001 ** Time*School Level Elementary -12.3458 0.7963 6485 -15.5 <.0001 ** Time*School Level High -9.1177 1.0446 6485 -8.73 <.0001 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch -1.3473 0.4681 6485 -2.88 0.004 ** Time*Minority -0.3481 0.4813 6485 -0.72 0.4695 Time*LEP 0.736 0.3996 6485 1.84 0.0655 Time*Students with Disabilities -0.1192 0.3384 6485 -0.35 0.7246 Time2*School Level Elementary 7.6205 0.7019 6485 10.86 <.0001 ** Time2*School Level High 6.9833 0.9198 6485 7.59 <.0001 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.5084 0.4124 6485 -1.23 0.2177 Time2*Minority 1.1352 0.423 6485 2.68 0.0073 ** Time2*LEP -0.4588 0.3486 6485 -1.32 0.1881 Time2*Students with Disabilities 0.1103 0.2963 6485 0.37 0.7097 Time3*School Level Elementary -1.0988 0.1541 6485 -7.13 <.0001 ** Time3*School Level High -1.4404 0.2018 6485 -7.14 <.0001 ** Time3*School Level Middle 0 . . Time3*Free Reduced Lunch 0.2132 0.09047 6485 2.36 0.0185 ** Time3*Minority -0.2906 0.0928 6485 -3.13 0.0017 ** Time3*LEP 0.08966 0.07613 6485 1.18 0.2389 Time3*Students with Disabilities -0.00899 0.06485 6485 -0.14 0.8898 Covariance Parameter Estimate SE z p 66.5352 2.4051 27.66 <.0001 ** Residual 13.2518 0.2446 54.17 <.0001 ** Note: p < .05; ** p < .01 For the elementary and middle schools with gifted students, the unconditional model was estimated to provide a baseline with which to compare the demographics model. The growth model was estimated to determine if time was fixed or random. Time in this dataset was also fixed because the variance of the slope was not significant ( z = 0.71, p = 0.2397). The results from the analysis in Model 4b indicated that the inter cept, school level, time, time2, time3, free or reduced lunch status, minority, students

PAGE 142

128 with disabilities, and gifted were all significant (see Table 21). Although the intercept for LEP was not significant, the interaction of LEP and time was significant. LEP was kept in the models estimated with the elementary and middle schools with gifted students. Interactions between time and elementary school level, free or reduced lunch status, minority, LEP, and gi fted were significant. Interactions between time2 and minority, LEP, and gifted were signif icant. Interactions between time3 and minority and gifted were significant. All model fit statistics indicated better m odel fit (see Table 29). Wh en examining the variance of FCAT Math in elementary and middle schools, adding demographics variables to the equation explained 91% of the between school variance and 65% more of the within school variance. Two sets of analyses were conducted on the rest of the models in order to examine the relationship of gifted with technology integration as one of the predictors of school achievement. Model 4b: Demographics by School Level (Elementary and Middle School only) Level 1: FCAT Math = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16*Gifted 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26*Gifted 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* LEP + 35* SWD + 36*Gifted Mixed-Effects Model: FCAT Math = 00 + 01*School Level + 02*SES + 03*Minority + 04*SWD + 05*LEP + 06*Gifted + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP*Time + 15* SWD*Time + 16*Gifted*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Gifted*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34* LEP*Time3 + 35* SWD*Time3 + 36*Gifted*Time3 + u0 + r Table 21. Model 4b: Math predicted by Time, School Level, an d Demographics Variables for Elementary and Middle Schools with Gifted Effect School Level Estimate SE df t p Intercept 677 0.3887 1819 1741.5 <.0001 ** Time -4.9775 0.7084 4941 -7.03 <.0001 ** Time2 8.5863 0.6249 4941 13.74 <.0001 ** Time3 -1.9783 0.1372 4941 -14.42 <.0001 ** School Level Elementary -35.4603 0.4495 4941 -78.88 <.0001 ** School Level Middle 0 . . Free Reduced Lunch -4.2951 0.2232 4941 -19.24 <.0001 ** Minority -6.3278 0.2621 4941 -24.14 <.0001 ** LEP 0.2415 0.2118 4941 1.14 0.2542

PAGE 143

129 Effect School Level Estimate SE df t p Students with Disabilities -1.4269 0.1572 4941 -9.08 <.0001 ** Gifted 3.827 0.1833 4941 20.88 <.0001 ** Time*School Level Elementary -11.2979 0.8287 4941 -13.63 <.0001 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch -1.1402 0.5609 4941 -2.03 0.0421 Time*Minority -1.1302 0.5721 4941 -1.98 0.0483 Time*LEP 1.0742 0.4422 4941 2.43 0.0152 Time*Students with Disabilities -0.4457 0.3808 4941 -1.17 0.2419 Time*Gifted 1.994 0.3971 4941 5.02 <.0001 ** Time2*School Level Elementary 7.0796 0.7313 4941 9.68 <.0001 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.3689 0.494 4941 -0.75 0.4552 Time2*Minority 1.4852 0.5018 4941 2.96 0.0031 ** Time2*LEP -0.7827 0.385 4941 -2.03 0.0421 Time2*Students with Disabilities 0.3768 0.3324 4941 1.13 0.2571 Time2*Gifted -1.227 0.3468 4941 -3.54 0.0004 ** Time3*School Level Elementary -1.0174 0.1606 4941 -6.33 <.0001 ** Time3*School Level Middle 0 . . Time3*Free Reduced Lunch 0.1603 0.1083 4941 1.48 0.1389 Time3*Minority -0.3465 0.1099 4941 -3.15 0.0016 ** Time3*LEP 0.1601 0.08381 4941 1.91 0.0562 Time3*Students with Disabilities -0.06905 0.07251 4941 -0.95 0.341 Time3*Gifted 0.2084 0.07572 4941 2.75 0.0059 ** Covariance Parameter Estimate SE z p 47.526 1.8734 25.37 <.0001 ** Residual 13.0255 0.2708 48.1 <.0001 ** Note: p < .05; ** p < .01 The next model added the variable that measures the School Learning Environment factors to the Demographics Model by School Level Model. These in cluded teacher qualifica tions and positive learning environment. This model was estimated twice, first without the gifted population but all school levels (see model 5a) and then with elementary and middle school levels and the gifted population (see model 5b). When school learning environment factors were added with the demographic and school level variables for all school levels, the parameter estimates for the intercept, time, time2, time3, elementary and high school relative to middle school, free or reduced lunch st atus, minority, students w ith disabilities, teacher qualifications, and positive learning environment were significant (see Table 22). Significant interactions with time included elementary and high school relative to middle school, free or reduced lunch status, and

PAGE 144

130 positive learning environment. Significant interactions with time2 and time3 were elementary and high school, minority, and positive learning environment. Adding the student learning environment variables explained an additional 1% of the between school va riance and explained 1% less of the within school variance for a total of 93% of all of the variance explai ned. All of the model fit indices indicated that this model fit the data better (see Table 28). Model 5a: School Learning Environment with Demographics by School Level (All School Levels without Gifted) Level 1: FCAT Math = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* SWD + 05* Teacher Qualifications + 06*Positive Learning Environment + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14*SWD + 15* Teacher Qualifications + 16*Positive Learning Environment 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* SWD + 25* Teacher Qualifications + 26*Positive Learning Environment 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* SWD + 35* Teacher Qualifications + 36*Positive Learning Environment Mixed-Effects Model: FCAT Math = 00 + 01*School Level + 02*SES + 03*Minority + 04* SWD + 05*Teacher Qualifications + 06* Positive Learning Environment + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14*SWD *Time + 15* Teacher Qualifications*Time + 16* Positive Learning Environment*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* SWD*Time2 + 25* Teacher Qualifications*Time2 + 26* Positive Learning Environment*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34*SWD*Time3 + 35* Teacher Qualifications*Time3 + 36* Positive Learning Environment*Time3 + u0 + r Table 22. Model 5a: Math Predicted by Demographics and Stud ent Learning Environment by School Level (All School Levels without Gifted and LEP) Effect School Level Estimate SE df t p Intercept 680.14 0.4403 2298 1544.6 <.0001 ** Time -5.504 0.8634 6867 -6.37 <.0001 ** Time2 9.2097 0.7489 6867 12.3 <.0001 ** Time3 -2.1354 0.1624 6867 -13.15 <.0001 ** School Level Elementary -39.8795 0.5171 6867 -77.12 <.0001 ** School Level High 26.3373 0.6155 6867 42.79 <.0001 ** School Level Middle 0 . . Free Reduced Lunch -4.1447 0.1922 6867 -21.56 <.0001 ** Minority -6.1259 0.2129 6867 -28.77 <.0001 ** Students with Disabilities -2.1066 0.1437 6867 -14.66 <.0001 ** Positive Learning Environment 1.6474 0.1513 6867 10.89 <.0001 ** Positive Teacher Qualifications 1.003 0.1138 6867 8.81 <.0001 **

PAGE 145

131 Effect School Level Estimate SE df t p Time*School Level Elementary -11.0835 1.064 6867 -10.42 <.0001 ** Time*School Level High -9.3217 1.0637 6867 -8.76 <.0001 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch -1.411 0.489 6867 -2.89 0.0039 ** Time*Minority -0.4513 0.442 6867 -1.02 0.3074 Time*Students with Disabilities -0.4879 0.3396 6867 -1.44 0.1509 Time*Positive Learning Environment -0.7349 0.4571 6867 -1.61 0.1079 Time*Positive Teacher Qualifications 0.8114 0.3249 6867 2.5 0.0125 Time2*School Level Elementary 6.3613 0.9335 6867 6.81 <.0001 ** Time2*School Level High 7.3264 0.9384 6867 7.81 <.0001 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.2134 0.4348 6867 -0.49 0.6236 Time2*Minority 1.0957 0.3913 6867 2.8 0.0051 ** Time2*Students with Disabilities 0.4279 0.2986 6867 1.43 0.1518 Time2*Positive Learning Environment 0.9942 0.409 6867 2.43 0.0151 Time2*Positive Teacher Qualifications -0.49 0.2848 6867 -1.72 0.0854 Time3*School Level Elementary -0.8004 0.204 6867 -3.92 <.0001 ** Time3*School Level High -1.5226 0.206 6867 -7.39 <.0001 ** Time3*School Level Middle 0 . . Time3*Free Reduced Lunch 0.13 0.09581 6867 1.36 0.1749 Time3*Minority -0.2709 0.08609 6867 -3.15 0.0017 ** Time3*Students with Disabilities -0.07848 0.06536 6867 -1.2 0.2299 Time3*Positive Learning Environment -0.242 0.09033 6867 -2.68 0.0074 ** Time3*Positive Teacher Qualifications 0.07683 0.06214 6867 1.24 0.2163 Covariance Parameter Estimate SE z p 57.0301 2.1007 27.15 <.0001 ** Residual 13.8759 0.2505 55.39 <.0001 ** Note: p < .05; ** p < .01 When the data were filtered to include only elementary and middle schools and gifted was also added to the equation, all intercept parameter estimates were significant (i.e., elementary school, time, time2, time3, free or reduced lunch status, mi nority, students with disabilitie s, teacher qualifications, and positive learning environment except for LEP). Significan t interactions with time included elementary, free or reduced lunch status, minority, LEP, gifted, and teacher qualificatio ns. Significant interactions with time2

PAGE 146

132 included elementary, minority, LEP, gifted, and positive learning environment. Significant interactions with time3 included elementary, minority, LEP, gifted, and positive learning environment (see Table 23). This model demonstrated better fit than the previous model by all model fit indices (see Table 29). It explained 1% more of the between school variance and the same amount of the within school variance as the previous mode l and explained 91% of all the variance. Model 5b: School Learning Environment with Demographics by School Level (Elementary and Middle Schools with Gifted) Level 1: FCAT Math = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16*Gifted + 17* Teacher Qualifications + 18* Positive Learning Environment 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26*Gifted + 27* Teacher Qualifications + 28* Positive Learning Environment 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* LEP + 35* SWD + 36*Gifted + 37* Teacher Qualifications + 38* Positive Learning Environment Mixed-Effects Model: FCAT Math = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Gifted*Time + 17* Teacher Qualifications*Time + 18* Positive Learning Environment*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Gifted*Time2 + 27* Teacher Qualifications*Time2 + 28* Positive Learning Environment*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34* LEP*Time3 + 35* SWD*Time3 + 36*Gifted*Time3 + 37* Teacher Qualifications*Time3 + 38* Positive Learning Environment*Time3 + u0 + r Table 23. Model 5b: Math Predicted by Demographics and Stud ent Learning Environment by School Level for Elementary and Middle School with Gifted Effect School Level Estimate SE df t p Intercept 678.69 0.4319 1819 1571.4 <.0001 ** Time -5.6609 0.9672 4933 -5.85 <.0001 ** Time2 9.2683 0.8305 4933 11.16 <.0001 ** Time3 -2.1431 0.179 4933 -11.97 <.0001 ** School Level Elementary -37.7309 0.5221 4933 -72.27 <.0001 ** School Level Middle 0 . . Free Reduced Lunch -4.3582 0.2218 4933 -19.65 <.0001 ** Minority -5.7845 0.2611 4933 -22.16 <.0001 ** LEP 0.05423 0.2072 4933 0.26 0.7935 Students with Disabilities -1.3258 0.1555 4933 -8.53 <.0001 ** Gifted 3.4228 0.1804 4933 18.97 <.0001 **

PAGE 147

133 Effect School Level Estimate SE df t p Positive Learning Environment 1.3689 0.1921 4933 7.13 <.0001 ** Positive Teacher Qualifications 0.8872 0.128 4933 6.93 <.0001 ** Time*School Level Elementary -9.8653 1.2258 4933 -8.05 <.0001 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch -1.2316 0.6035 4933 -2.04 0.0413 Time*Minority -1.3844 0.5835 4933 -2.37 0.0177 Time*LEP 1.2035 0.4567 4933 2.64 0.0084 ** Time*Students with Disabilities -0.441 0.3872 4933 -1.14 0.2549 Time*Gifted 1.9348 0.4058 4933 4.77 <.0001 ** Time*Positive Learning Environment -1.1384 0.6059 4933 -1.88 0.0603 Time*Positive Teacher Qualifications 0.8163 0.365 4933 2.24 0.0254 Time2*School Level Elementary 5.5149 1.0645 4933 5.18 <.0001 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch 0.00502 0.5357 4933 0.01 0.9925 Time2*Minority 1.6451 0.5102 4933 3.22 0.0013 ** Time2*LEP -1.0317 0.3986 4933 -2.59 0.0097 ** Time2*Students with Disabilities 0.3956 0.3379 4933 1.17 0.2418 Time2*Gifted -1.2193 0.3542 4933 -3.44 0.0006 ** Time2*Positive Learning Environment 1.3845 0.5353 4933 2.59 0.0097 ** Time2*Positive Teacher Qualifications -0.5752 0.3204 4933 -1.8 0.0726 Time3*School Level Elementary -0.6465 0.2312 4933 -2.8 0.0052 ** Time3*School Level Middle 0 . . Time3*Free Reduced Lunch 0.05508 0.1177 4933 0.47 0.6398 Time3*Minority -0.3765 0.1116 4933 -3.37 0.0007 ** Time3*LEP 0.2242 0.08686 4933 2.58 0.0099 ** Time3*Students with Disabilities -0.07603 0.07366 4933 -1.03 0.302 Time3*Gifted 0.2125 0.07729 4933 2.75 0.006 ** Time3*Positive Learning Environment -0.3302 0.1172 4933 -2.82 0.0049 ** Time3*Positive Teacher Qualifications 0.1034 0.07005 4933 1.48 0.1401 Covariance Parameter Estimate SE z p 42.01 1.6722 25.12 <.0001 ** Residual 12.9766 0.2698 48.1 <.0001 ** Note: p < .05; ** p < .01

PAGE 148

134 The next model added technology integration variables with the demographics, learning environment, and school level variables. These included student access to various types of software, teachers regularly using various types of software, frequency that students use various types of software, and technology support. This model was estimated twi ce, first without gifted population but all school levels (see model 6a) and then with elementary and middle school levels and gifted population (see model 6b). When the model was estimated with all school levels without gifted and LEP, the only significant technology parameter estimates were the percent of teachers who use technology for administrative purposes and the interaction of time, time2, and time3 with teachers’ use of technology for administrative purposes (see Table 24). Other significant parameter es timates included the interactions with time and high school and elementary school relative to middle sc hool, free or reduced lunch status, positive learning environment, and positive teacher qualif ications. Significant interactions with time2 and time3 included elementary and high school relative to middle school, minority, and positive learning environment. Only one model fit index indicated that this model had better fit (see Table 28). No additional variance was explained with this model. One technology integration indicator was retained in the final model for all school levels without gifted, the percent of teacher s who use technology for administrative purposes. Model 6a: Technology Integration with Demographics and Student Learning Environment by School Level (All School Levels without Gifted) Level 1: FCAT Math = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* SWD + 05* Teacher Qualifications + 06* Positive Learning Environment + 07*Access Content SW + 08*Access Office SW + 09*Access Ad Prod SW + 010*Teachers Use Deliver Instruction + 011*Teachers use Admin + 012*Frequency Students Use Content + 013*Frequency Students Use Tool + 014*Technical Support Human + 015*Technical Support Hardware + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* SWD + 15* Teacher Qualifications + 16* Positive Learning Environment + 17*Access Content SW + 18*Access Office SW + 19*Access Ad Prod SW + 110*Teachers Use Deliver Instruction + 111*Teachers use Admin + 112*Frequency Students Use Content + 113*Frequency Students Use Tool + 114*Technical Support Human + 115*Technical Support Hardware 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24SWD + 25* Teacher Qualifications + 26* Positive Learning Environment + 27*Access Content SW + 28*Access Office SW + 29*Access Ad Prod SW + 210*Teachers Use Deliver Instruction + 211*Teachers use Admin + 212*Frequency Students Use Content + 213*Frequency Students Use Tool + 214*Technical Support Human + 215*Technical Support Hardware 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* SWD + 35* Teacher Qualifications + 36* Positive Learning Environment + 37*Access Content SW + 38*Access Office SW + 319*Access Ad Prod SW + 310*Teachers Use Deliver Instruction + 311*Teachers use Admin + 314*Frequency Students Use Content + 315*Frequency Students Use Tool + 316*Technical Support Human + 317*Technical Support Hardware

PAGE 149

135 Mixed-Effects Model: FCAT Math = 00 + 01*School Level + 02*SES + 03*Minority + 04* SWD + 05* Teacher Qualifications + 06* Positive Learning Environment + 07*Access Content SW + 08*Access Office SW + 09*Access Ad Prod SW + 010*Teachers Use Deliver Instruction + 011*Teachers use Admin + 012*Frequency Students Use Content + 013*Frequency Students Use Tool + 014*Technical Support Human + 015*Technical Support Hardware + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* SWD *Time + 15*Teacher Qualifications*Time + 16* Positive Learning Environment*Time + 17*Access Content SW*Time + 18*Access Office SW*Time + 19*Access Ad Prod SW*Time + 110*Teachers Use Deliver Instruction*Time + 111*Teachers use Admin*Time + 112*Frequency Students Use Content*Time + 113*Frequency Students Use Tool*Time + 114*Technical Support Human*Time + 115*Technical Support Hardware*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* SWD*Time2 + 25*Teacher Qualifications*Time2 + 26* Positive Learning Environment*Time2 + 27*Access Content SW*Time2 + 28*Access Office SW*Time2 + 29*Access Ad Prod SW*Time2 + 210*Teachers Use Deliver Instruction*Time2 + 211*Teachers use Admin*Time2 + 212*Frequency Students Use Content*Time2 + 213*Frequency Students Use Tool*Time2 + 214*Technical Support Human*Time2 + 215*Technical Support Hardware*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34*SWD*Time3 + 35*Teacher Qualifications*Time3 + 36* Positive Learning Environment*Time3 + 37*Access Content SW*Time3 + 38*Access Office SW*Time3 + 39*Access Ad Prod SW*Time3 + 310*Teachers Use Deliver Instruction*Time3 + 311*Teachers use Admin*Time3 + 312*Frequency Students Use Content*Time3 + 313*Frequency Students Use Tool*Time3 + 314*Technical Support Human*Time3 + 315*Technical Support Hardware*Time3 + u0 + r Table 24. Model 6a: Math Predicted by Technology Integration with Demographics and Student Learning Environment by School Level (All Sc hool Levels without Gifted and LEP) Effect School Level Estimate SE df t p Intercept 680.17 0.4421 2298 1538.4 <.0001 ** Time -5.9419 0.9043 6831 -6.57 <.0001 ** Time2 9.5441 0.7855 6831 12.15 <.0001 ** Time3 -2.1982 0.1703 6831 -12.91 <.0001 ** School Level Elementary -39.9183 0.5215 6831 -76.55 <.0001 ** School Level High 26.261 0.6135 6831 42.8 <.0001 ** School Level Middle 0 . . Free Reduced Lunch -4.3101 0.1957 6831 -22.03 <.0001 ** Minority -6.0545 0.213 6831 -28.43 <.0001 ** Students with Disabilities -2.0685 0.1436 6831 -14.41 <.0001 ** Positive Learning Environment 1.6887 0.1517 6831 11.13 <.0001 ** Positive Teacher Qualifications 0.9759 0.1139 6831 8.57 <.0001 ** Access Content Software -0.01108 0.1061 6831 -0.1 0.9168 Access Office Software -0.00873 0.105 6831 -0.08 0.9337 Access Advanced Production Software -0.0507 0.1081 6831 -0.47 0.6391 Teachers Use To Deliver Instruction 0.1053 0.1169 6831 0.9 0.3674 Teachers Use For -0.2504 0. 1209 6831 -2.07 0.0383

PAGE 150

136 Effect School Level Estimate SE df t p Administrative Purposes Frequency that students use content software -0.07561 0.09745 6831 -0.78 0.4378 Frequency Students Use Tool-Based Software 0.03286 0.1045 6831 0.31 0.7532 Technical Support Human 0.04804 0.09561 6831 0.5 0.6154 Technical Support Hardware -0.04762 0.09287 6831 -0.51 0.6081 Time*School Level Elementary -10.4764 1.1296 6831 -9.27 <.0001 ** Time*School Level High -9.0311 1.0785 6831 -8.37 <.0001 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch -1.3148 0.5 6831 -2.63 0.0086 ** Time*Minority -0.2605 0.4475 6831 -0.58 0.5605 Time*Students with Disabilities -0.4827 0.3404 6831 -1.42 0.1562 Time*Positive Learning Environment -0.932 0.461 6831 -2.02 0.0432 Time*Positive Teacher Qualifications 0.7667 0.3288 6831 2.33 0.0197 Time*Access Content Software 0.1684 0.3766 6831 0.45 0.6547 Time*Access Office Software -0.3604 0.3705 6831 -0.97 0.3307 Time*Access Advanced Production Software 0.6149 0.3764 6831 1.63 0.1023 Time*Teachers Use To Deliver Instruction -0.7755 0.4154 6831 -1.87 0.062 Time*Teachers Use For Administrative Purposes 1.5909 0.4216 6831 3.77 0.0002 ** Time*Frequency that Students Use Content Software -0.1084 0.3626 6831 -0.3 0.7651 Time*Frequency Students Use Tool-Based Software -0.05096 0.3817 6831 -0.13 0.8938 Time*Technical Support Human 0.2935 0.3244 6831 0.9 0.3658 Time*Technical Support Hardware -0.1282 0.3315 6831 -0.39 0.6989 Time2*School Level Elementary 5.9175 0.9951 6831 5.95 <.0001 ** Time2*School Level High 7.0793 0.9522 6831 7.43 <.0001 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.2752 0.4461 6831 -0.62 0.5374 Time2*Minority 0.9581 0.3961 6831 2.42 0.0156 Time2*Students with Disabilities 0.4094 0.2994 6831 1.37 0.1715 Time2*Positive Learning Environment 1.1369 0.4123 6831 2.76 0.0058 ** Time2*Positive Teacher Qualifications -0.4446 0.2898 6831 -1.53 0.1251 Time2*Access Content Software -0.1988 0.3325 6831 -0.6 0.5499 Time2*Access Office 0.272 0.325 6831 0.84 0.4026

PAGE 151

137 Effect School Level Estimate SE df t p Software Time2*Access Advanced Production Software -0.5203 0.331 6831 -1.57 0.116 Time2*Teach Use Deliver Instruction 0.5984 0.3675 6831 1.63 0.1036 Time2*Teach Use Administrative Purposes -1.0999 0.3689 6831 -2.98 0.0029 ** Time2*Frequency Student Use Content Software 0.1849 0.3211 6831 0.58 0.5647 Time2*Frequency Students Use Tool-Based Software 0.03957 0.343 6831 0.12 0.9082 Time2*Technical Support Human -0.243 0.2868 6831 -0.85 0.397 Time2*Technical Support Hardware 0.0616 0.2916 6831 0.21 0.8327 Time3*School Level Elementary -0.7216 0.2177 6831 -3.32 0.0009 ** Time3*School Level High -1.4644 0.2091 6831 -7 <.0001 ** Time3*School Level Middle 0 . . Time3*Free Reduced Lunch 0.1424 0.09852 6831 1.45 0.1484 Time3*Minority -0.245 0.0871 6831 -2.81 0.0049 ** Time3*Students with Disabilities -0.07226 0.06554 6831 -1.1 0.2703 Time3*Positive Learning Environment -0.2697 0.091 6831 -2.96 0.003 ** Time3*Positive Teacher Qualifications 0.06669 0.06341 6831 1.05 0.293 Time3*Access Content Software 0.05345 0.07311 6831 0.73 0.4648 Time3*Access Office Software -0.04888 0.07111 6831 -0.69 0.4919 Time3*Access Advanced Production Software 0.1126 0.07247 6831 1.55 0.1203 Time3*Teach Use Deliver Instruction -0.1146 0.08082 6831 -1.42 0.1563 Time3*Teach Use Administrative Purposes 0.2113 0.08062 6831 2.62 0.0088 ** Time3*Frequency Student Use Content Software -0.03818 0.07053 6831 -0.54 0.5884 Time3*Frequency Students Use Tool-Based Software -0.01826 0.07623 6831 -0.24 0.8107 Time3*Technical Support Human 0.04521 0.0631 6831 0.72 0.4737 Time3*Technical Support Hardware -0.00677 0.06402 6831 -0.11 0.9158 Covariance Parameter Estimate SE z p 56.2368 2.0688 27.18 <.0001 ** Residual 13.7986 0.249 55.43 <.0001 ** Note: p < .05; ** p < .01

PAGE 152

138 Similar results were found with the elementary and middle school data with gifted. Percent of teachers who regularly use technology for administra tive purposes was a signifi cant technology parameter estimate at the intercept and with its interaction with time, time2, and time3. An additional technology integration variable that had significant interactions with time, time2, and time3 was percent of teachers who regularly use technology to deliver instruction (see Table 25). Other significant parameter estimates included the intercept, time, time2, time3, elementary school, free or reduced lunch status, minority, students with disabilities, gifted, positive learning environment, and pos itive teacher qualifications. Significant interactions with time included LEP, gifted, pos itive learning environmen t, and positive teacher qualifications. Significant interactions with time2 and time3 included minority, LEP, gifted, and positive learning environment. Only the -2 Log Likelihood index indicated that this model had better fit (see Table 29). Moreover, adding the technology integration indicators to the model did not explain any additional variance. Two technology integratio n variables, teachers use technology for administrative purposes and teachers use technology to deliver in struction, were the only technology integration indicators retained in the final model for the data with elementary and middle schools and gifted in order to determine if the model fit improved without the noise from the technology integration variables that were not significant. Model 6b: Technology Integration with Demographics and Student Learning Environment by School Level (Elementary and Middle Schools with Gifted) Level 1: FCAT Math = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 09*Access Content SW + 010*Access Office SW + 011*Access Ad Prod SW + 012*Teachers Use Deliver Instruction + 013*Teachers use Admin + 014*Frequency Students Use Content + 015*Frequency Students Use Tool + 016*Technical Support Human + 017*Technical Support Hardware + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16*Gifted + 17* Teacher Qualifications + 18* Positive Learning Environment + 19*Access Content SW + 110*Access Office SW + 111*Access Ad Prod SW + 112*Teachers Use Deliver Instruction + 113*Teachers use Admin + 114*Frequency Students Use Content + 115*Frequency Students Use Tool + 116*Technical Support Human + 117*Technical Support Hardware 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26*Gifted + 27* Teacher Qualifications + 28* Positive Learning Environment + 29*Access Content SW + 210*Access Office SW + 211*Access Ad Prod SW + 212*Teachers Use Deliver Instruction + 213*Teachers use Admin + 214*Frequency Students Use Content + 215*Frequency Students Use Tool + 216*Technical Support Human + 217*Technical Support Hardware 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* LEP + 35* SWD + 36*Gifted + 37* Teacher Qualifications + 38* Positive Learning Environment + 39*Access Content SW + 310*Access Office SW + 311*Access Ad Prod SW + 312*Teachers Use Deliver Instruction + 313*Teachers use Admin + 314*Frequency Students Use Content + 315*Frequency Students Use Tool + 316*Technical Support Human + 317*Technical Support Hardware

PAGE 153

139 Mixed-Effects Model: FCAT Math = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 09*Access Content SW + 010*Access Office SW + 011*Access Ad Prod SW + 012*Teachers Use Deliver Instruction + 013*Teachers use Admin + 014*Frequency Students Use Content + 015*Frequency Students Use Tool + 016*Technical Support Human + 017*Technical Support Hardware + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Gifted*Time + 17* Teacher Qualifications*Time + 18* Positive Learning Environment*Time + 19*Access Content SW*Time + 110*Access Office SW*Time + 111*Access Ad Prod SW*Time + 112*Teachers Use Deliver Instruction*Time + 113*Teachers use Admin*Time + 114*Frequency Students Use Content*Time + 115*Frequency Students Use Tool*Time + 116*Technical Support Human*Time + 117*Technical Support Hardware*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Gifted*Time2 + 27* Teacher Qualifications*Time2 + 28* Positive Learning Environment*Time2 + 29*Access Content SW*Time2 + 210*Access Office SW*Time2 + 211*Access Ad Prod SW*Time2 + 212*Teachers Use Deliver Instruction*Time2 + 213*Teachers use Admin*Time2 + 214*Frequency Students Use Content*Time2 + 215*Frequency Students Use Tool*Time2 + 216*Technical Support Human*Time2 + 217*Technical Support Hardware*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34* LEP*Time3 + 35* SWD*Time3 + 36*Gifted*Time3 + 37* Teacher Qualifications*Time3 + 38* Positive Learning Environment*Time3 + 39*Access Content SW*Time3 + 310*Access Office SW*Time3 + 311*Access Ad Prod SW*Time3 + 312*Teachers Use Deliver Instruction*Time3 + 313*Teachers use Admin*Time3 + 314*Frequency Students Use Content*Time3 + 315*Frequency Students Use Tool*Time3 + 316*Technical Support Human*Time3 + 317*Technical Support Hardware*Time3 + u0 + r Table 25. Model 6b: Math Predicted by Technology Integration with Demographics and Student Learning Environment by School Level for Elemen tary and Middle Schools with Gifted Effect School Level Estimate SE df t p Intercept 678.79 0.4361 1819 1556.4 <.0001 ** Time -6.3563 1.0164 4897 -6.25 <.0001 ** Time2 9.7723 0.8738 4897 11.18 <.0001 ** Time3 -2.2345 0.1882 4897 -11.87 <.0001 ** School Level Elementary -37.8503 0.5293 4897 -71.51 <.0001 ** School Level Middle 0 . . Free Reduced Lunch -4.4929 0.2256 4897 -19.91 <.0001 ** Minority -5.7611 0.2608 4897 -22.09 <.0001 ** LEP 0.04658 0.2069 4897 0.23 0.8219 Students with Disabilities -1.2816 0.1552 4897 -8.26 <.0001 ** Gifted 3.378 0.1807 4897 18.69 <.0001 ** Positive Learning Environment 1.4189 0.1925 4897 7.37 <.0001 ** Positive Teacher Qualifications 0.8593 0.1278 4897 6.72 <.0001 ** Access Content Software 0.07175 0.1222 4897 0.59 0.557 Access Office Software 0.02718 0.117 4897 0.23 0.8164 Access Advanced Production Software -0.2399 0.1227 4897 -1.96 0.0505 Teachers Use To Deliver 0. 14 0.1348 4897 1.04 0.299

PAGE 154

140 Effect School Level Estimate SE df t p Instruction Teachers Use For Administrative Purposes -0.3839 0.1419 4897 -2.71 0.0068 ** Frequency that Students Use Content Software 0.0308 0.1126 4897 0.27 0.7846 Frequency Students Use Tool-Based Software 0.1761 0.1182 4897 1.49 0.1364 Technical Support Human 0.0115 0.1114 4897 0.1 0.9178 Technical Support Hardware -0.1523 0.1045 4897 -1.46 0.145 Time*School Level Elementary -8.9621 1.2995 4897 -6.9 <.0001 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch -1.0653 0.616 4897 -1.73 0.0838 Time*Minority -1.0811 0.5887 4897 -1.84 0.0664 Time*LEP 1.1872 0.4576 4897 2.59 0.0095 ** Time*Students with Disabilities -0.4095 0.3877 4897 -1.06 0.291 Time*Gifted 2.0867 0.4101 4897 5.09 <.0001 ** Time*Positive Learning Environment -1.3589 0.6096 4897 -2.23 0.0258 Time*Positive Teacher Qualifications 0.7588 0.3693 4897 2.05 0.04 Time*Access Content Software 0.2175 0.4258 4897 0.51 0.6094 Time*Access Office Software -0.2559 0.4123 4897 -0.62 0.5349 Time*Access Advanced Production Software 0.8018 0.417 4897 1.92 0.0546 Time*Teachers Use To Deliver Instruction -1.1336 0.4761 4897 -2.38 0.0173 Time*Teachers Use For Administrative Purposes 2.1245 0.491 4897 4.33 <.0001 ** Time*Frequency that Students Use Content Software -0.4436 0.4246 4897 -1.04 0.2962 Time*Frequency Students Use Tool-Based Software -0.2507 0.4339 4897 -0.58 0.5634 Time*Technical Support Human 0.3625 0.3768 4897 0.96 0.336 Time*Technical Support Hardware 0.3317 0.3718 4897 0.89 0.3724 Time2*School Level Elementary 4.8721 1.1329 4897 4.3 <.0001 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.1321 0.5476 4897 -0.24 0.8094 Time2*Minority 1.4278 0.5149 4897 2.77 0.0056 ** Time2*LEP -1.0087 0.3998 4897 -2.52 0.0117 Time2*Students with Disabilities 0.3554 0.3383 4897 1.05 0.2936 Time2*Gifted -1.3246 0.3574 4897 -3.71 0.0002 ** Time2*Positive Learning Environment 1.5246 0.5383 4897 2.83 0.0046 ** Time2*Positive Teacher -0.5229 0. 326 4897 -1.6 0.1088

PAGE 155

141 Effect School Level Estimate SE df t p Qualifications Time2*Access Content Software -0.2945 0.3742 4897 -0.79 0.4313 Time2*Access Office Software 0.1636 0.3603 4897 0.45 0.6499 Time2*Access Advanced Production Software -0.5955 0.366 4897 -1.63 0.1038 Time2*Teach Use Deliver Instruction 0.9558 0.4188 4897 2.28 0.0225 Time2*Teach Use Administrative Purposes -1.5112 0.4249 4897 -3.56 0.0004 ** Time2*Frequency Student Use Content Software 0.4223 0.3747 4897 1.13 0.2597 Time2*Frequency Students Use Tool-Based Software 0.08314 0.3877 4897 0.21 0.8302 Time2*Technical Support Human -0.2902 0.3304 4897 -0.88 0.3798 Time2*Technical Support Hardware -0.2989 0.3253 4897 -0.92 0.3582 Time3*School Level Elementary -0.5332 0.2463 4897 -2.17 0.0304 Time3*School Level Middle 0 . . Time3*Free Reduced Lunch 0.08304 0.1205 4897 0.69 0.4908 Time3*Minority -0.3336 0.1126 4897 -2.96 0.0031 ** Time3*LEP 0.219 0.08715 4897 2.51 0.012 Time3*Students with Disabilities -0.06551 0.07373 4897 -0.89 0.3743 Time3*Gifted 0.2321 0.07791 4897 2.98 0.0029 ** Time3*Positive Learning Environment -0.3548 0.1178 4897 -3.01 0.0026 ** Time3*Positive Teacher Qualifications 0.09245 0.07149 4897 1.29 0.196 Time3*Access Content Software 0.07482 0.08215 4897 0.91 0.3624 Time3*Access Office Software -0.02649 0.07866 4897 -0.34 0.7363 Time3*Access Advanced Production Software 0.1212 0.08017 4897 1.51 0.1306 Time3*Teach Use Deliver Instruction -0.2005 0.09183 4897 -2.18 0.0291 Time3*Teach Use Administrative Purposes 0.2956 0.09225 4897 3.2 0.0014 Time3*Frequency Student Use Content Software -0.08079 0.08204 4897 -0.98 0.3248 Time3*Frequency Students Use Tool-Based Software -0.01248 0.08581 4897 -0.15 0.8843 Time3*Technical Support Human 0.0511 0.07236 4897 0.71 0.4801 Time3*Technical Support Hardware 0.06885 0.07113 4897 0.97 0.3331

PAGE 156

142 Covariance Parameter Estimate SE z p 41.4305 1.6504 25.1 <.0001 ** Residual 12.86 0.2674 48.09 <.0001 ** Note: p < .05; ** p < .01 The last models estimated in order to answer the second hypothesis predicting math achievement included all school levels, demographic, student learning environment, and significant technology integration variables. These models were different beca use the model fit to the data for all schools levels without gifted and LEP included one technology integr ation variable percent of teachers who regularly use technology for administrative purposes (see model 7a); while the model fitted to the data with elementary and middle school levels and gifted included two technology integration variables – percent of teachers who regularly use technology for administra tive purposes and percent of teachers who regularly use technology for delivery of instruction (see model 7b). For the model with all schools levels and no gifted or LEP, the same parameter estimates and in teractions identified in the previous models as significant were significant again (see Table 26). Although, there was no difference in the percentage of variance explained in this model than was in the Demographic Model with Student Learning Environment by school level or the Technology Integration with Demographic and Student Learning Environment Model by school level, the AIC, AICC, and BIC indices all indicated better model fit (see Table 28). The level-1 residuals for the final model for predicting FCAT Math using all school levels without gifted ranged between -23.67 and 21.57 with a standard deviation of 3.25. Although there were outliers, skewness was 0.06 and kurtosis was 2.45, which would indicate that the residuals were evenly distributed with most around the mean. Distribution of the empirical bayes intercepts ranged between -31.16 and 41.52 with standard deviation of 7.30. Skewness was 0.47, and kurtosis was 1.32, which indicated that the residuals at level-2 were within acceptable range.

PAGE 157

143 Final Model 7a: Significant Technology Integration Indicators with Demographics and Student Learning Environment by School Level (All School Levels without Gifted) Level 1: FCAT Math = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* SWD + 05*Teacher Qualifications + 06* Positive Learning Environment + 07*Teachers use Admin + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* SWD + 15*Teacher Qualifications + 16* Positive Learning Environment + 17*Teachers use Admin 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* SWD + 25*Teacher Qualifications + 26* Positive Learning Environment + 27*Teachers use Admin 3 = 30 + 31*School Level + 32*SES + 33*Minority + 04* SWD + 35*Teacher Qualifications + 36* Positive Learning Environment + 37*Teachers use Admin Mixed-Effects Model: FCAT Math = 00 + 01*School Level + 02*SES + 03*Minority + 04* SWD + 05*Teacher Qualifications + 06* Positive Learning Environment + 07*Teachers use Admin + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* SWD*Time + 15*Teacher Qualifications*Time + 16* Positive Learning Environment*Time + 17*Teachers use Admin*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* SWD*Time2 + 25*Teacher Qualifications*Time2 + 26* Positive Learning Environment*Time2 + 27*Teachers use Admin*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34* SWD*Time3 + 35*Teacher Qualifications*Time3 + 36* Positive Learning Environment*Time3 + 37*Teachers use Admin*Time3 + u0 + r Table 26. Final Model 7a: Math Predicted by Significant Technology Integration with Demographics and Student Learning Environment by School Level (All School Levels without Gifted and LEP) Effect School Level Estimate SE df t p Intercept 680.23 0.4397 2298 1546.9 <.0001 ** Time -5.997 0.873 6863 -6.87 <.0001 ** Time2 9.5432 0.7569 6863 12.61 <.0001 ** Time3 -2.1982 0.1641 6863 -13.4 <.0001 ** School Level Elementary -40.002 0.517 6863 -77.37 <.0001 ** School Level High 26.2999 0.6131 6863 42.9 <.0001 ** School Level Middle 0 . . Free Reduced Lunch -4.2872 0.1943 6863 -22.06 <.0001 ** Minority -6.0819 0.2125 6863 -28.63 <.0001 ** Students with Disabilities -2.0817 0.1435 6863 -14.51 <.0001 ** Positive Learning Environment 1.696 0.1513 6863 11.21 <.0001 ** Positive Teacher Qualifications 0.982 0.1138 6863 8.63 <.0001 ** Teachers Use For Administrative Purposes -0.2258 0.09797 6863 -2.3 0.0212 Time*School Level Elementary -10.409 1.0798 6863 -9.64 <.0001 ** Time*School Level High -9.2119 1.0628 6863 -8.67 <.0001 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch -1.2348 0.4906 6863 -2.52 0.0119 Time*Minority -0.2952 0.4438 6863 -0.67 0.5059

PAGE 158

144 Effect School Level Estimate SE df t p Time*Students with Disabilities -0.4659 0.3393 6863 -1.37 0.1698 Time*Positive Learning Environment -0.8874 0.4587 6863 -1.93 0.0531 Time*Positive Teacher Qualifications 0.8392 0.3246 6863 2.59 0.0098 ** Time*Teachers Use For Administrative Purposes 1.242 0.3389 6863 3.66 0.0002 ** Time2*School Level Elementary 5.9152 0.947 6863 6.25 <.0001 ** Time2*School Level High 7.2513 0.9377 6863 7.73 <.0001 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.3253 0.4361 6863 -0.75 0.4557 Time2*Minority 0.9858 0.3927 6863 2.51 0.0121 Time2*Students with Disabilities 0.3995 0.2983 6863 1.34 0.1806 Time2*Positive Learning Environment 1.0848 0.4104 6863 2.64 0.0082 ** Time2*Positive Teacher Qualifications -0.5057 0.2846 6863 -1.78 0.0756 Time2*Teach Use Administrative Purposes -0.8462 0.2983 6863 -2.84 0.0046 ** Time3*School Level Elementary -0.7174 0.2068 6863 -3.47 0.0005 ** Time3*School Level High -1.5074 0.2059 6863 -7.32 <.0001 ** Time3*School Level Middle 0 . . Time3*Free Reduced Lunch 0.1531 0.09614 6863 1.59 0.1113 Time3*Minority -0.2512 0.08633 6863 -2.91 0.0036 ** Time3*Students with Disabilities -0.07075 0.06531 6863 -1.08 0.2788 Time3*Positive Learning Environment -0.2574 0.09059 6863 -2.84 0.0045 ** Time3*Positive Teacher Qualifications 0.07983 0.06207 6863 1.29 0.1985 Time3*Teach Use Administrative Purposes 0.1671 0.0654 6863 2.56 0.0106 Covariance Parameter Estimate SE z p 56.4906 2.0766 27.2 <.0001 ** Residual 13.8405 0.2496 55.44 <.0001 ** Note: p < .05; ** p < .01 For the model with elementary and middle school levels and gifted, the same significant parameter estimates were identified as in the previous models (s ee Table 27). At the intercept, the estimates for time, time2, time3, elementary when compared to middle, free or reduced lunch status, minority, students with disabilities, gifted, positive learning environment, and positive teacher qua lifications were significant. At the intercept LEP was not significant. Variables that had significant interactions with time were elementary, LEP, gifted, positive learning environment, and positive teacher qualifications. Variables with significant

PAGE 159

145 interactions with time2 and time3 were elementary, minority, LEP, gifted, and positive learning environment. When examining the technology indicators, as in the previous model, the intercept of percent of teachers who regularly use tech nology for administrative purpos es was significant. Significant technology indicators with time in cluded both the percent of teachers who regularly use technology to deliver instruction and the percent of teachers who use technology for administrative purposes. Significant technology indicators with time2 and time3 included the percent of t eachers who use technology for administrative purposes. Although, there was no difference in the percentage of variance explained in this model than was in the Demographic Model with Student Learning Environment by school level or the Technology Integration with Demographic and Student Learning Environment Model by school level, the AIC, AICC, and BIC indices all indicated better model fit (see Table 29). The level-1 residuals for the final model for predicting FCAT Math using elementary and middle schools with gifted ranged between -23.46 and 22.13 with a standard deviation of 3.12. Although there were outliers, skewness was -0.01 and kurtosis was 2.74, which would indicate that the residuals were evenly distri buted with most around the mean. Distribution of the empirical bayes intercepts ranged between -23.26 and 28.25 with standard deviation of 6.18. Skewness was 0.31, and kurtosis was 0.44, which indicated that the residuals at level-2 were also normally distributed. Final Model 7b: Technology Integration with Demographics and Student Learning Environment by School Level (Elementary and Middle Schools with Gifted) Level 1: FCAT Math = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 09*Teachers Use Deliver Instruction + 010* Teachers use Admin + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16*Gifted + 17* Teacher Qualifications + 18* Positive Learning Environment + 19* Teachers Use Deliver Instruction + 110*Teachers use Admin 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26*Gifted + 27* Teacher Qualifications + 28* Positive Learning Environment + 29* Teachers Use Deliver Instruction + 210*Teachers use Admin 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* LEP + 35* SWD + 36*Gifted + 37* Teacher Qualifications + 38* Positive Learning Environment + 39* Teachers Use Deliver Instruction + 310*Teachers use Admin

PAGE 160

146 Mixed-Effects Model: FCAT Math = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 09* Teachers Use Deliver Instruction + 010*Teachers use Admin + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Gifted*Time + 17* Teacher Qualifications*Time + 18* Positive Learning Environment*Time + 19* Teachers Use Deliver Instruction *Time + 110*Teachers use Admin*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Gifted*Time2 + 27* Teacher Qualifications*Time2 + 28* Positive Learning Environment*Time2 + 29* Teachers Use Deliver Instruction*Time2 + 210* Teachers use Admin*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34* LEP*Time3 + 35* SWD*Time3 + 36*Gifted*Time3 + 37* Teacher Qualifications*Time3 + 38* Positive Learning Environment*Time3 + 39* Teachers Use Deliver Instruction*Time3 + 310* Teachers use Admin*Time3 + u0 + r Table 27. Final Model 7b: Math Predicted by Significant Technology Integration with Demographics and Student Learning Environment by School Level for Elementary and Middle Schools with Gifted Effect School Level Estimate SE df t p Intercept 678.81 0.4327 1819 1568.8 <.0001 ** Time -6.2536 0.9859 4925 -6.34 <.0001 ** Time2 9.6551 0.8469 4925 11.4 <.0001 ** Time3 -2.2131 0.1825 4925 -12.13 <.0001 ** School Level Elementary -37.8793 0.5238 4925 -72.31 <.0001 ** School Level Middle 0 . . Free Reduced Lunch -4.5121 0.224 4925 -20.14 <.0001 ** Minority -5.7575 0.2604 4925 -22.11 <.0001 ** LEP 0.04982 0.2065 4925 0.24 0.8093 Students with Disabilities -1.3011 0.1553 4925 -8.38 <.0001 ** Gifted 3.3718 0.1801 4925 18.72 <.0001 ** Positive Learning Environment 1.4159 0.1924 4925 7.36 <.0001 ** Positive Teacher Qualifications 0.8636 0.1279 4925 6.75 <.0001 ** Teachers Use To Deliver Instruction 0.1154 0.1269 4925 0.91 0.3632 Teachers Use For Administrative Purposes -0.359 0.1341 4925 -2.68 0.0075 ** Time*School Level Elementary -9.0999 1.2532 4925 -7.26 <.0001 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch -0.9663 0.6052 4925 -1.6 0.1104 Time*Minority -1.1046 0.5861 4925 -1.88 0.0595 Time*LEP 1.1634 0.4559 4925 2.55 0.0107 Time*Students with Disabilities -0.417 0.387 4925 -1.08 0.2812 Time*Gifted 2.1235 0.4074 4925 5.21 <.0001 ** Time*Positive Learning Environment -1.3146 0.608 4925 -2.16 0.0306 Time*Positive Teacher Qualifications 0.8314 0.3645 4925 2.28 0.0226

PAGE 161

147 Effect School Level Estimate SE df t p Time*Teachers Use To Deliver Instruction -0.882 0.448 4925 -1.97 0.049 Time*Teachers Use For Administrative Purposes 2.0823 0.4632 4925 4.5 <.0001 ** Time2*School Level Elementary 5.0284 1.0893 4925 4.62 <.0001 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.1757 0.5374 4925 -0.33 0.7438 Time2*Minority 1.4354 0.5128 4925 2.8 0.0051 ** Time2*LEP -0.9941 0.398 4925 -2.5 0.0125 Time2*Students with Disabilities 0.3661 0.3376 4925 1.08 0.2783 Time2*Gifted -1.3546 0.3552 4925 -3.81 0.0001 ** Time2*Positive Learning Environment 1.4849 0.5367 4925 2.77 0.0057 ** Time2*Positive Teacher Qualifications -0.5787 0.3199 4925 -1.81 0.0705 Time2*Teach Use Deliver Instruction 0.7232 0.3936 4925 1.84 0.0662 Time2*Teach Use Administrative Purposes -1.5144 0.4023 4925 -3.76 0.0002 ** Time3*School Level Elementary -0.56 0.2365 4925 -2.37 0.0179 Time3*School Level Middle 0 . . Time3*Free Reduced Lunch 0.09116 0.1182 4925 0.77 0.4404 Time3*Minority -0.3355 0.1122 4925 -2.99 0.0028 ** Time3*LEP 0.2157 0.08672 4925 2.49 0.0129 Time3*Students with Disabilities -0.06832 0.07359 4925 -0.93 0.3533 Time3*Gifted 0.2376 0.07747 4925 3.07 0.0022 ** Time3*Positive Learning Environment -0.3462 0.1174 4925 -2.95 0.0032 ** Time3*Positive Teacher Qualifications 0.1038 0.06993 4925 1.48 0.1378 Time3*Teach Use Deliver Instruction -0.1492 0.08613 4925 -1.73 0.0833 Time3*Teach Use Administrative Purposes 0.302 0.08757 4925 3.45 0.0006 ** Covariance Parameter Estimate SE z p 41.5929 1.6533 25.16 <.0001 ** Residual 12.9167 0.2683 48.14 <.0001 ** Note: p < .05; ** p < .01 The last step was to add in USDOE funded Magnet Schools and USDOE Technology Magnet Schools as variables in the model. Results of this model indicated that neither magnet school status nor technology magnet school status was a significant predic tor of FCAT Math with either the data with all school levels without gifted or with the data with elementary and middle schools and gifted.

PAGE 162

148 Table 28. Model Fit Indices for Models Predicting FCAT Ma th Scores for All Sch ool Levels (without Gifted and LEP) Model -2 Log Likelihood AIC (smaller is better) AICC (smaller is better) BIC (smaller is better) Model 1: Math Predicted by Average Math of All Schools in Florida 71441.4 71447.4 71447.4 71464.6 Model 2a: Time as a Predictor of Math 67179.8 67191.8 67191.9 67226.3 Model 2b: Time as a Predictor of Math Time Fixed 67187.7 67195.7 67195.7 67218.7 Quadratic Model 2c: Time2 as a Predictor of Math 66206.6 66216.6 66216.6 66245.3 Polynomial Model 2d: Time3 as a Predictor of Math 64550.4 64562.4 64562.4 64596.8 Model 3: Time, Time2, Time3, and School Level as Predictors of Math 60173.2 60201.2 60201.3 60281.6 Model 4a: Math predicted by Time, School Level, and Demographics Variables 57269 57321 57321.1 57470.2 Model 5a: Demographics and Student Learning Environment by School Level 56865 56933 56933.3 57128.2 Model 6a: Technology Integration with Demographics and Student Learning Environment by School Level 56795.4 56935.4 56936.5 57337.3 Final Model 7a: Significant Technology Integration with Demographics and Student Learning Environment by School Level 56826.4 56902.4 56902.7 57120.6 Table 29. Model Fit Indices for Models Predicting FCAT Math Scores for Elementary and Middle School Levels (with Gifted) Model -2 Log Likelihood AIC (smaller is better) AICC (smaller is better) BIC (smaller is better) Model 1: Math Predicted by Average Math of All Elementary and Middle Schools in Florida 59990.7 59996.7 59996.7 60013.5 Model 4b: Math predicted by Time, School Level, and Demographics Variables No High School includes gifted 41520.2 41580.2 41580.5 41745.5 Model 5b: Demographics and Teacher Qualifications by School Level 41293.9 41369.9 41370.4 41579.2 Model 6b: Technology Integration with Demographics and Teacher Qualifications by School Level 41224.5 41372.5 41374.2 41780.1

PAGE 163

149 Model -2 Log Likelihood AIC (smaller is better) AICC (smaller is better) BIC (smaller is better) Final Model 7b: Significant Technology Integration with Demographics and Student Learning Environment by School Level 41253.6 41345.6 41346.2 41598.9 The result of the analysis for all the models indicated that Hypothesis 2 was partially correct. When the sample included schools at all three school levels, there was a significant negative relationship between the percent of teachers who regularly use tech nology for administrative pu rposes and the intercept of school level FCAT Math achievement when all other school level, demographic, and school learning environment factors were controlled. Also, there were significant interactions between the percent of teachers who regularly use technology for administrative purposes and time, time2, and time3 with FCAT Math achievement. There was a significant positive interaction between time and time3 with the percent of teachers who regularly use technology for administra tive purposes with FCAT Math achievement, and a significant negative interaction between time2 and the percent of teachers who regularly use technology for administrative purposes with FCAT Math achievemen t. These interactions resulted in an s-shaped curvilinear trend. After controlling so that all other variables were held at the mean, the trend for each school level could be examined separately, by comparing schools with different levels in that teachers use technology for administrative purposes. Figure 15 illustrates th e relationship between the pe rcentage of teachers who regularly use technology for administrative purposes and average school FCAT Math score for high schools. Percentage of teachers who regularly use tech nology for administrative pu rposes were compared at one and two standard deviations below the mean, the mean, and one and two standard deviations above the mean. This allows the extreme cases of schools that have the percen tage of teachers who regularly use technology for administrative purposes, +2 standard deviations above the mean, and schools that have the percentage of teachers who regularly use technology fo r administrative purposes the least often, -2 standard deviations below the mean to be compared. Schools that had the 2 standard deviations, 1 standard deviation and at the mean percentage of teach ers who regularly use technology for administrative purposes started the study in 2003-04 with the highest FCAT Math scores (707) and schools that had 1 and 2 standard deviations above the mean of percentage of teach ers who regularly use technology for administrative

PAGE 164

150 purposes had started with the lowest FCAT Math scores (706). This difference of one point was significant because there were so many schools in the sample ; however, the practical importance was modest. The interaction between the percentage of teachers who re gularly use technology for administrative purposes and time, time2, and time3 with FCAT Math scores was significant, so the slopes of the trends at each level were significantly curvilinear and s-shaped. By 2004-0 5, all high schools decreased their average FCAT Math scores; however, schools that had 2 standard devi ations below the mean in the percentage of teachers who regularly use technology for administrative purposes had the greatest decline in scores. All high schools experienced gains in average FCAT Math scores between 2004-05 and 2005-06 (713 and 714) and then a decline between 2005-06 and 2006-07. High schools with 2 standard deviations above the mean decreased the least (one point to 713), while high schools with 2 standard deviations below the mean decreased the most (two points to 711). Relationship between Percent of Teachers Regularly Use Technology for Administrative Purposes and FCAT Math in High Schools702 704 706 708 710 712 714 716 2003200420052006 YearFCAT Math High Schools + 2SD High Schools + 1SD High Schools mean High Schools 1SD High Schools 2SD Figure 15. Relationship between the Percent of Teachers Who Regularly Use Technology for Administrative Purposes and FCAT Math in High Schools.

PAGE 165

151 Middle schools had a similar beginning pattern to high school, that is, after controlling for all other factors, schools that were tw o standard deviations above the mean in the percentage of teachers who regularly use technology for administrative purposes had the lowest FCAT Math scores in 2003-04 (680), while those with two standard deviations below the mean had the highest scores (681). Although this difference of one point was significant due to the large sample size, the practical importance is modest. There were significant interactions between time, time2, time3 and the percentage of teachers in middle schools that regularly use technology for administrativ e purposes with FCAT Math scores (see Figure 16). Between 2003-04 and 2004-05 middle schools with 2 standard deviations below the mean maintained their FCAT Math scores (681) while all other levels increased their scores (-1SD = 681; Mean, +1SD, and +2SD = 682). After this all schools increased their FCAT Math score between 2004-05 and 2005-06 (-1SD = 688; Mean, +1SD, and +2SD = 689). Then between 2005-06 and 2006-07 differences expanded with different trends. Middle schools with two standard deviations below the mean remained at 688, while middle schools one standard deviation below the mean decreased to 688. Middle schools at the mean and one standard deviation above the mean remained the same, and middle schools with two standard deviations above the mean gained one point (690). Although these changes were significant, the practical importance is modest.

PAGE 166

152 Relationship between Percent of Teachers Regularly Use Technology for Administrative Purposes and FCAT Math in Middle Schools678 680 682 684 686 688 690 692 2003200420052006 YearFCAT Math Middle Schools + 2SD Middle Schools + 1SD Middle Schools mean Middle Schools 1SD Middle Schools 2SD Figure 16. Relationship between the Percent of Teachers Who Regularly Use Technology for Administrative Purposes and FCAT Math in Middle Schools. Elementary schools experienced a similar pattern to middle schools (see Figure 17). Schools with the lowest percent of teachers who regularly use technology for administrative purposes began the study with the highest FCAT Math score (641), while all other levels or standard deviations had the same score (640). Although this difference of one point was sta tistically significant, it had no practical importance. Between 2003-04 and 2004-05, elementary schools at th e mean and one and two standard deviations below the mean experienced the greatest d ecline in FCAT Math scores (636), while elementary schools with one and two standard deviations above the mean experien ced the least decline in mean FCAT Math scores (637). Between 2004-05 and 2005-06 all schools experienced gains in mean school FCAT Math scores (646). Between 2005-06 and 2006-07 all elementary schools continue to make gains with schools one and two standard deviations above the mean at 652 and schools at the mean, one and two standard deviations below the mean at 651. Although significant the differences of one FCAT Math score is not of practical importance.

PAGE 167

153 Relationship between Percent of Teachers Regularly Use Technology for Administrative Purposes and FCAT Math in Elementary Schools634 636 638 640 642 644 646 648 650 652 654 2003200420052006 YearFCAT Math Elementary + 2SD Elementary + 1SD Elementary mean Elementary 1SD Elementary 2SD Figure 17. Relationship between the Percent of Teachers Who Regularly Use Technology for Administrative Purposes and FCAT Math in Elementary Schools. When the sample was restricted to just elementary and middle schools and percent of gifted students was included in the equation, both the intercept of gifted and the interactions of percent of gifted students in the school with time, time2, and time3. Thus, when all other factors were held equal, schools with highest percentages of gifted students began the study with the highest FCAT Math scores, and the trends were not constant (see Figure 18). In addition, the trends were different at elementary and middle school level.

PAGE 168

154 Relationship between Percent of Gifted Students on FCAT Math by School Level (Gifted Included)620 630 640 650 660 670 680 690 700 2003200420052006 YearFCAT Math Elementary + 2SD Elementary + 1SD Elementary mean Elementary 1SD Elementary 2SD Middle Schools + 2SD Middle Schools + 1SD Middle Schools mean Middle Schools 1SD Middle Schools 2SD Figure 18. Relationship between Percent of Gifted Studen ts on FCAT Math by School Level (Gifted Included). When examining the parameter estimates of the technology integration indicators within these data, there were significant main effects for relations hips between FCAT Math scores and the percent of teachers who regularly use technology for administra tive purposes. In addition, there were significant interactions between time and two technology integration indicators: th e percent of teachers who regularly use technology for delivery of inst ruction and the percent of teachers who regularly use technology for administrative purposes. The interactions between time2 and time3 and the percentage of teachers who regularly use technology for administrative purposes were significant. In order to visualize the significant relationships of each of these technology integration variables with FCAT Math, the trends are depicted in separate charts after contro lling for all other factors. There was a significant interaction with time and the percentage of teachers who regularly use technology for administrative purposes and FCAT Math scores for middle schools and elementary schools with gifted. Each school level was examined separately. One and two standard deviations above the mean, the mean, and one and two standard deviations below the mean of leve ls of percentages of teachers who

PAGE 169

155 regularly use technology for administ rative purposes were compared afte r controlling for all other factors. In 2003-04, middle schools with two and one standard deviation below the mean started with the highest average FCAT Math score (675), while schools at the mean and one and two standard deviations above the mean had the lowest FCAT Math score (674) (see Figu re 19). The scores for middle schools that were two standard deviations below the mean for percentages of teachers who use technology for administrative purposes decreased to 674 in 2004-05, and schools one standard deviation below the mean remained the same (675). Middle schools at the mean increased on e point to 675, at one standard deviation above the mean increased two points to 676, and schools at two standard deviations above the mean increase three points to 677. All middle schools increased mean FCAT Math scores between 2004-05 and 2005-06 (-2 SD = 682; -1 SD, Mean, +1 SD, and +2 SD = 683). Between 2005-06 and 2006-07 middle schools at two standard deviations below the mean for percentages of teachers who use technology for administrative purposes remained the same at 682, and schools at the mean and one standard deviation above the mean remained the same at 683. Schools that were one standard deviation below the mean decreased to 682, while middle schools that were two standard deviations above the mean increased to 684 (see Figure 19). Again, this significant difference of two points between the extremes did not have practical importance.

PAGE 170

156 Relationship between Percent of Teachers Use Technology for Administrative Purposes and FCAT Math in Middle Schools (Gifted Included)672 674 676 678 680 682 684 686 2003200420052006 YearFCAT Math Middle Schools + 2SD Middle Schools + 1SD Middle Schools mean Middle Schools 1SD Middle Schools 2SD Figure 19. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and FCAT Math in Middle Schools (Gifted Included). The trends for elementary schools followed a very similar pattern to that of middle schools (see Figure 20). In 2003-04, elementary schools with two standard deviations above th e mean of percentage of teachers who regularly use technology for administra tive purposes started with the lowest mean FCAT Math scores (640), while schools with two standard deviations below the mean started with the highest FCAT Math score (642). Elementary schools with all ot her levels of percentage of teachers who regularly use technology for administrative purposes started with mean FCAT Math score of 641. Although this difference was significant, the range of two points did not have practical importance. In 2004-05 the average school FCAT Math score declined; however, schools with the least or two standard deviations below the mean for percentage of teachers who regu larly use technology for administrative purposes declined six points (636), while elementary schools with two standard deviations above the mean declined only one point (639). The trend for elementary schools with all levels of percentage of teachers who regularly use technology for administrative purposes was up in 2005-06. In 2006-07 (-2 SD = 646; -1 SD,

PAGE 171

157 Mean +1 SD, and +2 SD = 647). Elementary school mean FCAT Math scores continued to increase in 2006-07, but at different rates (-2 SD = 651; -1 SD, and Mean = 652; +1 SD and +2 SD = 653). Relationship between Percent of Teachers Use Technology for Administrative Purposes and FCAT Math in Elementary Schools (Gifted Included)634 636 638 640 642 644 646 648 650 652 654 2003200420052006 YearFCAT Math Elementary + 2SD Elementary + 1SD Elementary mean Elementary 1SD Elementary 2SD Figure 20. Relationship between Percent of Teachers Wh o Regularly Use Technol ogy for Administrative Purposes and FCAT Math in Elementary Schools (Gifted Included). The differences in standard deviations of intercep t of the percentage of teachers who regularly use technology for delivery of instruction were not signific ant at the intercept for predicting FCAT Math scores for elementary schools and middle schools with gifted. However, the interaction between time and the percentage of teachers who regularly use technology fo r delivery of instruction with FCAT Math score was significant. The interactions with time2 and time3 were not significant. Both levels of school have similar trends. Charts were made for each level of school to visualize the relationship between the percentage of teachers who regularly use technology for delivery of instruction and FC AT Math achievement at one and two standard deviations above the mean, the mean, and one and two standard deviations below the mean. The trends for middle school level at two standard deviations above the mean, the mean, and two standard deviations below the mean of the percentage of teachers who regularly use technology for delivery of instruction were examined (see Figure 21). When controlling for all other variables, middle schools at

PAGE 172

158 two standard deviations above the mean of the perc entage of teachers who regu larly use technology for delivery of instruction had the highest FCAT Math scores (675) at the intercept, while all other levels had 674. Between 2003-04 and 2004-05 middle schools at two standard deviations above the mean of percent of teachers who regularly use techno logy for delivery of instruction ma intained their FCAT math scores (675), while middle schools at the mean and one standard deviation above the mean increased to 675. Schools with one and two standard deviations below the mean of per cent of teachers who regularly use technology for delivery of instruction had the greatest in crease in mean FCAT Math scores to 676. Between 2004-05 and 2005-06 middle schools at all levels of percent of teachers who regu larly use technology for delivery of instruction increased mean FCAT Math score to 683, which remained the same through 200607. Relationship between Percent of Teachers Use Technology for Delivery of Instruction and FCAT Math in Middle Schools (Gifted Included)673 674 675 676 677 678 679 680 681 682 683 684 2003200420052006 YearFCAT Math Middle Schools + 2SD Middle Schools + 1SD Middle Schools mean Middle Schools 1SD Middle Schools 2SD Figure 21. Relationship between the Percentage of Teachers Who Regularly Use Technology for Delivery of Instruction and FCAT Math in Middle Schools (Gifted Included). When controlling for all other variables, elementary schools at all levels for percent of teachers who regularly use technology for delivery of instructio n started with FCAT Math scores at 641 (see Figure

PAGE 173

159 22). Between 2003-04 and 2004-05 elementary schools that were at the mean and one and two standard deviations above the mean had decreases of 4 points in mean FCAT Math scores (637), while elementary schools that were one and two standard deviations below the mean in percent of teachers who regularly use technology for delivery of instruction 3 points in mean FCAT Math score (638). Between 2004-05 and 2005-06 elementary schools at all levels of percent of teachers who regularly use technology for delivery of instruction increased their mean FCAT Math to 647 and then increased again to 652 in 2006-07. Relationship between Percent of Teachers Use Technology for Delivery of Instruction and FCAT Math in Elementary Schools (Gifted Included)636 638 640 642 644 646 648 650 652 654 2003200420052006YearFCAT Math Elementary + 2SD Elementary + 1SD Elementary mean Elementary 1SD Elementary 2SD Figure 22. Relationship between the Percentage of Teachers Who Regularly Use Technology for Delivery of Instruction and FCAT Math in Elementary Schools (Gifted Included). Hypothesis 3 The third analysis conducted to answer the first research question used the FCAT Writing outcome data to test the following hypothesis: H3: After controlling for school level (elementary, middle, and high), school socio-economic status, minority, limited English proficiency, students with disabilities, gifted, teacher qualification, and learning environment quality, mean school writing achievement (FCAT rubric scores for Writing) will have a positive relationshi p with indicators of technology integration.

PAGE 174

160 The first step was to build the unconditional model. The unconditional model predicted the schools’ FCAT Writing from the average of FCAT Writing for all schools. There were no other predictors. The average FCAT for all schools was 3.809 points ( t (2263) = 654.42, p <.0001). This model explained 34% of the within schools variance and 66% of the between schools variance. Model 1: Unconditional Model Level 1: FCAT Writing = 0 + r Level 2: 0 = 00 + u0 Mixed-Effects Model: FCAT Writing = 00 + u0 + r The intraclass correlation coeffici ent (ICC) was computed to determine the proportion of variance in the FCAT Writing variable that is accounted for by the schools. The ICC was .66, which is high, indicates nested data, and supports using multi-level m odeling for the analysis. The model fit statistics from this model were used as the baseline for model comparisons (see Table 28). Next, time was added as a predictor to the equation to make the unconditional growth model (see Model 2a). The variance components from this analys is showed how much of the variance in the model was accounted for by time. Time was significant in this equation ( z = 14.6, p = <.0001), which indicated that there was variance in the slopes between schools. Therefore, time was set as a random effect, and the model was estimated. Model 2a: Unconditional Growth Model Level 1: FCAT Writing = 0 + 1*Time + r Level 2: 0 = 00 + u0 1 = 10 + u1 Mixed-Effects Model: FCAT Writing = 00 + 10*Time + u0 + u1*Time + r Both the intercept ( t (2263) = 582.39, p <.0001) and time ( t (2263) = 42.39, p <.0001) were significant parameters. The variance between schools increased by 14%, and the variance explained within schools increased (10%) when time was added to the model. To determine if the equation was not linear but curvilinear, time2 was added to the equation so the variance could be compared. Results indicated that time2 was significant ( t (6901) = 32.47, p <.0001); however, adding time2 did not explain any additional variance (see Model 2b). When time3 was added to the equation with time2, time3 was also significant ( t (6900) = 43.26, p <.0001), and all model fit indices improved. Although adding time3 increased the amount of

PAGE 175

161 variance between schools, it increased the variance explained within schools by an additional 2%. Consequently, both time2 and time3 were retained in the polynomial growth model equation (see Model 2c). Model 2b: Quadratic Growth Model Level 1: FCAT Writing = 0 + 1*Time + 2* Time2 + r Level 2: 0 = 00 + u0 1 = 10 + u1 2 = 20 Mixed-Effects Model: FCAT Writing = 00 + 10*Time + 20* Time2 + u0 + u1 + r Model 2c: Polynomial Growth Model Level 1: FCAT Writing = 0 + 1*Time + 2* Time2 + 3* Time3 + r Level 2: 0 = 00 + u0 1 = 10 + u1 2 = 20 2 = 30 Mixed-Effects Model: FCAT Writing = 00 + 10*Time + 20* Time2 + 30* Time3 + u0 + u1 + r Next, school level was added to the Polynomi al Growth Model to predict FCAT Writing (see Model 3). The significance of the parameter estimates de termined if school level was significantly related to the FCAT Writing and if there was an interaction w ith time. This model adjusted the mean school FCAT Writing and the slope of FCAT Writing growth for school level. The parameter estimates for the intercept, time, time2, time3, and elementary school level when compared to middle school were all significant. The interactions between time, time2, and time3 with high school level compared to middle school were all significant. The interaction between time3 and elementary school relative to middle school was also significant. All model fit indices indicated improved f it with this model (Table 39), even though they were negative, because lower fit statistics indicate better fit (Luke, 2004). The deviance for Model 1 is -332. The deviance for Model 2 is -334.5. A lower deviance always implies better fit…The level-2 slopes model (Model 3) [-353.8] is significantly better (Luke, 2004, p. 34) This model still did not account for any additional between school variance, but did account for an additional 2% of the within school variance from the Polynomial Growth Model. Model 3: School Level as Predictor Level 1: FCAT Writing = 0 + 1*Time + 2*Time2 + 3* Time3 + r Level 2: 0 = 00 + 01*School Level + u0 1 = 10 + 11*School Level + u1 2 = 20 + 21*School Level 3 = 30 + 31*School Level

PAGE 176

162 Mixed-Effects Model: FCAT Writing = 00 + 01*School Level + 10*Time + 11*School Level*Time + 20*Time2 + 21*School Level*Time2 + 30*Time3 + 31*School Level*Time3 + u0 + u1 + r Table 30. Model 3: Time, Time2, Time3, and School Level as Predictors of Writing Effect School Level Estimate SE df t p Intercept 3.7927 0.0145 2261 261.6 <.0001 ** Time -0.0876 0.02506 2261 -3.5 0.0005 ** Time2 0.1389 0.02202 4522 6.31 <.0001 ** Time3 -0.02407 0.00484 4522 -4.97 <.0001 ** School Level Elementary -0.1512 0.0165 4522 -9.16 <.0001 ** School Level High 0.03239 0.02179 4522 1.49 0.1372 School Level Middle 0 . . Time*School Level Elementary 0.03876 0.02852 4522 1.36 0.1742 Time*School Level High 0.1059 0.03767 4522 2.81 0.005 ** Time*School Level Middle 0 . . Time2*School Level Elementary 0.004747 0.02506 4522 0.19 0.8498 Time2*School Level High -0.1164 0.0331 4522 -3.52 0.0004 ** Time2*School Level Middle 0 . . Time3*School Level Elementary -0.01115 0.005508 4522 -2.02 0.043 Time3*School Level High 0.01945 0.007275 4522 2.67 0.0075 ** Time3*School Level Middle 0 . . Covariance Parameter Estimate SE z p 0.07343 0.00258 28.46 <.0001 ** -0.00483 0.000571 -8.47 <.0001 ** 0.003258 0.00022 14.78 <.0001 ** Residual 0.01843 0.000387 47.58 <.0001 ** Note: p < .05; ** p < .01 The next model added student demographic variables to the School Level Model. This model was estimated twice. The first time, the model was estimate d with high school as a school level and all of the demographic variables except gifted (see Model 4a). The second time, the data were filtered to exclude high school as a school level and keep the gifted variable with middle and elementary schools (see Model 4b). The model fit statistics of the demographic model with all three school levels was compared with the School Level as Predictor Model to determine if there was a better fit (see Table 39). The significance of the parameter estimates determined which of the demographic variables remained in the predictor equation (see Table 31). The variance estimates showed the amount of the total variance that was accounted for by

PAGE 177

163 each model. When all of the demographics variables ex cept gifted were added to the model, the intercept was significant and the average middle school started with FCAT Writing score of 3.75 ( t (2219) = 304.15, p <.0001). The parameter estimates for time, time2, time3, school level, free or reduced lunch status, minority, LEP, and students with disabilities were signifi cant (see Model 4a). Interactions with time and time2 were significant for high school level relative to middle school. Interactions with elementary school and time3 were significant. All model fit indices indicated better fit with the addition of these demographics variables (see Table 39). Adding the demographics variables with school level explained 36% of the between school variance and 48% of the within school va riance for a total of 40% of all variance explained. Model 4a: Demographics by School Level (including High School and no Gifted) Level 1: FCAT Writing = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + u1 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* LEP + 35* SWD Mixed-Effects Model: FCAT Writing = 00 + 01*School Level + 02*SES + 03*Minority + 04*SWD + 05*LEP + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14*SWD*Time + 15*LEP*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24*SWD*Time2 + 25*LEP*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34*SWD*Time3 + 35*LEP*Time3 + u0 + u1 + r Table 31. Model 4a: Writing predicted by Time, School Level, and Demographics Variables No Gifted Effect School Level Estimate SE df t p Intercept 3.7531 0.01234 2219 304.15 <.0001 ** Time -0.07979 0.02597 2198 -3.07 0.0021 ** Time2 0.1446 0.02269 4210 6.37 <.0001 ** Time3 -0.02588 0.004978 4210 -5.2 <.0001 ** School Level Elementary -0.08797 0.0142 4210 -6.19 <.0001 ** School Level High -0.06373 0.01853 4210 -3.44 0.0006 ** School Level Middle 0 . . Free Reduced Lunch -0.1338 0.007245 4210 -18.47 <.0001 ** Minority 0.02856 0.008061 4210 3.54 0.0004 ** LEP -0.02587 0.006946 4210 -3.72 0.0002 ** Disabilities -0.06912 0.005461 4210 -12.66 <.0001 ** Time*School Level Elementary 0.006534 0.03003 4210 0.22 0.8278 Time*School Level High 0.08497 0.03916 4210 2.17 0.0301 Time*School Level Middle 0 . . Time*Free Reduced Lunch -0.02532 0.01776 4210 -1.43 0.1541 Time*Minority -0.02819 0.01823 4210 -1.55 0.1221 Time*LEP 0.01336 0.01506 4210 0.89 0.3752

PAGE 178

164 Effect School Level Estimate SE df t p Time*Disabilities -0.00319 0.0128 4210 -0.25 0.8031 Time2*School Level Elementary 0.02574 0.02628 4210 0.98 0.3274 Time2*School Level High -0.09024 0.03425 4210 -2.63 0.0085 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch 0.0214 0.01553 4210 1.38 0.1684 Time2*Minority 0.01391 0.0159 4210 0.87 0.3817 Time2*LEP 0.007535 0.01304 4210 0.58 0.5633 Time2*Disabilities 0.001642 0.01113 4210 0.15 0.8827 Time3*School Level Elementary -0.01502 0.005769 4210 -2.6 0.0093 ** Time3*School Level High 0.01369 0.007515 4210 1.82 0.0685 Time3*School Level Middle 0 . . Time3*Free Reduced L -0.00472 0.003404 4210 -1.39 0.1658 Time3*Minority -0.00192 0.003486 4210 -0.55 0.5821 Time3*LEP -0.00247 0.002845 4210 -0.87 0.3859 Time3*Disabilities 7.53E-06 0.002432 4210 0 0.9975 Covariance Parameter Estimate SE z p 0.04318 0.00176 24.53 <.0001 ** -0.00471 0.0005 -9.43 <.0001 ** 0.003181 0.000227 14.04 <.0001 ** Residual 0.0183 0.000397 46.12 <.0001 ** Note: p < .05; ** p < .01 For the elementary and middle schools with gifted students, the unconditional model was estimated to provide a baseline with which to compare the demographics model. The growth model was estimated to determine if time was fixed or random. Time in this dataset was also random because there was significant variance in the slope ( z = 43.78, p = <.0001). The results from the analysis in Model 4b indicated that the inter cept, school level, time, time2, time3, free or reduced lunch status, minority, LEP, students with disabilities, and gifted were all significant (see Table 32). The only significant interaction was between time3 and elementary when compared to middle sch ool. All of the model fit statistics indicated better model fit (see Table 40). When examining the variance of FCAT Writing in elementary and middle schools, adding demographics variables to the equation explained 36% of the between school variance and 51% of the within school variance. Two sets of analyses were conducted on the rest of the models in order to examine the relationship of gifted with technolo gy integration as one of the predictors of school achievement.

PAGE 179

165 Model 4b: Demographics by School Level (Elementary and Middle School only) Level 1: FCAT Writing = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16*Gifted + u1 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26*Gifted 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* LEP + 35* SWD + 36*Gifted Mixed-Effects Model: FCAT Writing = 00 + 01*School Level + 02*SES + 03*Minority + 04*SWD + 05*LEP + 06*Gifted + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP*Time + 15* SWD*Time + 16*Gifted*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Gifted*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34* LEP*Time3 + 35* SWD*Time3 + 36*Gifted*Time3 + u0 + u1 + r Table 32. Model 4b: write predicted by Time, School Level, an d Demographics Variables No High School includes Gifted Effect School Level Estimate SE df t p Intercept 3.7384 0.01268 1792 294.8 <.0001 ** Time -0.05448 0.02749 1725 -1.98 0.0477 Time2 0.1186 0.02408 3154 4.93 <.0001 ** Time3 -0.02014 0.005284 3154 -3.81 0.0001 ** School Level Elementary -0.06702 0.01473 3154 -4.55 <.0001 ** School Level Middle 0 . . Free Reduced Lunch -0.117 0.008646 3154 -13.53 <.0001 ** Minority 0.02455 0.009529 3154 2.58 0.01 LEP -0.01788 0.007535 3154 -2.37 0.0177 Students with Disabilities -0.05299 0.006174 3154 -8.58 <.0001 ** Gifted 0.0714 0.0066 3154 10.82 <.0001 ** Time*School Level Elementary 0.005895 0.03216 3154 0.18 0.8545 Time*School Level Middle 0 . . Time*Free Reduced Lunch -0.01305 0.02192 3154 -0.6 0.5517 Time*Minority -0.0376 0.02235 3154 -1.68 0.0926 Time*LEP 0.01729 0.01716 3154 1.01 0.3136 Time*Students with Disabilities 0.002766 0.01481 3154 0.19 0.8519 Time*Gifted -0.00719 0.01543 3154 -0.47 0.6415 Time2*School Level Elementary 0.03046 0.02818 3154 1.08 0.2799 Time2*School Level Middle 0 . . Time2*Free Reduced Lunch 0.01128 0.01915 3154 0.59 0.5558 Time2*Minority 0.02529 0.01944 3154 1.3 0.1933 Time2*LEP 0.000242 0.01482 3154 0.02 0.987 Time2*Students with Disabilities -0.00294 0.01284 3154 -0.23 0.8188 Time2*Gifted 0.000789 0.01337 3154 0.06 0.9529

PAGE 180

166 Effect School Level Estimate SE df t p Time3*School Level Elementary -0.01651 0.006187 3154 -2.67 0.0077 ** Time3*School Level Middle 0 . . Time3*Free Reduced Lunch -0.00248 0.004192 3154 -0.59 0.5536 Time3*Minority -0.00475 0.004252 3154 -1.12 0.264 Time3*LEP -0.00061 0.003225 3154 -0.19 0.8496 Time3*Students with Disabilities 0.000877 0.002797 3154 0.31 0.7538 Time3*Gifted 0.000129 0.002916 3154 0.04 0.9648 Covariance Parameter Estimate SE z p 0.04291 0.00199 21.57 <.0001 ** -0.00545 0.000598 -9.11 <.0001 ** 0.003522 0.000279 12.61 <.0001 ** Residual 0.01905 0.000473 40.23 <.0001 ** Note: p < .05; ** p < .01 The next model added the variable that measures the School Learning Environment factors to the Demographics Model by School Level Model. These in cluded teacher qualifica tions and positive learning environment. This model was estimated twice, first without gifted population but all school levels (see model 5a) and then with elementary and middle school levels and gifted population (see model 5b). When school learning environment factors were added with the demographic and school level variables for all school levels, the parameter estimates for the intercept, time2, time3, elementary and high school relative to middle school, free or reduced lunch status, minority, LEP, students with disabilities, teacher qualifications, and positive learning environment were significant (see Table 33). Time was the only variable that was not significant. The only signific ant interactions between time, time2, and time3 were with high school relative to middle school. Adding the student learning environment variables explained an additional 5% of the between school variance and maintained the same within school variance for a total of 43% of all of the variance explained. All of the model fit indices indicated that this model fit the data better (see Table 39). Model 5a: School Learning Environment with Demographics by School Level (All School Levels without Gifted) Level 1: FCAT Writing = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06* Teacher Qualifications + 07*Positive Learning Environment + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14*LEP + 15*SWD + 16* Teacher Qualifications + 17*Positive Learning Environment + u1

PAGE 181

167 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26* Teacher Qualifications + 27*Positive Learning Environment 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* LEP + 35* SWD + 6* Teacher Qualifications + 37*Positive Learning Environment Mixed-Effects Model: FCAT Writing = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Teacher Qualifications + 07* Positive Learning Environment + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14*LEP*Time + 15*SWD *Time + 16* Teacher Qualifications*Time + 17* Positive Learning Environment*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26* Teacher Qualifications*Time2 + 27* Positive Learning Environment*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34*LEP*Time3 + 35*SWD*Time3 + 36* Teacher Qualifications*Time3 + 37* Positive Learning Environment*Time3 + u0 + u1 + r Table 33. Model 5a: Demographics and Student Learning Envi ronment by School Level (All School Levels without Gifted) Effect School Level Estimate SE df t p Intercept 3.8014 0.01415 2219 268.72 <.0001 ** Time -0.04363 0.03273 2198 -1.33 0.1826 Time2 0.106 0.02804 4202 3.78 0.0002 ** Time3 -0.01751 0.006065 4202 -2.89 0.0039 ** School Level Elementary -0.1521 0.01703 4202 -8.93 <.0001 ** School Level High -0.06334 0.01832 4202 -3.46 0.0006 ** School Level Middle 0 . . Free Reduced Lunch -0.1205 0.007406 4202 -16.27 <.0001 ** Minority 0.03718 0.008151 4202 4.56 <.0001 ** LEP -0.03167 0.006865 4202 -4.61 <.0001 ** Students with Disabilities -0.06198 0.005491 4202 -11.29 <.0001 ** Positive Learning Environment 0.03944 0.006373 4202 6.19 <.0001 ** Positive Teacher Qualifications 0.03127 0.004659 4202 6.71 <.0001 ** Time*School Level Elementary -0.03113 0.04054 4202 -0.77 0.4426 Time*School Level High 0.09486 0.04006 4202 2.37 0.0179 Time*School Level Middle 0 . . Time*Free Reduced Lunch -0.01133 0.01936 4202 -0.59 0.5584 Time*Minority -0.02507 0.01873 4202 -1.34 0.1807 Time*LEP 0.00713 0.0154 4202 0.46 0.6434 Time*Students with Disabilities 0.002793 0.01313 4202 0.21 0.8315 Time*Positive Learning Environment 0.02237 0.01797 4202 1.24 0.2133 Time*Positive Teacher Qualifications 0.004514 0.0123 4202 0.37 0.7137 Time2*School Level Elementary 0.06425 0.03515 4202 1.83 0.0676

PAGE 182

168 Effect School Level Estimate SE df t p Time2*School Level High -0.09504 0.03502 4202 -2.71 0.0067 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch 0.01051 0.01707 4202 0.62 0.538 Time2*Minority 0.01014 0.01629 4202 0.62 0.5338 Time2*LEP 0.01104 0.01337 4202 0.83 0.4091 Time2*Students with Disabilities -0.00405 0.01139 4202 -0.36 0.7223 Time2*Positive Learning Environment -0.01875 0.01594 4202 -1.18 0.2395 Time2*Positive Teacher Qualifications -0.00872 0.01068 4202 -0.82 0.4139 Time3*School Level Elementary -0.02335 0.007659 4202 -3.05 0.0023 ** Time3*School Level High 0.01441 0.007677 4202 1.88 0.0605 Time3*School Level Middle 0 . . Time3*Free Reduced L -0.00231 0.003752 4202 -0.62 0.5379 Time3*Minority -0.00121 0.003568 4202 -0.34 0.734 Time3*LEP -0.00312 0.002921 4202 -1.07 0.2862 Time3*Students with Disabilities 0.001183 0.002488 4202 0.48 0.6343 Time3*Positive Learning Environment 0.003886 0.003508 4202 1.11 0.2681 Time3*Positive Teacher Qualifications 0.002476 0.002325 4202 1.06 0.287 Covariance Parameter Estimate SE z p 0.04012 0.001663 24.13 <.0001 ** -0.0045 0.000484 -9.31 <.0001 ** 0.003094 0.000224 13.81 <.0001 ** Residual 0.01828 0.000396 46.15 <.0001 ** Note: p < .05; ** p < .01 When the data were filtered to include only elementary and middle schools and gifted was also added to the equation, all intercept parameter estimates, elementary school, time2, time3, free or reduced lunch status, minor ity, LEP, students with disabilities, teacher qualifications, and positive learning environment were significant. Time was the only parameter that was not significant. The only significant interaction was between time3 and elementary (see Table 34). This model demonstrated better fit than the previous model by all model fit indices (see Table 40). It explained 2% more of the between school variance and the same amount of the within school variance than the previous model and explained 45% of all the variance.

PAGE 183

169 Model 5b: School Learning Environment with Demographics by School Level (Elementary and Middle Schools with Gifted) Level 1: FCAT Writing = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16*Gifted + 17* Teacher Qualifications + 18* Positive Learning Environment + u1 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26*Gifted + 27* Teacher Qualifications + 28* Positive Learning Environment 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* LEP + 35* SWD + 36*Gifted + 37* Teacher Qualifications + 38* Positive Learning Environment Mixed-Effects Model: FCAT Writing = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Gifted*Time + 17* Teacher Qualifications*Time + 18* Positive Learning Environment*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Gifted*Time2 + 27* Teacher Qualifications*Time2 + 28* Positive Learning Environment*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34* LEP*Time3 + 35* SWD*Time3 + 36*Gifted*Time3 + 37* Teacher Qualifications*Time3 + 38* Positive Learning Environment*Time3 + u0 + u1 + r Table 34. Model 5b: Demographics and Student Learning Envir onment by School Level for Elementary and Middle School with Gifted Effect School Level Estimate SE df t p Intercept 3.7837 0.01579 1792 239.59 <.0001 ** Time -0.0523 0.03765 1725 -1.39 0.165 Time2 0.1017 0.03201 3146 3.18 0.0015 ** Time3 -0.01523 0.00689 3146 -2.21 0.0272 School Level Elementary -0.1281 0.01949 3146 -6.58 <.0001 ** School Level Middle 0 . . Free Reduced Lunch -0.107 0.008804 3146 -12.15 <.0001 ** Minority 0.03147 0.009764 3146 3.22 0.0013 ** LEP -0.02388 0.00757 3146 -3.16 0.0016 ** Students with Disabilities -0.04995 0.006211 3146 -8.04 <.0001 ** Gifted 0.06232 0.006676 3146 9.34 <.0001 ** Positive Learning Environment 0.03653 0.008343 3146 4.38 <.0001 ** Positive Teacher Qualifications 0.02362 0.005412 3146 4.36 <.0001 ** Time*School Level Elementary 0.008295 0.04781 3146 0.17 0.8623 Time*School Level Middle 0 . . Time*Free Reduced Lunch -0.00783 0.02362 3146 -0.33 0.7402 Time*Minority -0.03798 0.02282 3146 -1.66 0.0962 Time*LEP 0.01399 0.01772 3146 0.79 0.4298

PAGE 184

170 Effect School Level Estimate SE df t p Time*Students with Disabilities 0.005403 0.01507 3146 0.36 0.72 Time*Gifted -0.00462 0.01577 3146 -0.29 0.7694 Time*Positive Learning Environment 0.000179 0.02374 3146 0.01 0.994 Time*Positive Teacher Qualifications -0.00434 0.0142 3146 -0.31 0.7597 Time2*School Level Elementary 0.04715 0.04112 3146 1.15 0.2515 Time2*School Level Middle 0 . . Time2*Free Reduced Lunch 0.00436 0.02081 3146 0.21 0.8341 Time2*Minority 0.02404 0.01977 3146 1.22 0.2242 Time2*LEP 0.003689 0.01536 3146 0.24 0.8102 Time2*Students with Disabilities -0.00594 0.01306 3146 -0.45 0.6494 Time2*Gifted 0.000945 0.01365 3146 0.07 0.9448 Time2*Positive Learning Environment -0.00946 0.0208 3146 -0.45 0.6492 Time2*Positive Teacher Qualifications -0.0025 0.01237 3146 -0.2 0.8398 Time3*School Level Elementary -0.02192 0.008914 3146 -2.46 0.014 Time3*School Level Middle 0 . . Time3*Free Reduced Lunch -0.00051 0.004566 3146 -0.11 0.9118 Time3*Minority -0.00447 0.004319 3146 -1.03 0.3011 Time3*LEP -0.0015 0.003346 3146 -0.45 0.6538 Time3*Students with Disabilities 0.001534 0.002843 3146 0.54 0.5895 Time3*Gifted -0.00019 0.002977 3146 -0.06 0.9486 Time3*Positive Learning Environment 0.003115 0.004542 3146 0.69 0.4928 Time3*Positive Teacher Qualifications 0.001369 0.002702 3146 0.51 0.6124 Covariance Parameter Estimate SE z p 0.04163 0.001942 21.44 <.0001 ** -0.00532 0.00059 -9.03 <.0001 ** 0.003493 0.000278 12.56 <.0001 ** Residual 0.01896 0.000471 40.25 <.0001 ** Note: p < .05; ** p < .01 The next model added technology integration variables with the demographics, learning environment, and school level variables. These included student access to various types of software, teachers regularly using various types of software, frequency that students use various types of software, and technology support. This model was estimated twi ce, first without gifted population but all school levels (see model 6a) and then with elementary and middle school levels and gifted population (see model

PAGE 185

171 6b). When the model was estimated with all school levels without gifted, the only significant technology parameter estimate was the frequency that students use content software (see Table 35). Other significant parameter estimates included the intercept, time2, time3, elementary and high school relative to middle school, free or reduced lunch status, minority, LEP, students with disabilities, positive learning environment, and positive teacher qualifications. Time was the only parameter that was not significant. The interactions of time and time2 with high school relative to middle school and the interaction of time3 with elementary school relative to middle school were also significant. Only -2 Log Likelihood index of model fit indicated that this model had better fit (see Table 39). This model explained no additional variance. One technology integration indicator was retained in the final model for all school levels without gifted, the frequency that students use content software. Model 6a: Technology Integration with Demographics and Student Learning Environment by School Level (All School Levels without Gifted) Level 1: FCAT Writing = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06* Teacher Qualifications + 07* Positive Learning Environment + 08*Access Content SW + 09*Access Office SW + 010*Access Ad Prod SW + 011*Teachers Use Deliver Instruction + 012*Teachers use Admin + 013*Frequency Students Use Content + 014*Frequency Students Use Tool + 015*Technical Support Human + 016*Technical Support Hardware + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16* Teacher Qualifications + 17* Positive Learning Environment + 18*Access Content SW + 19*Access Office SW + 110*Access Ad Prod SW + 111*Teachers Use Deliver Instruction + 112*Teachers use Admin + 113*Frequency Students Use Content + 114*Frequency Students Use Tool + 115*Technical Support Human + 116*Technical Support Hardware + u1 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26* Teacher Qualifications + 27* Positive Learning Environment + 28*Access Content SW + 29*Access Office SW + 210*Access Ad Prod SW + 211*Teachers Use Deliver Instruction + 212*Teachers use Admin + 213*Frequency Students Use Content + 214*Frequency Students Use Tool + 215*Technical Support Human + 216*Technical Support Hardware 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* LEP + 35* SWD + 36* Teacher Qualifications + 37* Positive Learning Environment + 38*Access Content SW + 39*Access Office SW + 310*Access Ad Prod SW + 311*Teachers Use Deliver Instruction + 312*Teachers use Admin + 313*Frequency Students Use Content + 314*Frequency Students Use Tool + 315*Technical Support Human + 316*Technical Support Hardware

PAGE 186

172 Mixed-Effects Model: FCAT Writing = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06* Teacher Qualifications + 07* Positive Learning Environment + 08*Access Content SW + 09*Access Office SW + 010*Access Ad Prod SW + 011*Teachers Use Deliver Instruction + 012*Teachers use Admin + 013*Frequency Students Use Content + 014*Frequency Students Use Tool + 015*Technical Support Human + 016*Technical Support Hardware + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Teacher Qualifications*Time + 17* Positive Learning Environment*Time + 18*Access Content SW*Time + 19*Access Office SW*Time + 110*Access Ad Prod SW*Time + 111*Teachers Use Deliver Instruction*Time + 112*Teachers use Admin*Time + 113*Frequency Students Use Content*Time + 114*Frequency Students Use Tool*Time + 115*Technical Support Human*Time + 116*Technical Support Hardware*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Teacher Qualifications*Time2 + 27* Positive Learning Environment*Time2 + 28*Access Content SW*Time2 + 29*Access Office SW*Time2 + 210*Access Ad Prod SW*Time2 + 211*Teachers Use Deliver Instruction*Time2 + 212*Teachers use Admin*Time2 + 213*Frequency Students Use Content*Time2 + 214*Frequency Students Use Tool*Time2 + 215*Technical Support Human*Time2 + 216*Technical Support Hardware*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34* LEP*Time3 + 35* SWD*Time3 + 36*Teacher Qualifications*Time3 + 37* Positive Learning Environment*Time3 + 38*Access Content SW*Time3 + 39*Access Office SW*Time3 + 310*Access Ad Prod SW*Time3 + 311*Teachers Use Deliver Instruction*Time3 + 312*Teachers use Admin*Time3 + 313*Frequency Students Use Content*Time3 + 314*Frequency Students Use Tool*Time3 + 315*Technical Support Human*Time3 + 316*Technical Support Hardware*Time3 + u0 + u1 + r Table 35. Model 6a: Technology Integration w ith Demographics and Student Learni ng Environment by School Level (All School Levels without Gifted) Effect School Level Estimate SE df t p Intercept 3.7976 0.01437 2219 264.29 <.0001 ** Time -0.05663 0.03441 2198 -1.65 0.1 Time2 0.1181 0.02954 4166 4 <.0001 ** Time3 -0.01994 0.006387 4166 -3.12 0.0018 ** School Level Elementary -0.147 0.01741 4166 -8.44 <.0001 ** School Level High -0.06442 0.01839 4166 -3.5 0.0005 ** School Level Middle 0 . . Free Reduced Lunch -0.1196 0.007533 4166 -15.88 <.0001 ** Minority 0.03726 0.008172 4166 4.56 <.0001 ** LEP -0.03057 0.006881 4166 -4.44 <.0001 ** Students with Disabilities -0.06104 0.005481 4166 -11.14 <.0001 ** Positive Learning Environment 0.03802 0.006388 4166 5.95 <.0001 ** Positive Teacher Qualifications 0.03128 0.004661 4166 6.71 <.0001 ** Access Content Software 0.007328 0.004649 4166 1.58 0.115 Access Office Software -0.00217 0.004609 4166 -0.47 0.6378 Access Advanced Production Software -0.00567 0.004704 4166 -1.21 0.2279 Teachers Use to Deliver 0.002398 0.005145 4166 0.47 0.6412

PAGE 187

173 Effect School Level Estimate SE df t p Instruction Teachers Use for Administrative Purposes -0.00102 0.005295 4166 -0.19 0.8468 Frequency that Students Use Content Software -0.01284 0.004256 4166 -3.02 0.0026 ** Frequency Students Use Tool-Based Software 0.007223 0.004579 4166 1.58 0.1147 Technical Support Human -0.00369 0.004219 4166 -0.88 0.3816 Technical Support Hardware -0.00452 0.004065 4166 -1.11 0.2662 Time*School Level Elementary -0.01378 0.04319 4166 -0.32 0.7496 Time*School Level High 0.09666 0.04068 4166 2.38 0.0175 Time*School Level Middle 0 . . Time*Free Reduced Lunch -0.00916 0.01978 4166 -0.46 0.6434 Time*Minority -0.01948 0.01896 4166 -1.03 0.3043 Time*LEP 0.004152 0.01548 4166 0.27 0.7886 Time*Students with Disabilities 0.005154 0.01315 4166 0.39 0.6951 Time*Positive Learning Environment 0.0242 0.01812 4166 1.34 0.1818 Time*Positive Teacher Qualifications 0.002198 0.01247 4166 0.18 0.8601 Time*Access Content Software -0.0102 0.01453 4166 -0.7 0.4827 Time*Access Office Software 0.01321 0.01443 4166 0.92 0.3602 Time*Access Advanced Production Software 0.0116 0.01442 4166 0.8 0.4212 Time*Teachers Use to Deliver Instruction 0.003405 0.01614 4166 0.21 0.8329 Time*Teachers Use for Administrative Purposes 0.01203 0.01641 4166 0.73 0.4636 Time*Frequency that Students Use Content Software 0.000921 0.01408 4166 0.07 0.9478 Time*Frequency Students Use Tool-Based Software -0.00387 0.01488 4166 -0.26 0.7948 Time*Technical Support Human 0.0101 0.01262 4166 0.8 0.4239 Time*Technical Support Hardware 0.007637 0.01285 4166 0.59 0.5523 Time2*School Level Elementary 0.0477 0.03762 4166 1.27 0.2049 Time2*School Level High -0.09453 0.03559 4166 -2.66 0.0079 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch 0.006524 0.01747 4166 0.37 0.7089 Time2*Minority 0.006105 0.01649 4166 0.37 0.7113 Time2*LEP 0.01324 0.01345 4166 0.98 0.3248 Time2*Students with Disabilities -0.00664 0.01141 4166 -0.58 0.5609 Time2*Positive Learning Environment -0.02039 0.01606 4166 -1.27 0.2043 Time2*Positive Teacher -0.00613 0.01089 4166 -0.56 0.5733

PAGE 188

174 Effect School Level Estimate SE df t p Qualifications Time2*Access Content Software 0.007102 0.01259 4166 0.56 0.5726 Time2*Access Office Software -0.00558 0.01247 4166 -0.45 0.6549 Time2*Access Advanced Production Software -0.00765 0.01249 4166 -0.61 0.5403 Time2*Teachers Use to Deliver Instruction -0.00457 0.014 4166 -0.33 0.7442 Time2*Teachers Use for Administrative Purposes -0.00971 0.01412 4166 -0.69 0.4918 Time2*Frequency that Students Use Content Software 0.006991 0.01225 4166 0.57 0.5681 Time2*Frequency Students Use Tool-Based Software -0.00509 0.01309 4166 -0.39 0.6975 Time2*Technical Support Human -0.00567 0.01099 4166 -0.52 0.6056 Time2*Technical Support Hardware 0.002376 0.0111 4166 0.21 0.8306 Time3*School Level Elementary -0.02003 0.008207 4166 -2.44 0.0147 Time3*School Level High 0.01433 0.007806 4166 1.84 0.0664 Time3*School Level Middle 0 . . Time3*Free Reduced Lunch -0.00133 0.003847 4166 -0.35 0.7298 Time3*Minority -0.00042 0.00361 4166 -0.12 0.9065 Time3*LEP -0.0036 0.002939 4166 -1.22 0.2209 Time3*Students with Disabilities 0.001799 0.002492 4166 0.72 0.4704 Time3*Positive Learning Environment 0.004256 0.003534 4166 1.2 0.2286 Time3*Positive Teacher Qualifications 0.001818 0.002379 4166 0.76 0.4446 Time3*Access Content Software -0.00119 0.002757 4166 -0.43 0.6666 Time3*Access Office Software 0.000633 0.002725 4166 0.23 0.8164 Time3*Access Advanced Production Software 0.001412 0.002726 4166 0.52 0.6044 Time3*Teachers Use to Deliver Instruction 0.001177 0.003068 4166 0.38 0.7012 Time3*Teachers Use for Administrative Purposes 0.0021 0.003077 4166 0.68 0.4949 Time3*Frequency that Students Use Content Software -0.00182 0.002679 4166 -0.68 0.497 Time3*Frequency Students Use Tool-Based Software 0.00149 0.002896 4166 0.51 0.6069 Time3*Technical Support Human 0.001105 0.002411 4166 0.46 0.6469 Time3*Technical Support Hardware -0.00164 0.002431 4166 -0.68 0.4995

PAGE 189

175 Covariance Parameter Estimate SE z p 0.03988 0.001656 24.08 <.0001 ** -0.00436 0.000479 -9.09 <.0001 ** 0.003014 0.000222 13.59 <.0001 ** Residual 0.01818 0.000395 46.08 <.0001 ** Note: p < .05; ** p < .01 No technology indicator was significant with the elementary and middle school data with gifted. The parameter estimates for the intercept, elementary school relative to middle school, time2, time3, free or reduced lunch status, minority, LEP, students with disa bilities, gifted, teacher qualifications, and positive learning environment were significant. Time was not si gnificant. The only interaction that was significant was time3and elementary relative to middle school (see Table 36). Only the -2 Log Likelihood index indicated that this model had better fit (see Table 40). Adding the technology integration indicators to the model did not explain any additional variance in the m odel. Because there were no significant technology indicators, all technology indicators were dropped from the model when magnet school status was added. Model 6b: Technology Integration with Demographics and Student Learning Environment by School Level (Elementary and Middle Schools with Gifted) Level 1: FCAT Writing = 0 + 1*Time + 2*Time2 + 3*Time3 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 09*Access Content SW + 010*Access Office SW + 011*Access Ad Prod SW + 012*Teachers Use Deliver Instruction + 013*Teachers use Admin + 014*Frequency Students Use Content + 015*Frequency Students Use Tool + 016*Technical Support Human + 017*Technical Support Hardware + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16*Gifted + 17* Teacher Qualifications + 18* Positive Learning Environment + 19*Access Content SW + 110*Access Office SW + 111*Access Ad Prod SW + 112*Teachers Use Deliver Instruction + 113*Teachers use Admin + 114*Frequency Students Use Content + 115*Frequency Students Use Tool + 116*Technical Support Human + 117*Technical Support Hardware + u1 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26*Gifted + 27* Teacher Qualifications + 28* Positive Learning Environment + 29*Access Content SW + 210*Access Office SW + 211*Access Ad Prod SW + 212*Teachers Use Deliver Instruction + 213*Teachers use Admin + 214*Frequency Students Use Content + 215*Frequency Students Use Tool + 216*Technical Support Human + 217*Technical Support Hardware 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* LEP + 35* SWD + 36*Gifted + 37* Teacher Qualifications + 38* Positive Learning Environment + 39*Access Content SW + 310*Access Office SW + 311*Access Ad Prod SW + 312*Teachers Use Deliver Instruction + 313*Teachers use Admin + 314*Frequency Students Use Content + 315*Frequency Students Use Tool + 316*Technical Support Human + 317*Technical Support Hardware

PAGE 190

176 Mixed-Effects Model: FCAT Writing = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 09*Access Content SW + 010*Access Office SW + 011*Access Ad Prod SW + 012*Teachers Use Deliver Instruction + 013*Teachers use Admin + 014*Frequency Students Use Content + 015*Frequency Students Use Tool + 016*Technical Support Human + 017*Technical Support Hardware + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Gifted*Time + 17* Teacher Qualifications*Time + 18* Positive Learning Environment*Time + 19*Access Content SW*Time + 110*Access Office SW*Time + 111*Access Ad Prod SW*Time + 112*Teachers Use Deliver Instruction*Time + 113*Teachers use Admin*Time + 114*Frequency Students Use Content*Time + 115*Frequency Students Use Tool*Time + 116*Technical Support Human*Time + 117*Technical Support Hardware*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Gifted*Time2 + 27* Teacher Qualifications*Time2 + 28* Positive Learning Environment*Time2 + 29*Access Content SW*Time2 + 210*Access Office SW*Time2 + 211*Access Ad Prod SW*Time2 + 212*Teachers Use Deliver Instruction*Time2 + 213*Teachers use Admin*Time2 + 214*Frequency Students Use Content*Time2 + 215*Frequency Students Use Tool*Time2 + 216*Technical Support Human*Time2 + 217*Technical Support Hardware*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34* LEP*Time3 + 35* SWD*Time3 + 36*Gifted*Time3 + 37* Teacher Qualifications*Time3 + 38* Positive Learning Environment*Time3 + 39*Access Content SW*Time3 + 310*Access Office SW*Time3 + 311*Access Ad Prod SW*Time3 + 312*Teachers Use Deliver Instruction*Time3 + 313*Teachers use Admin*Time3 + 314*Frequency Students Use Content*Time3 + 315*Frequency Students Use Tool*Time3 + 316*Technical Support Human*Time3 + 317*Technical Support Hardware*Time3 + u0 + u1 + r Table 36. Model 6b: Technology Integration w ith Demographics and Student Learni ng Environment by School Level for Elementary and Middle Schools with Gifted Effect School Level Estimate SE df t p Intercept 3.7821 0.01609 1792 235.03 <.0001 ** Time -0.06294 0.03962 1725 -1.59 0.1124 Time2 0.1118 0.03374 3110 3.32 0.0009 ** Time3 -0.0174 0.007258 3110 -2.4 0.0166 School Level Elementary -0.1259 0.01994 3110 -6.31 <.0001 ** School Level Middle 0 . . Free Reduced Lunch -0.1057 0.008973 3110 -11.78 <.0001 ** Minority 0.03066 0.009804 3110 3.13 0.0018 ** LEP -0.02294 0.007585 3110 -3.02 0.0025 ** Students with Disabilities -0.04931 0.006202 3110 -7.95 <.0001 ** Gifted 0.06242 0.006706 3110 9.31 <.0001 ** Positive Learning Environment 0.03538 0.008356 3110 4.23 <.0001 ** Positive Teacher Qualifications 0.02375 0.005408 3110 4.39 <.0001 ** Access Content Software 0.008497 0.005475 3110 1.55 0.1207 Access Office Software 0.000562 0.005219 3110 0.11 0.9143 Access Advanced Production Software -0.00391 0.005461 3110 -0.72 0.4744 Teachers Use to Deliver 0.001141 0.006011 3110 0.19 0.8494

PAGE 191

177 Effect School Level Estimate SE df t p Instruction Teachers Use for Administrative Purposes -0.00236 0.006299 3110 -0.37 0.7077 Frequency that Students Use Content Software -0.00972 0.005023 3110 -1.93 0.0531 Frequency Students Use Tool-Based Software 0.003688 0.005274 3110 0.7 0.4844 Technical Support Human -0.00595 0.004975 3110 -1.2 0.2317 Technical Support Hardware -0.00712 0.004676 3110 -1.52 0.128 Time*School Level Elementary 0.02292 0.05075 3110 0.45 0.6516 Time*School Level Middle 0 . . Time*Free Reduced Lunch -0.01032 0.02419 3110 -0.43 0.6697 Time*Minority -0.03125 0.02308 3110 -1.35 0.1758 Time*LEP 0.009833 0.01779 3110 0.55 0.5806 Time*Students with Disabilities 0.00877 0.01511 3110 0.58 0.5618 Time*Gifted -0.00676 0.01596 3110 -0.42 0.672 Time*Positive Learning Environment 0.004399 0.02391 3110 0.18 0.854 Time*Positive Teacher Qualifications -0.00789 0.0144 3110 -0.55 0.5837 Time*Access Content Software -0.00417 0.01691 3110 -0.25 0.8054 Time*Access Office Software 0.000143 0.01635 3110 0.01 0.993 Time*Access Advanced Production Software 0.001623 0.01645 3110 0.1 0.9214 Time*Teachers Use to Deliver Instruction 0.009942 0.01891 3110 0.53 0.5991 Time*Teachers Use for Administrative Purposes 0.00786 0.01947 3110 0.4 0.6864 Time*Frequency that Students Use Content Software -0.0018 0.01684 3110 -0.11 0.9148 Time*Frequency Students Use Tool-Based Software 0.005861 0.01729 3110 0.34 0.7346 Time*Technical Support Human 0.01917 0.01486 3110 1.29 0.197 Time*Technical Support Hardware 0.02102 0.01479 3110 1.42 0.1555 Time2*School Level Elementary 0.03295 0.04384 3110 0.75 0.4524 Time2*School Level Middle 0 . . Time2*Free Reduced Lunch 0.003323 0.02133 3110 0.16 0.8762 Time2*Minority 0.01923 0.02 3110 0.96 0.3363 Time2*LEP 0.007586 0.01544 3110 0.49 0.6232 Time2*Students with Disabilities -0.00932 0.01309 3110 -0.71 0.4764 Time2*Gifted 0.003252 0.0138 3110 0.24 0.8138 Time2*Positive Learning Environment -0.01381 0.02094 3110 -0.66 0.5097 Time2*Positive Teacher 0.000841 0.01262 3110 0.07 0.9469

PAGE 192

178 Effect School Level Estimate SE df t p Qualifications Time2*Access Content Software 0.000846 0.01462 3110 0.06 0.9539 Time2*Access Office Software 0.002699 0.01409 3110 0.19 0.8482 Time2*Access Advanced Production Software 0.00179 0.01425 3110 0.13 0.9 Time2*Teachers Use to Deliver Instruction -0.01033 0.01637 3110 -0.63 0.5282 Time2*Teachers Use for Administrative Purposes -0.00621 0.01661 3110 -0.37 0.7087 Time2*Frequency that Students Use Content Software 0.01002 0.01465 3110 0.68 0.4941 Time2*Frequency Students Use Tool-Based Software -0.01237 0.01519 3110 -0.81 0.4155 Time2*Technical Support Human -0.01336 0.01287 3110 -1.04 0.2994 Time2*Technical Support Hardware -0.00575 0.01275 3110 -0.45 0.6521 Time3*School Level Elementary -0.01891 0.009519 3110 -1.99 0.0471 Time3*School Level Middle 0 . . Time3*Free Reduced Lunch -0.00006 0.004686 3110 -0.01 0.9906 Time3*Minority -0.00348 0.004368 3110 -0.8 0.4263 Time3*LEP -0.00242 0.003365 3110 -0.72 0.4716 Time3*Students with Disabilities 0.002351 0.002851 3110 0.82 0.4096 Time3*Gifted -0.00073 0.003006 3110 -0.24 0.8093 Time3*Positive Learning Environment 0.00411 0.004572 3110 0.9 0.3687 Time3*Positive Teacher Qualifications 0.000586 0.002766 3110 0.21 0.8322 Time3*Access Content Software 0.00017 0.003205 3110 0.05 0.9577 Time3*Access Office Software -0.00079 0.003073 3110 -0.26 0.798 Time3*Access Advanced Production Software -0.00076 0.003116 3110 -0.24 0.808 Time3*Teachers Use to Deliver Instruction 0.002396 0.003582 3110 0.67 0.5036 Time3*Teachers Use for Administrative Purposes 0.001487 0.003603 3110 0.41 0.6798 Time3*Frequency that Students Use Content Software -0.00255 0.003202 3110 -0.79 0.4267 Time3*Frequency Students Use Tool-Based Software 0.002975 0.003354 3110 0.89 0.3751 Time3*Technical Support Human 0.002882 0.002816 3110 1.02 0.3063 Time3*Technical Support Hardware -0.0003 0.002783 3110 -0.11 0.9148

PAGE 193

179 Covariance Parameter Estimate SE z p 0.04139 0.001936 21.38 <.0001 ** -0.00519 0.000585 -8.87 <.0001 ** 0.003402 0.000275 12.36 <.0001 ** Residual 0.01885 0.000469 40.17 <.0001 ** Note: p < .05; ** p < .01 The last model estimated in order to answer the third hypothesis for predicting FCAT Writing achievement included all school levels, demographic, student learning environment, and significant technology integration variables. These models were different because the model fit to the data for all schools levels without gifted included one technology integration variable – frequency that students use content software (see model 7a). The final model fitte d to the data with elementary and middle school levels and gifted included no technology integration variables. For the model with all schools levels and without gifted, the same parameter estimates and interactions identified in the previous models as significant were significant again (s ee Table 37). Although there was no difference in the percentage of variance explained in this model than was in the Demographic Model with Student Learning Environment by school level or the Technology Integration with Demographic and Student Learning Environment Model by school level, the AIC, AICC, and BIC indices all indicated better model fit (see Table 39). The level-1 residuals for the final model for predicting FCAT Wr iting using all school levels without gifted ranged between -0.53 and 0.48 with a standard deviation of 0.11. Although there were outliers, skewness was -0.09 and kurtosis was 0.87, which would indicate that the residuals were evenly distributed. Distribution of the empirical bayes intercepts ranged between -0.58 and 0.70 with standard deviation of 0.18. Skewness was 0.38, and kurtosis was 0.56, which indicated that the intercept residuals at level-2 were also normally distributed. Distribution of the empirical bayes slopes ranged between -0.19 and 0.16 with standard deviation of 0.04. Skewness was 0.17, and kurtosis was 1.18, which indicated that the slope residuals at level-2 were within acceptable range. Final Model 7a: Significant Technology Integration Indicators with Demographics and Student Learning Environment by School Level (All School Levels without Gifted) Level 1: FCAT Writing = 0 + 1*Time + 2*Time2 + 3*Time3 + r

PAGE 194

180 Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Teacher Qualifications + 07* Positive Learning Environment + 08*Frequency Students Use Content + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16* Teacher Qualifications + 17* Positive Learning Environment + 18*Frequency Students Use Content + u1 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26* Teacher Qualifications + 27* Positive Learning Environment + 28*Frequency Students Use Content 3 = 30 + 31*School Level + 32*SES + 33*Minority + 34* LEP + 35* SWD + 36* Teacher Qualifications + 37* Positive Learning Environment + 38*Frequency Students Use Content Mixed-Effects Model: FCAT Writing = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06* Teacher Qualifications + 07* Positive Learning Environment + 08*Frequency Students Use Content + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Teacher Qualifications*Time + 17* Positive Learning Environment*Time + 18*Frequency Students Use Content*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Teacher Qualifications*Time2 + 27* Positive Learning Environment*Time2 + 28*Frequency Students Use Content*Time2 + 30*Time3 + 31*School Level*Time3 + 32*SES*Time3 + 33*Minority*Time3 + 34* LEP*Time3 + 35* SWD*Time3 + 36*Teacher Qualifications*Time3 + 37* Positive Learning Environment*Time3 + 38*Frequency Students Use Content*Time3 + 310*Teachers use Admin*Time3 + u0 + u1 + r Table 37. Final Model 7a: Significant Technology Integration with Demographics and Student Learning Environment by School Level (A ll School Levels without Gifted) Effect School Level Estimate SE df t p Intercept 3.7984 0.01416 2219 268.16 <.0001 ** Time -0.04304 0.03294 2198 -1.31 0.1915 Time2 0.1071 0.0282 4198 3.8 0.0001 ** Time3 -0.0178 0.006098 4198 -2.92 0.0035 ** School Level Elementary -0.1481 0.01707 4198 -8.67 <.0001 ** School Level High -0.06466 0.0183 4198 -3.53 0.0004 ** School Level Middle 0 . . Free Reduced Lunch -0.1201 0.007399 4198 -16.23 <.0001 ** Minority 0.03788 0.008142 4198 4.65 <.0001 ** LEP -0.03106 0.006858 4198 -4.53 <.0001 ** Students with Disabilities -0.06171 0.005483 4198 -11.26 <.0001 ** Positive Learning Environment 0.03906 0.006364 4198 6.14 <.0001 ** Positive Teacher Qualifications 0.03128 0.004652 4198 6.72 <.0001 ** Frequency that Students Use Content Software -0.01147 0.004053 4198 -2.83 0.0047 ** Time*School Level Elementary -0.03231 0.04084 4198 -0.79 0.429 Time*School Level High 0.1003 0.04014 4198 2.5 0.0125 Time*School Level Middle 0 . . Time*Free Reduced Lunch -0.00986 0.01949 4198 -0.51 0.6128

PAGE 195

181 Effect School Level Estimate SE df t p Time*Minority -0.02561 0.01874 4198 -1.37 0.1718 Time*LEP 0.006756 0.01541 4198 0.44 0.6611 Time*Students with Disabilities 0.003128 0.01314 4198 0.24 0.8118 Time*Positive Learning Environment 0.02404 0.01798 4198 1.34 0.1813 Time*Positive Teacher Qualifications 0.004747 0.01232 4198 0.39 0.7 Time*Frequency that Students Use Content Software 0.0049 0.01276 4198 0.38 0.701 Time2*School Level Elementary 0.06313 0.03538 4198 1.78 0.0744 Time2*School Level High -0.09909 0.03509 4198 -2.82 0.0048 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch 0.008596 0.0172 4198 0.5 0.6173 Time2*Minority 0.01034 0.0163 4198 0.63 0.5259 Time2*LEP 0.01119 0.01338 4198 0.84 0.4029 Time2*Students with Disabilities -0.00447 0.0114 4198 -0.39 0.6948 Time2*Positive Learning Environment -0.0201 0.01595 4198 -1.26 0.2077 Time2*Positive Teacher Qualifications -0.00913 0.0107 4198 -0.85 0.3935 Time2*Frequency that Students Use Content Software 0.002287 0.01102 4198 0.21 0.8355 Time3*School Level Elementary -0.02303 0.007703 4198 -2.99 0.0028 ** Time3*School Level High 0.0152 0.007694 4198 1.98 0.0483 Time3*School Level Middle 0 . . Time3*Free Reduced Lunch -0.0019 0.003783 4198 -0.5 0.6164 Time3*Minority -0.00123 0.003571 4198 -0.34 0.7315 Time3*LEP -0.00314 0.002923 4198 -1.08 0.282 Time3*Students with Disabilities 0.001281 0.00249 4198 0.51 0.6069 Time3*Positive Learning Environment 0.004171 0.003511 4198 1.19 0.235 Time3*Positive Teacher Qualifications 0.002561 0.002331 4198 1.1 0.272 Time3*Frequency that Students Use Content Software -0.00074 0.002405 4198 -0.31 0.7598 Covariance Parameter Estimate SE z p 0.03993 0.001658 24.09 <.0001 ** -0.00441 0.000481 -9.15 <.0001 ** 0.003048 0.000223 13.67 <.0001 ** Residual 0.01829 0.000396 46.13 <.0001 ** Note: p < .05; ** p < .01 The last step was to add in USDOE funded Magnet Schools and USDOE Technology Magnet Schools as variables in the model. Results of this m odel indicated that neither the interaction between time

PAGE 196

182 nor time2 with U.S. technology magnet school was a significant predictor of FCAT Writing with the data at all school levels without gifted; how ever, the interaction between time and time2 with U.S. magnet school status was a significant predictor of FCAT Writing (see Table 38). Neither U.S. magnet school status nor U.S. technology magnet school status were significant with the data with elementary and middle schools and gifted. Table 38. Model 8a: Magnet Schools with Significant Technology Integration Demographics and Student Learning Environment by School Level (A ll School Levels without Gifted) Effect School Level Estimate SE df t p Intercept 3.7697 0.0689 2217 54.71 <.0001 ** Time -0.07478 0.1494 2196 -0.5 0.6166 Time2 0.06203 0.1299 4194 0.48 0.6329 Time3 0.001923 0.02848 4194 0.07 0.9462 School Level Elementary -0.1491 0.01709 4194 -8.73 <.0001 ** School Level High -0.06492 0.0183 4194 -3.55 0.0004 ** School Level Middle 0 . . Free Reduced Lunch -0.1197 0.0074 4194 -16.18 <.0001 ** Minority 0.03851 0.00818 4194 4.71 <.0001 ** LEP -0.03138 0.006869 4194 -4.57 <.0001 ** Students with Disabilities -0.0617 0.00549 4194 -11.24 <.0001 ** Positive Learning Environment 0.03928 0.006365 4194 6.17 <.0001 ** Positive Teacher Qualifications 0.03131 0.00465 4194 6.73 <.0001 ** Frequency that Students Use Content Software -0.0117 0.004055 4194 -2.89 0.0039 ** Not a Technology Magnet School US -0.00367 0.07864 4194 -0.05 0.9627 Technology Magnet School US 0 . . Not a US Magnet School 0.03393 0.04088 4194 0.83 0.4066 US Magnet School 0 . . Time*School Level Elementary -0.02877 0.04083 4194 -0.7 0.4811 Time*School Level High 0.1026 0.0401 4194 2.56 0.0105 Time*School Level Middle 0 . . Time*Free Reduced Lunch -0.01147 0.01949 4194 -0.59 0.5563 Time*Minority -0.02806 0.01877 4194 -1.49 0.1351 Time*LEP 0.008347 0.01542 4194 0.54 0.5884 Time*Students with Disabilities 0.004177 0.01315 4194 0.32 0.7507 Time*Positive 0.02444 0.01797 4194 1.36 0.1739

PAGE 197

183 Effect School Level Estimate SE df t p Learning Environment Time*Positive Teacher Qualifications 0.004482 0.0123 4194 0.36 0.7157 Time*Frequency that Students Use Content Software 0.00452 0.01275 4194 0.35 0.723 Time*Not a Technology Magnet School US 0.2215 0.1702 4194 1.3 0.1932 Time*Technology Magnet School US 0 . . Time*Not a US Magnet School -0.1955 0.08854 4194 -2.21 0.0273 Time*US Magnet School 0 . . Time2*School Level Elementary 0.06002 0.03536 4194 1.7 0.0897 Time2*School Level High -0.1012 0.03506 4194 -2.89 0.0039 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch 0.01012 0.0172 4194 0.59 0.5563 Time2*Minority 0.01261 0.01633 4194 0.77 0.4399 Time2*LEP 0.009714 0.01339 4194 0.73 0.4681 Time2*Students with Disabilities -0.00503 0.01141 4194 -0.44 0.6594 Time2*Positive Learning Environment -0.02067 0.01595 4194 -1.3 0.1949 Time2*Positive Teacher Qualifications -0.00874 0.01069 4194 -0.82 0.4134 Time2*Frequency that Students Use Content Software 0.002916 0.01101 4194 0.26 0.7911 Time2*Not a Technology Magnet School US -0.112 0.1484 4194 -0.75 0.4505 Time2*Technology Magnet School US 0 . . Time*Not a US Magnet School 0.1623 0.07758 4194 2.09 0.0365 Time*US Magnet School 0 . . Time3*School Level Elementary -0.02234 0.007699 4194 -2.9 0.0037 ** Time3*School Level High 0.01563 0.007686 4194 2.03 0.042 Time3*School Level Middle 0 . . Time3*Free Reduced Lunch -0.00224 0.003782 4194 -0.59 0.5541 Time3*Minority -0.00172 0.003575 4194 -0.48 0.63 Time3*LEP -0.00282 0.002924 4194 -0.96 0.3348

PAGE 198

184 Effect School Level Estimate SE df t p Time3*Students with Disabilities 0.001341 0.002492 4194 0.54 0.5905 Time3*Positive Learning Environment 0.004286 0.00351 4194 1.22 0.2222 Time3*Positive Teacher Qualifications 0.00246 0.002328 4194 1.06 0.2909 Time3*Frequency that Students Use Content Software -0.0009 0.002403 4194 -0.37 0.7078 Time3*Not a Technology Magnet School US 0.01144 0.03259 4194 0.35 0.7257 Time3*Technology Magnet School US 0 . . Time3*Not a US Magnet School -0.03231 0.01706 4194 -1.89 0.0582 Time3*US Magnet School 0 . . Covariance Parameter Estimate SE z p 0.03995 0.001657 24.11 <.0001 ** -0.00442 0.000481 -9.18 <.0001 ** 0.003055 0.000223 13.71 <.0001 ** Residual 0.01824 0.000395 46.13 <.0001 ** Note: p < .05; ** p < .01 Table 39. Model Fit Indices for Models Predicting FCAT Writin g Scores for All School Levels (without Gifted and LEP) Model -2 Log Likelihood AIC (smaller is better) AICC (smaller is better) BIC (smaller is better) Model 1: Writing Predicted by Average Writing of All Schools in Florida 346.8 352.8 352.8 370 Model 2a: Time as a Predictor of Writing -2168 -2156 -2156 -2121.7 Quadratic Model 2b: Time2 as a Predictor of Writing -2173.7 -2159.7 -2159.6 -2119.6 Polynomial Model 2c: Time3 as a Predictor of Writing -2341.7 -2325.7 -2325.7 -2279.9 Model 3: Time, Time2, Time3, and School Level as Predictors of Writing -2847.8 -2815.8 -2815.8 -2724.2 Model 4a: Writing predicted by Time, School Level, and Demographics Variables -3939.6 -3875.6 -3875.3 -3693

PAGE 199

185 Model -2 Log Likelihood AIC (smaller is better) AICC (smaller is better) BIC (smaller is better) Model 5a: Demographics and Student Learning Environment by School Level -4121.8 -4041.8 -4041.4 -3813.5 Model 6a: Technology Integration with Demographics and Student Learning Environment by School Level -4176.2 -4024.2 -4022.8 -3590.5 Final Model 7a: Significant Technology Integration with Demographics and Student Learning Environment by School Level -4132.5 -4044.5 -4044 -3793.4 Model 8a: Magnet Schools with Significant Technology Integration Demographics and Student Learning Environment by School Level (All School Levels without Gifted) -4146 -4042 -4041.3 -3745.3 Table 40. Model Fit Indices for Models Predicting FCAT Writing Scores for Elementary and Middle School Levels (with Gifted) Model -2 Log Likelihood AIC (smaller is better) AICC (smaller is better) BIC (smaller is better) Model 1: Writing Predicted by Average Writing of All Elementary and Middle Schools in Florida -261.6 -251.6 -251.6 -223.8 Model 4b: Writing predicted by Time, School Level, and Demographics Variables No High School includes gifted -2815.5 -2751.5 -2751.1 -2575.7 Model 5b: Demographics and Teacher Qualifications by School Level -2885.7 -2805.7 -2805.2 -2586 Model 6b: Technology Integration with Demographics and Teacher Qualifications by School Level -2933.4 -2781.4 -2779.7 -2364 Final Model 7b: Significant Technology Integration with Demographics and Student Learning Environment by School Level -2898.5 -2802.5 -2801.8 -2538.9 The result of the analysis for all the models indicated that Hypothesis 3 was partially correct. When the sample included schools at all three school levels, there was a significant negative relationship between the frequency that students use content softwa re and the intercept of school level FCAT Writing achievement when all other school level, demographic, and school learning environment factors were controlled. These interactions resulted in an s-shaped curvilinear trend.

PAGE 200

186 After controlling so that all other variables were held at the mean, the trend for each school level could be examined separately, by comparing schools with different levels that students use content software. Figure 23 illustrates the relationship between the frequency that students use content software and average school FCAT Writing score fo r high schools. The frequency that students use content software was compared at one and two standard deviations be low the mean, the mean, and one and two standard deviations above the mean. This allows the extreme cas es of schools that have the frequency that students use content software the most often, +2 standard deviations above the mean, to be compared with schools that have students who use content software the least often, -2 standard deviations below the mean. Schools that had the 2 standard deviations above the mean in frequency that students use content software started the study in 2003-04 with the lowest FCAT Writing scores (3.71) and schools that had 2 standard deviations below the mean in frequency that students use content software started with the highest FCAT Writing scores (3.76). This difference of 0.05 point was significant because there were so many schools in the sample; however, the practical importance was modest. The differences between these extremes narrowed over time. In 2005-06 all levels of frequency that students used content software had the same mean FCAT Writing score. In 2006-07 schools with 2 st andard deviations above the mean in frequency that students used content software had the highest scores and schools with 2 standard deviations below the mean in frequency that students used content software had the lowest scores.

PAGE 201

187 Relationship between Frequency Students Use Content Software and FCAT Writing in High Schools3.50 3.60 3.70 3.80 3.90 4.00 4.10 4.20 4.30 4.40 4.50 2003200420052006 YearFCAT Writing High Schools + 2SD High Schools + 1SD High Schools mean High Schools 1SD High Schools 2SD Figure 23. Relationship between Frequency that students use content software and FCAT Writing in high schools. Middle schools had a similar beginning pattern to high school, that is, after controlling for all other factors, schools that were two standard deviations above the mean in the frequency that students use content software had the lowest FC AT Writing scores in 2003-04 (3.78), while those with two standard deviations below the mean had the highest scores (3.82). Although this difference of 0.04 point was significant due to the large sample size, the practical importance is modest (see Figure 24). In 2005-06 the mean FCAT Writing score for schools at all levels of frequency that students use content software was the same (4.00). In 2006-07 schools with 1 and 2 standard deviations above the mean had the highest score (4.16), while schools that were 2 standard deviations below the mean had the lowest score (4.14). The slope of these changes was not significant.

PAGE 202

188 Relationship between Frequency Students Use Content Software and FCAT Writing in Middle Schools3.50 3.60 3.70 3.80 3.90 4.00 4.10 4.20 4.30 4.40 4.50 2003200420052006 YearFCAT Writing Middle Schools + 2SD Middle Schools + 1SD Middle Schools mean Middle Schools 1SD Middle Schools 2SD Figure 24. Relationship between Frequency that Students Use Content Software and FCAT Writing in Middle Schools. Elementary schools experienced a similar pattern to middle schools (see Figure 25). Schools with the lowest level of frequency that students use content software began the study with the highest FCAT Writing score (3.67), while schools with the highest le vel of frequency that students use content software had the lowest FCAT Writing score (3.63). Although this difference of 0.04 point was statistically significant, it had modest practical importance. In 2005-06 and 2006-07 schools with one and two standard deviations above the mean had the highest score (3.86) while school at the mean and with one and two standard deviations below the mean had the lowest score (3.85). However the slope of the interaction between time and the frequency that students use content software was not significant.

PAGE 203

189 Relationship between Frequency Students Use Content Software and FCAT Writing in Elementary Schools3.50 3.60 3.70 3.80 3.90 4.00 4.10 4.20 4.30 4.40 4.50 2003200420052006 YearFCAT Writing Elementary + 2SD Elementary + 1SD Elementary mean Elementary 1SD Elementary 2SD Figure 25. Relationship between Frequency that students use content software and FCAT Writing in elementary schools. When the sample was restricted to just elementary and middle schools and percent of gifted students was included in the equation, both the intercept of gifted and the interactions of percent of gifted students in the school with time, time2, and time3 were significant. Thus, when all other factors were held equal, schools with highest percentages of gifted students began the study with the highest FCAT Writing scores, and the trends were not linear (see Figure 26). In addition, the trends were different at elementary and middle school level. There were no significant t echnology integration indicators with this dataset.

PAGE 204

190 Relationship between Percent of Gifted Students on FCAT Writing by School Level (Gifted Included) 3.50 3.60 3.70 3.80 3.90 4.00 4.10 4.20 4.30 4.40 4.50 2003200420052006 YearFCAT Writing Elementary + 2SD Elementary + 1SD Elementary mean Elementary 1SD Elementary 2SD Middle Schools + 2SD Middle Schools + 1SD Middle Schools mean Middle Schools 1SD Middle Schools 2SD Figure 26. Relationship between Percent of Gifted Students on FCAT Writing by School Level (Gifted Included). Research Question 2 What is the relationship between indicators of technology integration and changes in mediating outcomes of absence rate and student misconduct, when controlling for school level, school socioeconomic status, minority, limited E nglish proficiency, students with disabilities, gifted, teacher qualification, and learning environment quality? The second research question was answered by conducting multi-level models with the student misconduct and absence rates. Absence rate was measured by the percent of students who were absent more than 21 days per year. Miscon duct was measured with a composite variable created from the sum of the mean percent of students in in-school suspensions, mean percent of students in out-of-school suspensions, and the mean number of crime incidents per student. Hypothesis 1 The first analysis conducted to answer the seco nd research question used the student absences outcome data to test the following hypothesis:

PAGE 205

191 H1: After controlling for school level (elementary, middle, and high), school socio-economic status, minority, limited English proficiency, students with disabilities, gifted, teacher qualification, and learning environment quality, mean school percent of students with more than 21 days absent variable will have a negative relationship with indicators of technology integration. The first step was to build the unconditional model. The unconditional model predicted the schools’ percent of students absent more than 21 days from the average of the percent of students absent more than 21 days for all schools. There were no ot her predictors. The average percent of students absent more than 21 days for all schools was 9.00 ( t (2311) = 76.03, p <.0001). Model 1: Unconditional Model Level 1: Student Absences = 0 + r Level 2: 0 = 00 + u0 Mixed-Effects Model: Student Absences = 00 + u0 + r The intraclass correlation coeffici ent (ICC) was computed to determine the proportion of variance in the percent of students absent more than 21 days variable that is accounted for by the schools. The ICC was .76, which is high and supports using multi-leve l modeling for the analysis. The model fit statistics from this model were used as the baseline for model comparisons. The next step added time to the predictor equation (see Model 2a). The variance components from this analysis showed how much of the variance in the model was accounted for by time. The variance in the slopes between schools was significant. Therefore, tim e was set as a random effect, and the model was estimated. Both the intercept ( t (2311) = 74.16, p <.0001) and time ( t (2311) = 11.29, p <.0001) were significant parameters. However, time explained 18% of the variance between schools and accounted for 18% of the variance within schools. Model 2a: Unconditional Growth Model Level 1: Student Absences = 0 + 1*Time + r Level 2: 0 = 00 + u0 1 = 10 + u1 Mixed-Effects Model: Student Absences = 00 + 10*Time + u0 + u1*Time + r To determine if the equation was not linear but curvilinear, time2 was added to the equation so the variance could be compared. Results indicated that time2 was significant ( t (2311) = -6.56, p <.0001), and it increased the within school variance explained by 2% over the Growth Model (see Model 2b). Because the trends included on three points in time, time3 was not added to the equation. Consequently, time and time2 were retained in the quadratic growth model equation.

PAGE 206

192 Model 2b: Quadratic Growth Model Level 1: Student Absences = 0 + 1*Time + 2* Time2 + r Level 2: 0 = 00 + u0 1 = 10 + u1 2 = 20 Mixed-Effects Model: Student Absences = 00 + 10*Time + 20* Time2 + u0 + u1 + r Next, school level was added to the Quadratic Growth Model to predict misconduct (See Model 3). The significance of the parameter estimates determined if school level was significantly related to the percent of students absent more than 21 days and if th ere was an interaction with time. This model adjusted the mean school percent of students absent more than 21 days and the slope of percent of students absent more than 21 days for school level. The parameter estimates of elementary school and high school relative to middle school and time and time2 were significant at the intercept. All of the interactions between time and time2 with elementary and high school relative to mi ddle school were significant (see Table 41). All model fit indices indicated improved fit with this model (see Table 52). This model accounted for an additional 27% of the between school variance and an additional 1% of the within school variance from the Quadratic Growth Model. Model 3: School Level as Predictor Level 1: Student Absences = 0 + 1*Time + 2*Time2 + r Level 2: 0 = 00 + 01*School Level + u0 1 = 10 + 11*School Type + u1 2 = 20 + 21*School Type Mixed-Effects Model: Student Absences = 00 + 01*School Type + 10*Time + 11*School Type*Time + 20*Time2 + 21*School Type*Time2 + u0 + u1 + r Table 41. Model 3: Time, Time Squared, and School Type as Predictors of Student Absences Effect School Level Estimate SE df t p Intercept 10.8563 0.2291 2309 47.39 <.0001 ** Time 0.8426 0.3286 2309 2.56 0.0104 Time2 -0.3558 0.1554 2309 -2.29 0.0221 School Level Elementary -4.571 0.2606 2309 -17.54 <.0001 ** School Level High 3.035 0.3457 2309 8.78 <.0001 ** School Level Middle 0 . . Time*School Level Elementary 1.144 0.3737 2309 3.06 0.0022 ** Time*School Level High -1.0712 0.4959 2309 -2.16 0.0309 Time*School Level Middle 0 . . Time2*School Level Elementary -0.3293 0.1768 2309 -1.86 0.0626 Time2*School Level High 0.8016 0.2346 2309 3.42 0.0006 **

PAGE 207

193 Effect School Level Estimate SE df t p Time2*School Level Middle 0 . . Covariance Parameter Estimate SE z p 16.2216 0.6765 23.98 <.0001 ** 2.3649 0.2456 9.63 <.0001 ** 1.4602 0.1823 8.01 <.0001 ** Residual 7.1823 0.2112 34 <.0001 ** Note: p < .05; ** p < .01 The next model added student demographic variables to the School Level Model. This model was run twice. The first time, the model was estimated w ith elementary, middle, and high school as school levels and all of the demographic variables except gift ed, because gifted is not a designation at the high school level (see Model 4a). The second time, the data were filtered to exclude high school as a school level and keep the gifted variable with middle and elementary schools (see Model 4b). The model fit statistics of the demographic model with all three sch ool levels was compared with the School Level as Predictor Model to determine if there was a better fit (see Table 52). The significance of the parameter estimates determined which of the demographic variables remained in the predictor equation (see Table 42). The variance estimates showed the amount of the total variance that was accounted for by each model. When all of the demographic variables except gifted were added to the model (see Model 4a), the intercept was significant and the average middl e school started with 11.18 ( t (2259) = 53.38, p <.0001) percent of students absent more than 21 days. The parameter estima tes for elementary and high school level relative to middle, free or reduced lunch status, minority, limited English proficiency (LEP), and students with disabilities were significant, while the parameter estimate for time and time2 were not significant. There were significant interactions between time and elemen tary relative to middle school, free or reduced lunch status, and LEP, while interactions between time and minority and students with disabilities were not significant. Time2 had significant interactions with elementary and high schools relative to middle school. No other demographic variables had significant interactions with time2. All model fit indices indicated better fit with the addition of these demographics variables (see Table 52). Adding the demographics variables with school level explained 64% of the between school variance and 18% of the within school variance for a total of 53% of all variance explained.

PAGE 208

194 Model 4a: Demographics by School Level (including High School and no Gifted) Level 1: Student Absences = 0 + 1*Time + 2*Time2 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + u0 1 = 10 + 11*School Type + 12*SES + 13*Minority + 14* LEP + 15* SWD + u1 2 = 20 + 21*School Type + 22*SES + 23*Minority + 24* LEP + 25* SWD Mixed-Effects Model: Student Absences = 00 + 01*School Level + 02*SES + 03*Minority + 04*SWD + 05*LEP + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14*SWD*Time + 15*LEP*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24*SWD*Time2 + 25*LEP*Time2 + u0 + u1 + r Table 42. Model 4a: Student Absences Predicted by Time, Sc hool Type, and Demographics Variables (No Gifted) Effect School Level Estimate SE df t p Intercept 11.1805 0.2094 2259 53.38 <.0001 ** Time 0.5675 0.3473 2203 1.63 0.1024 Time2 -0.2634 0.1646 2092 -1.6 0.1097 School Level Elementary -5.125 0.2409 2092 -21.27 <.0001 ** School Level High 4.241 0.3157 2092 13.43 <.0001 ** School Level Middle 0 . . Free Reduced Lunch 1.1772 0.1234 2092 9.54 <.0001 ** Minority 0.8634 0.1365 2092 6.32 <.0001 ** LEP -0.2918 0.1173 2092 -2.49 0.013 Students with Disabilities 1.1264 0.09194 2092 12.25 <.0001 ** Time*School Level Elementary 1.5591 0.4011 2092 3.89 0.0001 ** Time*School Level High -0.7698 0.5253 2092 -1.47 0.1429 Time*School Level Middle 0 . . Time*Free Reduced Lunch 0.645 0.2358 2092 2.74 0.0063 ** Time*Minority -0.1982 0.2428 2092 -0.82 0.4143 Time*LEP -0.4186 0.2024 2092 -2.07 0.0387 Time*Students with Disabilities -0.04916 0.1712 2092 -0.29 0.7741 Time2*School Level Elementary -0.5769 0.1903 2092 -3.03 0.0025 ** Time2*School Level High 0.6845 0.2492 2092 2.75 0.0061 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.1385 0.1123 2092 -1.23 0.2176 Time2*Minority -0.07399 0.115 2092 -0.64 0.5202 Time2*LEP 0.1246 0.09537 2092 1.31 0.1916 Time2*Students with Disabilities 0.09235 0.08101 2092 1.14 0.2544 Covariance Parameter Estimate SE z p 10.5658 0.5508 19.18 <.0001 ** 2.0808 0.2254 9.23 <.0001 ** 0.954 0.1841 5.18 <.0001 ** Residual 7.4587 0.2288 32.61 <.0001 ** Note: p < .05; ** p < .01

PAGE 209

195 The results from the analysis in Model 4b indicated that the intercept, elementary relative to middle school, time, time2, free or reduced lunch status, minority, LE P, students with disabilities, and gifted were all significant (see Table 43). Interactions betw een time and elementary rela tive to middle school, free or reduced lunch status, minority, LEP, students with disa bilities, and gifted were significant. Interactions between time2 and elementary relative to middle school and free and reduced lunch status were significant. Because the parameter for gifted was significant in this model, an unconditional model using the same population with high schools filtered out, predicting stud ent absences with average student absences for all schools was estimated in order to compare the fit of this model. All of the model fit statistics indicated better model fit (see Table 53). When examining the va riance of student absences in elementary and middle schools, adding demographics variables to the equation explained 78% of the between school variance and 22% of the within school variance. Two sets of analyses were conducted on the rest of the models in order to examine the relationship of gifted with technology integration as one of the predictors of student absences. Model 4b: Demographics by School Level (Elementary and Middle School only) Level 1: Student Absences = 0 + 1*Time + 2*Time2 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + u0 1 = 10 + 11*School Type + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16*Gifted + u1 2 = 20 + 21*School Type + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26*Gifted Mixed-Effects Model: Student Absences = 00 + 01*School Level + 02*SES + 03*Minority + 04*SWD + 05*LEP + 06*Gifted + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP*Time + 15* SWD*Time + 16*Gifted*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Gifted*Time2 + u0 + u1 + r Table 43. Model 4b: Student Absences predicted by Time, School Level, and Demographics Variables for Elementary and Middle Schools with Gifted Effect School Level Estimate SE df t p Intercept 11.1249 0.1643 1825 67.73 <.0001 ** Time 0.7269 0.3297 1704 2.2 0.0276 Time2 -0.3203 0.1563 1537 -2.05 0.0406 School Level Elementary -5.1788 0.1909 1537 -27.13 <.0001 ** School Level Middle 0 . .

PAGE 210

196 Effect School Level Estimate SE df t p Free Reduced Lunch 1.2449 0.1154 1537 10.79 <.0001 ** Minority 0.4983 0.1251 1537 3.98 <.0001 ** LEP -0.3201 0.09813 1537 -3.26 0.0011 ** Students with Disabilities 0.8943 0.0809 1537 11.05 <.0001 ** Gifted -0.4848 0.08652 1537 -5.6 <.0001 ** Time*School Level Elementary 1.4777 0.3853 1537 3.83 0.0001 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch 0.9815 0.2601 1537 3.77 0.0002 ** Time*Minority -0.6107 0.2661 1537 -2.29 0.0219 Time*LEP -0.5341 0.2064 1537 -2.59 0.0098 ** Time*Students with Disabilities 0.0817 0.1776 1537 0.46 0.6456 Time*Gifted -0.00272 0.1858 1537 -0.01 0.9883 Time2*School Level Elementary -0.5112 0.1828 1537 -2.8 0.0052 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.2825 0.1237 1537 -2.28 0.0225 Time2*Minority 0.1419 0.1262 1537 1.12 0.2611 Time2*LEP 0.1705 0.09744 1537 1.75 0.0803 Time2*Students with Disabilities 0.06487 0.0841 1537 0.77 0.4406 Time2*Gifted 0.04103 0.08722 1537 0.47 0.6382 Covariance Parameter Estimate SE z p 4.2996 0.3812 11.28 <.0001 ** 2.532 0.1913 13.24 <.0001 ** 1.119 0.1856 6.03 <.0001 ** Residual 6.2437 0.2201 28.37 <.0001 ** Note: p < .05; ** p < .01 The next model added the variables that measure the school learning environment factors to the Demographics Model by School Level Model. These in cluded teacher qualifica tions and positive learning environment. The composite variable for positive learning environment used in all the analyses of the other outcomes included the variable percent of students abse nt less than 21 days. In order to prevent collinearity, the composite variable for positive learning environmen t was recalculated without the variable for student absences before the model was estimated. This model was estimated twice, first without gifted population but included all school levels (see model 5a) and then with elementary and middle school levels and gifted population (see model 5b). When school learning environment factors were added with the demographic and school level variables for all school levels, the parameter estimates for the intercept, elementary and high school relative to middle school, free or reduced lunch status, minority, students with disabilities, and

PAGE 211

197 positive learning environment were significant, while time, time2, LEP, and teacher qualifications were not significant (see Table 44). Signif icant interactions with time and time2 included elementary schools relative to middle schools, free and reduced lunch status, and LE P, and positive learning environment. Adding the student learning environment variables explained an additional 2% of the between school variance and explained an additional 1% of the within school varian ce for a total of 55% of all of the variance explained. All of the model fit indices indicated that th is model fit the data better (see Table 52). Model 5a: School Learning Environment with Demographics by School Level (all school levels without gifted and LEP) Level 1: Student Absences = 0 + 1*Time + 2*Time2 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04*SWD + 05* Teacher Qualifications + 06*Positive Learning Environment + u0 1 = 10 + 11*School Type + 12*SES + 13*Minority + 14*LEP + 15*SWD + 16* Teacher Qualifications + 17*Positive Learning Environment + u1 2 = 20 + 21*School Type + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26* Teacher Qualifications + 27*Positive Learning Environment Mixed-Effects Model: Student Absences = 00 + 01*School Level + 02*SES + 03*Minority + 04* SWD + 05*Teacher Qualifications + 06* Positive Learning Environment + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14*SWD *Time + 15* Teacher Qualifications*Time + 16* Positive Learning Environment*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* SWD*Time2 + 25* Teacher Qualifications*Time2 + 26* Positive Learning Environment*Time2 + u0 + u1 + r Table 44. Model 5a: Absences Predicted by Demographics and Student Learning Environment by School Level (All School Levels without Gifted) Effect School Level Estimate SE df t p Intercept 9.9297 0.2366 2259 41.97 <.0001 ** Time 0.5105 0.4363 2203 1.17 0.2422 Time2 -0.1434 0.2061 2086 -0.7 0.4865 School Level Elementary -3.452 0.2837 2086 -12.17 <.0001 ** School Level High 3.9992 0.3116 2086 12.83 <.0001 ** School Level Middle 0 . . Free Reduced Lunch 1.0116 0.1253 2086 8.07 <.0001 ** Minority 0.5305 0.138 2086 3.84 0.0001 ** LEP -0.1373 0.1163 2086 -1.18 0.2379 Students with Disabilities 0.9648 0.09183 2086 10.51 <.0001 ** Positive Learning Environment -1.1039 0.1035 2086 -10.66 <.0001 ** Positive Teacher Qualifications 0.08278 0.07872 2086 1.05 0.2931 Time*School Level Elementary 1.5174 0.5355 2086 2.83 0.0046 ** Time*School Level High -0.6382 0.5278 2086 -1.21 0.2267 Time*School Level Middle 0 . .

PAGE 212

198 Effect School Level Estimate SE df t p Time*Free Reduced Lunch 0.506 0.2499 2086 2.02 0.043 Time*Minority -0.08575 0.2482 2086 -0.35 0.7298 Time*LEP -0.4033 0.2049 2086 -1.97 0.0492 Time*Students with Disabilities -0.05828 0.1732 2086 -0.34 0.7365 Time*Positive Learning Environment 0.142 0.2296 2086 0.62 0.5365 Time*Positive Teacher Qualifications -0.4028 0.1641 2086 -2.46 0.0142 Time2*School Level Elementary -0.6285 0.2548 2086 -2.47 0.0137 Time2*School Level High 0.6511 0.2502 2086 2.6 0.0093 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.07063 0.1197 2086 -0.59 0.5553 Time2*Minority -0.1168 0.1175 2086 -0.99 0.3202 Time2*LEP 0.1246 0.09677 2086 1.29 0.1981 Time2*Students with Disabilities 0.1098 0.08186 2086 1.34 0.1801 Time2*Positive Learning Environment -0.0519 0.1112 2086 -0.47 0.6407 Time2*Positive Teacher Qualifications 0.1795 0.07802 2086 2.3 0.0215 Covariance Parameter Estimate SE z p 9.9418 0.5279 18.83 <.0001 ** 2.0392 0.2204 9.25 <.0001 ** 0.9566 0.1821 5.25 <.0001 ** Residual 7.3519 0.2252 32.64 <.0001 ** Note: p < .05; ** p < .01 When the data were filtered to include only elementary and middle schools and gifted was also added to the equation, parameter estimates for the inte rcept, elementary relative to middle school, free or reduced lunch status, students with disabilities, gifted and positive learning environment were significant, while time, time2, minority, LEP, and positive teacher qualifica tions were not signif icant. Significant interactions with time included elementary relative to middle school, free or reduced lunch status, minority, and LEP. Significant interactions with time2 only included elementary relative to middle school (see Table 45). This model demonstrated better fit than the previo us model by all model fit indices (see Table 53). It explained 2% more of the between school variance and 1% more of the within school variance than the previous model for a total 63% of all the variance.

PAGE 213

199 Model 5b: School Learning Environment with Demographics by School Level (Elementary and Middle Schools with Gifted) Level 1: Student Absences = 0 + 1*Time + 2*Time2 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + u0 1 = 10 + 11*School Type + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16*Gifted + 17* Teacher Qualifications + 18* Positive Learning Environment + u1 2 = 20 + 21*School Type + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26*Gifted + 27* Teacher Qualifications + 28* Positive Learning Environment Mixed-Effects Model: Student Absences = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Gifted*Time + 17* Teacher Qualifications*Time + 18* Positive Learning Environment*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Gifted*Time2 + 27* Teacher Qualifications*Time2 + 28* Positive Learning Environment*Time2 + u0 + u1 + r Table 45. Model 5b: Absences Predicted by Demographics and Student Learning Environment by School Level for Elementary and Middle School with Gifted Effect School Level Estimate SE df t p Intercept 10.018 0.1995 1825 50.22 <.0001 ** Time 0.3751 0.4472 1704 0.84 0.4017 Time2 -0.03742 0.211 1531 -0.18 0.8592 School Level Elementary -3.686 0.2455 1531 -15.01 <.0001 ** School Level Middle 0 . . Free Reduced Lunch 1.1083 0.1156 1531 9.59 <.0001 ** Minority 0.2226 0.1266 1531 1.76 0.079 LEP -0.1688 0.09751 1531 -1.73 0.0836 Students with Disabilities 0.8061 0.07999 1531 10.08 <.0001 ** Gifted -0.3716 0.08643 1531 -4.3 <.0001 ** Positive Learning Environment -0.9781 0.1051 1531 -9.31 <.0001 ** Positive Teacher Qualifications -0.02594 0.07138 1531 -0.36 0.7163 Time*School Level Elementary 1.8769 0.5614 1531 3.34 0.0008 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch 0.7751 0.2713 1531 2.86 0.0043 ** Time*Minority -0.5854 0.2722 1531 -2.15 0.0317 Time*LEP -0.4407 0.2112 1531 -2.09 0.0371 Time*Students with Disabilities 0.03888 0.1789 1531 0.22 0.8279 time*Gifted 0.03961 0.189 1531 0.21 0.834 Time*Positive Learning Environment -0.1871 0.2685 1531 -0.7 0.486 Time*Positive Teacher Qualifications -0.3163 0.168 1531 -1.88 0.06 Time2*School Level Elementary -0.8227 0.2671 1531 -3.08 0.0021 **

PAGE 214

200 Effect School Level Estimate SE df t p Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.168 0.1302 1531 -1.29 0.197 Time2*Minority 0.1397 0.1288 1531 1.08 0.2784 Time2*LEP 0.1279 0.1 1531 1.28 0.2014 Time2*Students with Disabilities 0.09683 0.08471 1531 1.14 0.2532 time2*Gifted 0.02173 0.08874 1531 0.24 0.8066 Time2*Positive Learning Environment 0.1471 0.1301 1531 1.13 0.2584 Time2*Positive Teacher Qualifications 0.1462 0.08045 1531 1.82 0.0693 Covariance Parameter Estimate SE z p 3.9099 0.367 10.65 <.0001 ** 2.5557 0.1869 13.68 <.0001 ** 1.0726 0.1834 5.85 <.0001 ** Residual 6.187 0.2182 28.35 <.0001 ** Note: p < .05; ** p < .01 The next model added technology integration variables with the demographics, learning environment, and school level variables. These va riables included student access to various types of software, teachers regularly using various types of so ftware, frequency that stude nts use various types of software, and technology support. This model was estimated twice, first without gifted population but all school levels (see model 6a) and then with elementary and middle school levels and gifted population (see model 6b). When the model was estimated with all school levels without gifted, the significant technology parameter estimates were frequency that students use co ntent software at the intercept and the interactions between time and time2 with teachers who regularly use tech nology for administrative purposes and technical support – human (see Table 46). Other significant parameter estimates included the intercept, elementary and high school relative to middle school, free or reduced lunch status, minority, students with disabilities, and positive learning e nvironment, while LEP and positive teacher qualifications were not significant. Significant interactions with time included elementary relative to mi ddle school, free or reduced lunch status, LEP, and positive teacher qualifications. Signif icant interactions with time2 included elementary and high relative to middle school and po sitive teacher qualificati ons. All model fit indices indicated that this model had better fit (see Table 52). No additional variance was explained with this model. Two technology integration in dicators were retained in the final model for all school levels without

PAGE 215

201 gifted, percent of teacher s who use technology for administrativ e purposes and technology support human. Model 6a: Technology Integration with Demographics and Student Learning Environment by School Level (All School Levels without Gifted) Level 1: Student Absences = 0 + 1*Time + 2*Time2 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06* Teacher Qualifications + 07* Positive Learning Environment + 08*Access Content SW + 09*Access Office SW + 010*Access Ad Prod SW + 011*Teachers Use Deliver Instruction + 012*Teachers use Admin + 013*Frequency Students Use Content + 014*Frequency Students Use Tool + 015*Technical Support Human + 016*Technical Support Hardware + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16* Teacher Qualifications + 17* Positive Learning Environment + 18*Access Content SW + 19*Access Office SW + 110*Access Ad Prod SW + 111*Teachers Use Deliver Instruction + 112*Teachers use Admin + 113*Frequency Students Use Content + 114*Frequency Students Use Tool + 115*Technical Support Human + 116*Technical Support Hardware + u1 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26* Teacher Qualifications + 27* Positive Learning Environment + 28*Access Content SW + 29*Access Office SW + 210*Access Ad Prod SW + 211*Teachers Use Deliver Instruction + 212*Teachers use Admin + 213*Frequency Students Use Content + 214*Frequency Students Use Tool + 215*Technical Support Human + 216*Technical Support Hardware Mixed-Effects Model: Student Absences = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06* Teacher Qualifications + 07* Positive Learning Environment + 08*Access Content SW + 09*Access Office SW + 010*Access Ad Prod SW + 011*Teachers Use Deliver Instruction + 012*Teachers use Admin + 013*Frequency Students Use Content + 014*Frequency Students Use Tool + 015*Technical Support Human + 016*Technical Support Hardware + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Teacher Qualifications*Time + 17* Positive Learning Environment*Time + 18*Access Content SW*Time + 19*Access Office SW*Time + 110*Access Ad Prod SW*Time + 111*Teachers Use Deliver Instruction*Time + 112*Teachers use Admin*Time + 113*Frequency Students Use Content*Time + 114*Frequency Students Use Tool*Time + 115*Technical Support Human*Time + 116*Technical Support Hardware*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Teacher Qualifications*Time2 + 27* Positive Learning Environment*Time2 + 28*Access Content SW*Time2 + 29*Access Office SW*Time2 + 210*Access Ad Prod SW*Time2 + 211*Teachers Use Deliver Instruction*Time2 + 212*Teachers use Admin*Time2 + 213*Frequency Students Use Content*Time2 + 214*Frequency Students Use Tool*Time2 + 215*Technical Support Human*Time2 + 216*Technical Support Hardware*Time2 + u0 + u1 + r Table 46. Model 6a: Technology Integration w ith Demographics and Student Learni ng Environment by School Level (All School Levels without Gifted) Effect School Level Estimate SE df t p Intercept 10.0858 0.2408 2259 41.89 <.0001 ** Time 0.2367 0.4587 2203 0.52 0.6059 Time2 -0.00366 0.2179 2059 -0.02 0.9866 School Level Elementary -3.6634 0.2905 2059 -12.61 <.0001 **

PAGE 216

202 Effect School Level Estimate SE df t p School Level High 4.0514 0.3135 2059 12.92 <.0001 ** School Level Middle 0 . . Free Reduced Lunch 0.956 0.1276 2059 7.49 <.0001 ** Minority 0.5077 0.1386 2059 3.66 0.0003 ** LEP -0.1384 0.1168 2059 -1.19 0.2361 Students with Disabilities 0.9716 0.09176 2059 10.59 <.0001 ** Positive Learning Environment -1.0714 0.1036 2059 -10.34 <.0001 ** Positive Teacher Qualifications 0.07752 0.0787 2059 0.98 0.3247 Access Content Software 0.0834 0.0794 2059 1.05 0.2937 Access Office Software -0.0249 0.07898 2059 -0.32 0.7525 Access Advanced Production Software -0.02961 0.0804 2059 -0.37 0.7127 Teachers Use to Deliver Instruction -0.1421 0.08797 2059 -1.62 0.1064 Teachers Use for Administrative Purposes -0.1513 0.0909 2059 -1.66 0.0962 Frequency that Students Use Content Software 0.143 0.07278 2059 1.96 0.0496 Frequency Students Use Tool-Based Software 0.0432 0.07857 2059 0.55 0.5825 Technical Support Human -0.048 0.07203 2059 -0.67 0.5053 Technical Support Hardware -0.08185 0.06965 2059 -1.18 0.2401 Time*School Level Elementary 1.9079 0.5698 2059 3.35 0.0008 ** Time*School Level High -0.6158 0.5347 2059 -1.15 0.2496 Time*School Level Middle 0 . . Time*Free Reduced Lunch 0.5228 0.2548 2059 2.05 0.0403 Time*Minority -0.00161 0.2509 2059 -0.01 0.9949 Time*LEP -0.4421 0.2057 2059 -2.15 0.0317 Time*Students with Disabilities -0.05408 0.1734 2059 -0.31 0.7551 Time*Positive Learning Environment 0.03395 0.2319 2059 0.15 0.8836 Time*Positive Teacher Qualifications -0.4122 0.1656 2059 -2.49 0.0129 Time*Access Content Software 0.08198 0.1906 2059 0.43 0.6671 Time*Access Office Software 0.01079 0.19 2059 0.06 0.9547 Time*Access Advanced Production Software 0.1234 0.1903 2059 0.65 0.5167 Time*Teachers Use to Deliver Instruction -0.00222 0.2112 2059 -0.01 0.9916 Time*Teachers Use for Administrative Purposes 0.4922 0.2165 2059 2.27 0.0231 Time*Frequency that Students Use Content Software -0.09701 0.1844 2059 -0.53 0.5989 Time*Frequency Students Use Tool-Based Software -0.1651 0.1929 2059 -0.86 0.3921

PAGE 217

203 Effect School Level Estimate SE df t p Time*Technical Support Human 0.3383 0.1661 2059 2.04 0.0419 Time*Technical Support Hardware 0.1014 0.1687 2059 0.6 0.5478 Time2*School Level Elementary -0.8298 0.2732 2059 -3.04 0.0024 ** Time2*School Level High 0.6136 0.2543 2059 2.41 0.0159 Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.06898 0.1226 2059 -0.56 0.5738 Time2*Minority -0.1537 0.1189 2059 -1.29 0.1963 Time2*LEP 0.1511 0.0973 2059 1.55 0.1207 Time2*Students with Disabilities 0.1053 0.08208 2059 1.28 0.1999 Time2*Positive Learning Environment -0.00892 0.1124 2059 -0.08 0.9368 Time2*Positive Teacher Qualifications 0.1856 0.07924 2059 2.34 0.0193 Time2*Access Content Software -0.1053 0.09173 2059 -1.15 0.2511 Time2*Access Office Software -0.02278 0.0915 2059 -0.25 0.8034 Time2*Access Advanced Production Software -0.0043 0.09154 2059 -0.05 0.9625 Time2*Teachers Use to Deliver Instruction 0.01714 0.1024 2059 0.17 0.8671 Time2*Teachers Use for Administrative Purposes -0.2106 0.1039 2059 -2.03 0.0428 Time2*Frequency that Students Use Content Software 0.04685 0.08961 2059 0.52 0.6012 Time2*Frequency Students Use Tool-Based Software 0.05756 0.09445 2059 0.61 0.5423 Time2*Technical Support Human -0.2003 0.07971 2059 -2.51 0.0121 Time2*Technical Support Hardware -0.1527 0.08118 2059 -1.88 0.0601 Covariance Parameter Estimate SE z p 9.9757 0.5294 18.84 <.0001 ** 2.0476 0.2189 9.35 <.0001 ** 0.8108 0.1786 4.54 <.0001 ** Residual 7.3218 0.2248 32.57 <.0001 ** Note: p < .05; ** p < .01 Similar results were found with the elementary and middle school data with gifted. There were no significant technology parameter estima tes at the intercept, and only the interaction of time with the percent of teachers who use technology for admi nistrative purposes was significant (see

PAGE 218

204 Table 47). Other significant parameter estimates included the intercept, time, time2, elementary relative to middle school, free or reduced lunch status, student s with disabilities, gifted, and positive learning environment, while minority, LEP, and positive teacher qualifications were not significant. Significant interactions with time included elementary relative to middle school, free and reduced lunch status, LEP, and positive learning environment. Th e only significant interaction with time2 was elementary relative to middle school. Three of the fit indices indicated that this model had better fit (see Table 53), even though adding the technology integration indicators to the model did not explain any additional variance. Teachers who use technology for administrative purposes was the only technology integration indicator retained in the final model for the data with elem entary and middle schools and gifted. Model 6b: Technology Integration with Demographics and Student Learning Environment by School Level (Elementary and Middle Schools with Gifted) Level 1: Student Absences = 0 + 1*Time + 2*Time2 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 09*Access Content SW + 010*Access Office SW + 011*Access Ad Prod SW + 012*Teachers Use Deliver Instruction + 013*Teachers use Admin + 014*Frequency Students Use Content + 015*Frequency Students Use Tool + 016*Technical Support Human + 017*Technical Support Hardware + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16*Gifted + 17* Teacher Qualifications + 18* Positive Learning Environment + 19*Access Content SW + 110*Access Office SW + 111*Access Ad Prod SW + 112*Teachers Use Deliver Instruction + 113*Teachers use Admin + 114*Frequency Students Use Content + 115*Frequency Students Use Tool + 116*Technical Support Human + 117*Technical Support Hardware + u1 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26*Gifted + 27* Teacher Qualifications + 28* Positive Learning Environment + 29*Access Content SW + 210*Access Office SW + 211*Access Ad Prod SW + 212*Teachers Use Deliver Instruction + 213*Teachers use Admin + 214*Frequency Students Use Content + 215*Frequency Students Use Tool + 216*Technical Support Human + 217*Technical Support Hardware Mixed-Effects Model: Student Absences = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 09*Access Content SW + 010*Access Office SW + 011*Access Ad Prod SW + 012*Teachers Use Deliver Instruction + 013*Teachers use Admin + 014*Frequency Students Use Content + 015*Frequency Students Use Tool + 016*Technical Support Human + 017*Technical Support Hardware + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Gifted*Time + 17* Teacher Qualifications*Time + 18* Positive Learning Environment*Time + 19*Access Content SW*Time + 110*Access Office SW*Time + 111*Access Ad Prod SW*Time + 112*Teachers Use Deliver Instruction*Time + 113*Teachers use Admin*Time + 114*Frequency Students Use Content*Time + 115*Frequency Students Use Tool*Time + 116*Technical Support Human*Time + 117*Technical Support Hardware*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Gifted*Time2 + 27* Teacher Qualifications*Time2 + 28* Positive Learning Environment*Time2 + 29*Access Content SW*Time2 + 210*Access Office SW*Time2 + 211*Access Ad Prod SW*Time2 + 212*Teachers Use Deliver Instruction*Time2 + 213*Teachers use Admin*Time2 + 214*Frequency Students Use Content*Time2 + 215*Frequency Students Use Tool*Time2 + 216*Technical Support Human*Time2 + 217*Technical Support Hardware*Time2 + u0 + u1 + r

PAGE 219

205 Table 47. Model 6b: Technology Integration w ith Demographics and Student Learni ng Environment by School Level for Elementary and Middle Schools with Gifted Effect School Level Estimate SE df t p Intercept 10.1578 0.2041 1825 49.76 <.0001 ** Time 0.1526 0.4716 1704 0.32 0.7463 Time2 0.05106 0.2239 1504 0.23 0.8197 School Level Elementary -3.8731 0.2524 1504 -15.35 <.0001 ** School Level Middle 0 . . Free Reduced Lunch 1.0593 0.1181 1504 8.97 <.0001 ** Minority 0.1874 0.1275 1504 1.47 0.1418 LEP -0.1655 0.09803 1504 -1.69 0.0915 Students with Disabilities 0.804 0.08001 1504 10.05 <.0001 ** Gifted -0.3717 0.08711 1504 -4.27 <.0001 ** Positive Learning Environment -0.9479 0.1052 1504 -9.02 <.0001 ** Positive Teacher Qualifications -0.0299 0.07133 1504 -0.42 0.6752 Access Content Software 0.05763 0.07475 1504 0.77 0.4409 Access Office Software -0.0016 0.07158 1504 -0.02 0.9821 Access Advanced Production Software -0.00514 0.07468 1504 -0.07 0.9452 Teachers Use to Deliver Instruction -0.1332 0.08247 1504 -1.61 0.1065 Teachers Use for Administrative Purposes -0.1593 0.08651 1504 -1.84 0.0658 Frequency that Students Use Content Software 0.1016 0.06888 1504 1.47 0.1405 Frequency Students Use Tool-Based Software 0.0086 0.07264 1504 0.12 0.9058 Technical Support Human -0.04038 0.06755 1504 -0.6 0.5501 Technical Support Hardware -0.02848 0.06417 1504 -0.44 0.6572 Time*School Level Elementary 2.1855 0.5967 1504 3.66 0.0003 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch 0.8276 0.2779 1504 2.98 0.0029 ** Time*Minority -0.5322 0.2758 1504 -1.93 0.0539 Time*LEP -0.4703 0.2127 1504 -2.21 0.0272 Time*Students with Disabilities 0.03309 0.18 1504 0.18 0.8542 Time*Gifted 0.09982 0.1918 1504 0.52 0.6029 Time*Positive Learning Environment -0.2978 0.2718 1504 -1.1 0.2735 Time*Positive Teacher Qualifications -0.3361 0.1703 1504 -1.97 0.0486 Time*Access Content Software 0.09871 0.1942 1504 0.51 0.6113 Time*Access Office Software -0.1085 0.1889 1504 -0.57 0.566 Time*Access Advanced -0.05071 0.1913 1504 -0.27 0.791

PAGE 220

206 Effect School Level Estimate SE df t p Production Software Time*Teachers Use to Deliver Instruction -0.06853 0.2172 1504 -0.32 0.7524 Time*Teachers Use for Administrative Purposes 0.5253 0.226 1504 2.32 0.0202 Time*Frequency that Students Use Content Software 0.06608 0.1931 1504 0.34 0.7323 Time*Frequency Students Use Tool-Based Software 0.03403 0.196 1504 0.17 0.8622 Time*Technical Support Human 0.1805 0.1733 1504 1.04 0.2977 Time*Technical Support Hardware -0.0968 0.1702 1504 -0.57 0.5696 Time2*School Level Elementary -0.9501 0.2862 1504 -3.32 0.0009 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.1702 0.134 1504 -1.27 0.2041 Time2*Minority 0.1187 0.1309 1504 0.91 0.3649 Time2*LEP 0.1464 0.101 1504 1.45 0.1474 Time2*Students with Disabilities 0.09916 0.08548 1504 1.16 0.2462 Time2*Gifted -0.0053 0.09034 1504 -0.06 0.9532 Time2*Positive Learning Environment 0.181 0.1318 1504 1.37 0.1698 Time2*Positive Teacher Qualifications 0.1539 0.08213 1504 1.87 0.0611 Time2*Access Content Software -0.09308 0.09481 1504 -0.98 0.3264 Time2*Access Office Software 0.03233 0.09217 1504 0.35 0.7258 Time2*Access Advanced Production Software 0.08214 0.09286 1504 0.88 0.3766 Time2*Teachers Use to Deliver Instruction 0.04502 0.1069 1504 0.42 0.6738 Time2*Teachers Use for Administrative Purposes -0.1971 0.1096 1504 -1.8 0.0723 Time2*Frequency that Students Use Content Software -0.05648 0.09527 1504 -0.59 0.5534 Time2*Frequency Students Use Tool-Based Software -0.01862 0.09759 1504 -0.19 0.8488 Time2*Technical Support Human -0.1182 0.0838 1504 -1.41 0.1587 Time2*Technical Support Hardware -0.05149 0.08301 1504 -0.62 0.5352 Covariance Parameter Estimate SE z p 3.887 0.3689 10.54 <.0001 ** 2.5845 0.1854 13.94 <.0001 ** 0.8673 0.1805 4.8 <.0001 ** Residual 6.2248 0.2207 28.21 <.0001 ** Note: p < .05; ** p < .01

PAGE 221

207 The last models estimated in order to answer the first hypothesis included all school levels, demographic, student learning environment, and significant technology integration variables. These models were different because the model fit to the data fo r all schools levels without gifted included two technology integration variables percent of teachers who regularly use technology for administrative purposes and the level of technology support human (see model 8a); while the model fitted to the data with elementary and middle school levels and gifted included only one technology integration variable – the percent of teachers who regular ly use technology for administrativ e purposes (see model 8b). For the model with all schools levels and no gifted, the same parameter estimates and interactions identified in the previous models as significant were significant again (see Table 48). Although, th ere was no difference in the percentage of variance explained in this model than there was in the Demographic Model with Student Learning Environment by school level or the Technology Integration with Demographic and Student Learning Environment Model by school level, the BIC index indicated better model fit (see Table 52). The level-1 residuals for the final model for predicting stud ent absences using all school levels without gifted ranged between -19.04 and 22.48 with a standard devi ation of 2.22. Although th ere were outliers, skewness was 0.47 and kurtosis was 13.23, which would indicate that most of the residuals were centered at the mean. Distribution of the empirical bayes intercepts ranged between -12.50 and 15.13 with standard deviation of 2.86. Skewness was 0.84, and kurtosis was 2.81, which indicated that most of the intercept residuals at level-2 were centered at the mean. Distribu tion of the empirical bayes slopes ranged between -2.77 and 4.77 with standard deviation of 0.75. Skewness was 1.31, and kurtosis was 4.44, which indicated that most of the slope residuals at level-2 were not normally distribute d. Because the residuals for student absences outcome were not normally distributed the results of the analysis may be biased. Final Model 7a: Significant Technology Integration Indicators with Demographics and Student Learning Environment by School Level (All School Levels without Gifted) Level 1: Student Absences = 0 + 1*Time + 2*Time2 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Teacher Qualifications + 07* Positive Learning Environment + 08*Teachers use Admin + 09*Technical Support Hardware + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16* Teacher Qualifications + 17* Positive Learning Environment + 18* Teachers use Admin + 19* Technical Support Hardware + u1 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26* Teacher Qualifications + 27* Positive Learning Environment + 28*Teachers use Admin + 29* Technical Support Hardware

PAGE 222

208 Mixed-Effects Model: Student Absences = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06* Teacher Qualifications + 07* Positive Learning Environment + 08*Teachers use Admin + 09*Technical Support Hardware + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD*Time + 16*Teacher Qualifications*Time + 17* Positive Learning Environment*Time + 18* Teachers use Admin*Time + 19* Technical Support Hardware*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Teacher Qualifications*Time2 + 27* Positive Learning Environment*Time2 + 28*Teachers use Admin*Time2 + 29* Technical Support Hardware*Time2 + u0 + u1 + r Table 48. Final Model 7a: Significant Technology Integration with Demographics and Student Learning Environment by School Level (A ll School Levels without Gifted) Effect School Level Estimate SE df t p Intercept 9.9922 0.2377 2259 42.03 <.0001 ** Time 0.2026 0.4427 2203 0.46 0.6473 Time2 0.01098 0.2093 2080 0.05 0.9582 School Level Elementary -3.5418 0.2857 2080 -12.4 <.0001 ** School Level High 3.9887 0.3115 2080 12.8 <.0001 ** School Level Middle 0 . . Free Reduced Lunch 0.9618 0.1267 2080 7.59 <.0001 ** Minority 0.514 0.1381 2080 3.72 0.0002 ** LEP -0.1247 0.1167 2080 -1.07 0.2853 Students with Disabilities 0.9744 0.09174 2080 10.62 <.0001 ** Positive Learning Environment -1.0886 0.1036 2080 -10.51 <.0001 ** Positive Teacher Qualifications 0.07804 0.07884 2080 0.99 0.3224 Teachers Use for Administrative Purposes -0.1873 0.07461 2080 -2.51 0.0121 Technical Support Human -0.03634 0.07108 2080 -0.51 0.6092 Time*School Level Elementary 1.9454 0.5456 2080 3.57 0.0004 ** Time*School Level High -0.6503 0.5268 2080 -1.23 0.2172 Time*School Level Middle 0 . . Time*Free Reduced Lunch 0.5353 0.2511 2080 2.13 0.0331 Time*Minority 0.002798 0.2491 2080 0.01 0.991 Time*LEP -0.4589 0.2054 2080 -2.23 0.0256 Time*Students with Disabilities -0.03988 0.1728 2080 -0.23 0.8176 Time*Positive Learning Environment 0.05275 0.2302 2080 0.23 0.8188 Time*Positive Teacher Qualifications -0.4224 0.1644 2080 -2.57 0.0102 Time*Teachers Use for Administrative Purposes 0.5167 0.1749 2080 2.95 0.0032 ** Time*Technical Support Human 0.3417 0.1636 2080 2.09 0.0369

PAGE 223

209 Effect School Level Estimate SE df t p Time2*School Level Elementary -0.8454 0.2598 2080 -3.25 0.0012 ** Time2*School Level High 0.6583 0.2497 2080 2.64 0.0085 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.07835 0.1203 2080 -0.65 0.5151 Time2*Minority -0.1574 0.1179 2080 -1.34 0.182 Time2*LEP 0.1596 0.09701 2080 1.65 0.1 Time2*Students with Disabilities 0.09823 0.08169 2080 1.2 0.2293 Time2*Positive Learning Environment -0.01234 0.1114 2080 -0.11 0.9118 Time2*Positive Teacher Qualifications 0.1989 0.07828 2080 2.54 0.0111 Time2*Teachers Use for Administrative Purposes -0.2375 0.08382 2080 -2.83 0.0046 ** Time2*Technical Support Human -0.2152 0.07864 2080 -2.74 0.0063 ** Covariance Parameter Estimate SE z p 9.9732 0.5278 18.9 <.0001 ** 2.0306 0.2198 9.24 <.0001 ** 0.9645 0.1811 5.33 <.0001 ** Residual 7.3005 0.2238 32.62 <.0001 ** Note: p < .05; ** p < .01 For the model with elementary and middle school levels and gifted, the same significant parameter estimate, percent of teachers who regularly use techno logy for administrative purposes, was identified as in the previous model (see Table 49). Interactions between time and time2 with percent of teachers who regularly use technology for administ rative purposes were significant. Although there was no difference in the percentage of variance explained in this model than was in the Demographic Model with Student Learning Environment by school level or the Technology Integration with Demographic and Student Learning Environment Model by school level, the BIC index all indicated better model fit (see Table 52). The level-1 residuals for the final model for predicting student absences using elementary and middle schools with gifted ranged between -16.49 and 22.23 with a standard deviation of 2.12. There were outliers. Skewness was 0.75 and kurtosis was 13.71, which would indicate that the residuals were not normally distributed and most were centered at the mean. Di stribution of the empirical bayes intercepts ranged between -5.88 and 12.94 with standard deviation of 1.87. Skewness was 1.68, and kurtosis was 5.88, which indicated that the intercept residuals at level-2 were not distributed no rmally. Distribution of the empirical

PAGE 224

210 bayes slopes ranged between -3.17 and 6.57 with standard deviation of 1.03. Skewness was 1.24, and kurtosis was 3.93, which indicated that most of the slope residuals at level-2 were not normally distributed. Because the residuals for student absences outcome were not normally distributed, the results of the analysis may be biased. Final Model 7b: Technology Integration with Demographics and Student Learning Environment by School Level (Elementary and Middle Schools with Gifted) Level 1: Student Absences = 0 + 1*Time + 2*Time2 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 09*Teachers use Admin + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16*Gifted + 17* Teacher Qualifications + 18* Positive Learning Environment + 19* Teachers use Admin + u1 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26*Gifted + 27* Teacher Qualifications + 28* Positive Learning Environment + 29*Teachers use Admin Mixed-Effects Model: Student Absences = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 09*Teachers use Admin + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Gifted*Time + 17* Teacher Qualifications*Time + 18* Positive Learning Environment*Time + 19*Teachers use Admin*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Gifted*Time2 + 27* Teacher Qualifications*Time2 + 28* Positive Learning Environment*Time2 + 29*Teachers use Admin*Time2 + u0 + u1 + r Table 49. Final Model 7b: Significant Technology Integration with Demographics and Student Learning Environment by School Level for Elemen tary and Middle Schools with Gifted Effect School Level Estimate SE df t p Intercept 10.0879 0.2007 1825 50.27 <.0001 ** Time 0.1225 0.4547 1704 0.27 0.7877 Time2 0.06814 0.2147 1528 0.32 0.751 School Level Elementary -3.7797 0.2473 1528 -15.28 <.0001 ** School Level Middle 0 . . Free Reduced Lunch 1.052 0.1167 1528 9.01 <.0001 ** Minority 0.2045 0.1268 1528 1.61 0.1069 LEP -0.1649 0.09745 1528 -1.69 0.0908 Students with Disabilities 0.8098 0.07991 1528 10.13 <.0001 ** Gifted -0.3837 0.08644 1528 -4.44 <.0001 ** Positive Learning Environment -0.9598 0.1051 1528 -9.13 <.0001 ** Positive Teacher Qualifications -0.03555 0.07137 1528 -0.5 0.6185 Teachers Use for -0.2047 0.06924 1528 -2.96 0.0032 **

PAGE 225

211 Effect School Level Estimate SE df t p Administrative Purposes Time*School Level Elementary 2.2127 0.5723 1528 3.87 0.0001 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch 0.8636 0.272 1528 3.18 0.0015 ** Time*Minority -0.5038 0.273 1528 -1.85 0.0652 Time*LEP -0.452 0.2109 1528 -2.14 0.0322 Time*Students with Disabilities 0.05196 0.1786 1528 0.29 0.7712 Time*Gifted 0.07717 0.1892 1528 0.41 0.6834 Time*Positive Learning Environment -0.268 0.2694 1528 -0.99 0.32 Time*Positive Teacher Qualifications -0.3045 0.1678 1528 -1.81 0.0698 Time*Teachers Use for Administrative Purposes 0.5563 0.1774 1528 3.14 0.0017 ** Time2*School Level Elementary -0.9621 0.2724 1528 -3.53 0.0004 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.2061 0.1306 1528 -1.58 0.1146 Time2*Minority 0.1078 0.1292 1528 0.83 0.4045 Time2*LEP 0.1335 0.09985 1528 1.34 0.1813 Time2*Students with Disabilities 0.0913 0.08461 1528 1.08 0.2807 Time2*Gifted 0.005296 0.08883 1528 0.06 0.9525 Time2*Positive Learning Environment 0.1788 0.1304 1528 1.37 0.1707 Time2*Positive Teacher Qualifications 0.1419 0.08034 1528 1.77 0.0776 Time2*Teachers Use for Administrative Purposes -0.2272 0.08581 1528 -2.65 0.0082 ** Covariance Parameter Estimate SE z p 3.9259 0.3666 10.71 <.0001 ** 2.5493 0.1864 13.68 <.0001 ** 1.0783 0.1828 5.9 <.0001 ** Residual 6.1579 0.2172 28.34 <.0001 ** Note: p < .05; ** p < .01 The last step was to add in USDOE funded Magnet Schools and USDOE Technology Magnet Schools as variables in the model. Results of this model indicated that having U.S. technology magnet school status was significant at the interc ept; however, the interactions with time and time2 with having U.S. technology magnet school status were not significant in both datasets (see Table 50 and Table 51).

PAGE 226

212 The parameter estimates for having U.S. magnet school stat us at the intercept as well as the interaction with time with having U.S. magnet school status were significant predictors of percent of students absent more than 21 days in the dataset that included all school levels without gifted (see Table 50), while only the interactions with time and time2 with having U.S. magnet school status were significant predictors of percent of students absent more than 21 days in the dataset that included elementary and middle school with gifted (see Table 51). U.S. technology magnet schools were predicted to start the study with 3.24 lower percent of students absent more than 21 days w ith the all school level without gifted dataset and 2.54 lower percent of students absent more than 21 days in the elementary and middle school with gifted dataset. On the other hand, U.S. magnet schools had the reverse relationship. U.S. magnet schools were predicted to start the study with 1.93 percent higher of students absent more than 21 days with the all school level without gifted dataset and 0.66 percent higher of students absent more than 21 days in the elementary and middle school with gifted dataset. Table 50. Model 8a: Magnet Schools with Significant Technology Integration Demographics and Student Learning Environment by School Level (A ll School Levels without Gifted) Effect School Level Estimate SE df t p Intercept 8.6371 1.1461 2257 7.54 <.0001 ** Time 1.1814 1.9339 2201 0.61 0.5413 Time2 -0.3936 0.9141 2078 -0.43 0.6668 School Level Elementary -3.5135 0.2856 2078 -12.3 <.0001 ** School Level High 4.0148 0.3108 2078 12.92 <.0001 ** School Level Middle 0 . . Free Reduced Lunch 0.952 0.1266 2078 7.52 <.0001 ** Minority 0.498 0.1386 2078 3.59 0.0003 ** LEP -0.1168 0.1167 2078 -1 0.3172 Students with Disabilities 0.9847 0.09173 2078 10.73 <.0001 ** Positive Learning Environment -1.0835 0.1036 2078 -10.46 <.0001 ** Positive Teacher Qualifications 0.07696 0.07878 2078 0.98 0.3287 Teachers Use for Administrative Purposes -0.1817 0.07471 2078 -2.43 0.0151 Technical Support Human -0.03495 0.07107 2078 -0.49 0.6229 Not a Technology Magnet School US 3.2352 1.3168 2078 2.46 0.0141 Technology Magnet School US 0 . . Not a US Magnet School -1.927 0.7027 2078 -2.74 0.0062 **

PAGE 227

213 Effect School Level Estimate SE df t p US Magnet School 0 . . Time*School Level Elementary 2.0352 0.5456 2078 3.73 0.0002 ** Time*School Level High -0.6263 0.5261 2078 -1.19 0.234 Time*School Level Middle 0 . . Time*Free Reduced Lunch 0.4938 0.2512 2078 1.97 0.0494 Time*Minority -0.07251 0.25 2078 -0.29 0.7718 Time*LEP -0.4107 0.2058 2078 -2 0.0461 Time*Students with Disabilities -0.03385 0.1729 2078 -0.2 0.8448 Time*Positive Learning Environment 0.03441 0.23 2078 0.15 0.8811 Time*Positive Teacher Qualifications -0.4172 0.1641 2078 -2.54 0.0111 Time*Teachers Use for Administrative Purposes 0.49 0.1751 2078 2.8 0.0052 ** Time*Technical Support Human 0.3471 0.1637 2078 2.12 0.0341 Time*Not a Technology Magnet School US 2.5563 2.2082 2078 1.16 0.2471 Time*Technology Magnet School US 0 . . Time*Not a US Magnet School -3.6699 1.179 2078 -3.11 0.0019 ** Time*US Magnet School 0 . . Time2*School Level Elementary -0.8759 0.2598 2078 -3.37 0.0008 ** Time2*School Level High 0.653 0.2495 2078 2.62 0.0089 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.06408 0.1204 2078 -0.53 0.5947 Time2*Minority -0.1317 0.1183 2078 -1.11 0.2658 Time2*LEP 0.1419 0.09719 2078 1.46 0.1445 Time2*Students with Disabilities 0.09691 0.08175 2078 1.19 0.236 Time2*Positive Learning Environment -0.00351 0.1113 2078 -0.03 0.9749 Time2*Positive Teacher Qualifications 0.1957 0.07818 2078 2.5 0.0124 Time2*Teachers Use for Administrative Purposes -0.228 0.08394 2078 -2.72 0.0067 ** Time2*Technical Support Human -0.2173 0.0787 2078 -2.76 0.0058 ** Time2*Not a Technology Magnet School US -0.609 1.0451 2078 -0.58 0.5602 Time2*Technology Magnet School US 0 . . Time2*Not a US Magnet School 1.0575 0.5595 2078 1.89 0.0589 Time2*US Magnet School 0 . .

PAGE 228

214 Covariance Parameter Estimate SE z p 9.8976 0.5249 18.86 <.0001 ** 1.9702 0.2188 9.01 <.0001 ** 0.9356 0.1801 5.2 <.0001 ** Residual 7.2828 0.2232 32.62 <.0001 ** Note: p < .05; ** p < .01 Table 51. Model 8b: Magnet Schools with Significant Technology Integration Demographics and Student Learning Environment by School Level for Elemen tary and Middle Schools with Gifted Effect School Level Estimate SE df t p Intercept 8.2082 1.0128 1823 8.1 <.0001 ** Time 0.5601 2.1257 1702 0.26 0.7922 Time2 0.06375 1.0177 1526 0.06 0.9501 School Level Elementary -3.7833 0.2471 1526 -15.31 <.0001 ** School Level Middle 0 . . Free Reduced Lunch 1.0535 0.1166 1526 9.03 <.0001 ** Minority 0.2131 0.1271 1526 1.68 0.0938 LEP -0.1695 0.09752 1526 -1.74 0.0824 Students with Disabilities 0.8196 0.07993 1526 10.25 <.0001 ** Gifted -0.3833 0.08628 1526 -4.44 <.0001 ** Positive Learning Environment -0.949 0.105 1526 -9.04 <.0001 ** Positive Teacher Qualifications -0.04138 0.07128 1526 -0.58 0.5616 Teachers Use for Administrative Purposes -0.1931 0.06926 1526 -2.79 0.0054 ** Not a Technology Magnet School US 2.5372 1.1544 1526 2.2 0.0281 Technology Magnet School US 0 . . Not a US Magnet School -0.6561 0.5993 1526 -1.09 0.2738 US Magnet School 0 . . Time*School Level Elementary 2.2867 0.5714 1526 4 <.0001 ** Time*School Level Middle 0 . . Time*Free Reduced Lunch 0.8053 0.2723 1526 2.96 0.0031 ** Time*Minority -0.5818 0.2733 1526 -2.13 0.0334 Time*LEP -0.3943 0.2111 1526 -1.87 0.0619 Time*Students with Disabilities 0.05492 0.1785 1526 0.31 0.7583 Time*Gifted 0.06121 0.1888 1526 0.32 0.7459 Time*Positive Learning Environment -0.2902 0.2689 1526 -1.08 0.2807 Time*Positive Teacher Qualifications -0.2931 0.1675 1526 -1.75 0.0804 Time*Teachers Use for 0.5285 0.1777 1526 2.97 0.003 **

PAGE 229

215 Effect School Level Estimate SE df t p Administrative Purposes Time*Not a Technology Magnet School US 3.6062 2.3931 1526 1.51 0.132 Time*Technology Magnet School US 0 . . Time*Not a US Magnet School -4.1758 1.2171 1526 -3.43 0.0006 ** Time*US Magnet School 0 . . Time2*School Level Elementary -0.9889 0.272 1526 -3.64 0.0003 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.1877 0.1308 1526 -1.44 0.1514 Time2*Minority 0.1312 0.1294 1526 1.01 0.3106 Time2*LEP 0.1137 0.09997 1526 1.14 0.2554 Time2*Students with Disabilities 0.08898 0.08456 1526 1.05 0.2928 Time2*Gifted 0.009079 0.08873 1526 0.1 0.9185 Time2*Positive Learning Environment 0.1934 0.1302 1526 1.49 0.1377 Time2*Positive Teacher Qualifications 0.1367 0.08021 1526 1.7 0.0886 Time2*Teachers Use for Administrative Purposes -0.2203 0.08595 1526 -2.56 0.0105 Time2*Not a Technology Magnet School US -1.0704 1.1448 1526 -0.94 0.3499 Time2*Technology Magnet School US 0 . . Time2*Not a US Magnet School 1.1144 0.5782 1526 1.93 0.0541 Time2*US Magnet School 0 . . Covariance Parameter Estimate SE z p 3.9043 0.365 10.7 <.0001 ** 2.5135 0.1851 13.58 <.0001 ** 1.0275 0.1808 5.68 <.0001 ** Residual 6.1369 0.2165 28.35 <.0001 ** Note: p < .05; ** p < .01 Table 52. Model Fit Indices for Models Predicting FCAT Stud ent Absences Scores for All School Levels (without Gifted) Model -2 Log Likelihood AIC (smaller is better) AICC (smaller is better) BIC (smaller is better) Model 1: Absences Predicted by Average Absences of All Schools in Florida 40464.6 40470.6 40470.6 40487.8 Model 2a: Time as a Predictor of Abse nces 40065.4 40077.4 40077.4 40111.9

PAGE 230

216 Model -2 Log Likelihood AIC (smaller is better) AICC (smaller is better) BIC (smaller is better) Quadratic Model 2b: Time2 as a Predictor of Absences 40022.8 40036.8 40036.8 40077 Model 3: Time, Time2, and School Level as Predictors of Absences 39266 39292 39292 39366.7 Model 4a: Absences predicted by Time, School Level, and Demographics Variables 36544.7 36594.7 36594.9 36737.8 Model 5a: Demographics and Student Learning Environment by School Level 36389.7 36451.7 36452 36629.2 Model 6a: Technology Integration with Demographics and Student Learning Environment by School Level 36304.4 36420.4 36421.5 36752.4 Final Model 7a: Significant Technology Integration with Demographics and Student Learning Environment by School Level 36366 36440 36440.4 36651.8 Model 8a: Magnet Schools with Significant Technology Integration Demographics and Student Learning Environment by School Level (All School Levels without Gifted) 36323.1 36409.1 36409.7 36655.3 Table 53. Model Fit Indices for Models Predicting FCAT Studen t Absences Scores for Elementary and Middle School Levels (with Gifted) Model -2 Log Likelihood AIC (smaller is better) AICC (smaller is better) BIC (smaller is better) Model 1: Absences Predicted by Average Absences of All Elementa ry and Middle Schools in Florida 33095.3 33101.3 33101.3 33118 Model 4b: Absences predicted by Time, School Level, and Demographics Variables No High School includes gifted 26929.5 26979.5 26979.7 27117.2 Model 5b: Demographics and Teacher Qualifications by School Level 26804.9 26866.9 26867.3 27037.7 Model 6b: Technology Integration with Demographics and Teacher Qualifications by School Level 26739.3 26855.3 26856.7 27174.9 Final Model 7b: Significant Technology Integration with Demographics and Student Learning Environment by School Level 26792.4 26860.4 26860.9 27047.8 Model 8b: Magnet Schools with Significant Technology Integration Demographics and Student Learning Environment by School Level for Elementary and Middle Schools with Gifted 26752 26832 26832.6 27052.4

PAGE 231

217 The result of the analysis for all the models indicated that Hypothesis 1 for Research Question 2 was partially correct. When the sample included schools at all three school levels there was a significant relationship at the intercept between the percent of teachers who regularly use technology for administrative purposes (-0.1873) and percent of students absent more than 21 days at the intercept when all other school level, demographic, and school learning environment factors were controlled. Also, there were significant interactions between time (0.5167) and time2 (-0.2375) with the pe rcent of teachers who regularly use technology for administrative purposes for predicting the percent of students absent more than 21 days and significant interactions between time (0.3417) and time2 (-0.2152) with technology support human for predicting the percent of students absent more than 21 days. The interactions of time and time2 with the percent of teachers who regularly use te chnology for administrative purposes resulted in a curvilinear trend. After controlling so that all other variables were held at the mean, the trend for each school level could be examined separately, by co mparing schools with different levels that teachers use technology for administrative purposes. Figure 27 illustrates the relati onship between the percent of teachers who regularly use technology for administrative purposes and the percent of students absent more than 21 days for high schools. Percent of teachers who regularly use technology for administrative purposes was compared at one and two standard deviations below the mean, the mean, and one and two standard deviations above the mean. This allowed the extreme cases of schools that had the highest percent of teachers who regularly use technology for administrative purposes, +2 standard deviations above the mean, and schools that had the lowest percent of teachers regularl y use technology for administrative pur poses, -2 standard deviations below the mean to be compared. High schools that had the highest percen tage of teachers who regularly use technology for administrative purposes started the study in 2003-04 with 13.61% of students absent more than 21 days and schools that had the lowest percent of teachers who regularly use technology for administrative purposes had started with 14.36% of students absent more than 21 days. This difference of 0.75% was significant because there were so many schools in the sample; however, the practical importance was modest. The interaction between the percent of teachers who regularly use technology for administrative purposes and time and time2 with percent of students absent more than 21 days was significant, so the slopes of the trends were curv ilinear. In 2004-05, schools at one and two standard

PAGE 232

218 deviations below the mean of percent of teachers who use technology for administrative purposes experienced a decrease in the percent of students absent more than 21 days, while high schools that had two standard deviations above the mean of percent of teachers who use technology for administrative purposes experienced the greatest increase in the percentage of students absent more than 21 days (0.78%). By 200506, all high schools gained in their percent of students absent for more than 21 days, and high schools with the lowest percent of teachers who regularly use technology for administrative purposes had the highest increase in percent of students absent more than 21 days (see Figure 27). Relationship between Percent of Teachers that Regularly Use Technology for Administrative Purposes and Student Absences in High Schools0 2 4 6 8 10 12 14 16 18 20 200320042005 YearPercent of Students with 21+ Days Absent High Schools + 2SD High Schools + 1SD High Schools mean High Schools 1SD High Schools 2SD Figure 27. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Student Absences in High Schools. Middle schools had a similar beginning pattern to hi gh school, that is after controlling for all other factors, schools that were two stan dard deviations above the mean in the percent of teachers who regularly use technology for administrative purposes had the lowest percent of students absent more than 21 days in 2003-04 (9.62%), while those with two standard deviations below the mean had the highest levels (10.37%). Although this difference of 0.75% was sign ificant due to the large sample size, the practical

PAGE 233

219 importance is modest. Because th e interactions between time and time2 and the percent of teachers who regularly use technology for administrative purposes with percent of students absent more than 21 days were significant, the trends were curvilinear. Betw een 2003-04 and 2004-05 middle schools at one and two standard deviations below the mean in percent of teachers using tech nology for administrative purposes experienced decreases in the percent of students absent more than 21 days (0.07% and 0.35%, respectively), while schools that were two standard deviations above the mean experienced the greatest increases in percent of students abse nt more than 21 days (0.77%). These trends reversed in 2005-06, with middle schools with the highest percentage of teach ers who regularly use tech nology for administrative purposes having a decline in the percentage of students with more than 21 days absent (0.16%) and schools at two standard deviations below the mean of perc entage of teachers who regul arly use technology for administrative purposes having the greates t increases (0.63%) (see Figure 28). Relationship between Percent of Teachers that Regularly Use Technology for Administrative Purposes and Student Absences in Middle Schools0 2 4 6 8 10 12 14 16 18 20 200320042005 YearPercent of Students with 21+ Days Absent Middle Schools + 2SD Middle Schools + 1SD Middle Schools mean Middle Schools 1SD Middle Schools 2SD Figure 28. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Student Absences in Middle Schools.

PAGE 234

220 The trends for elementary schools were similar to those of high and middle schools. Schools with the highest percent of teachers who regularly use technology for administrative purposes began the study with the lowest percent of students absent more th an 21 days (6.08%), while schools with the lowest percent of teachers who regularly use technology for administrative purposes began the study with the highest percent of students absent more than 21 days (6.83%). Although this difference of 0.75% was statistically significant, it had modest practical im portance. Between 2003-04 and 2004-05, elementary schools experienced an increase in percent of students absent more than 21 days with elementary schools that had the greatest percentage of teachers who regularly use techno logy for administrative purposes associated with the greatest increase in the percent of students absent more than 21 days (0.87%). At the end of the study in 2005-06, the trends changed. Elementary schools at two standard deviations below the mean in percentage of teachers who regularly use technology for administra tive purposes experienced increases in the percentage of students who were absent more than 21 days (0.04%), while schools at two standard deviations above the mean in percentage of teachers who regularly use technology for administrative purposes experienced the grea test decreases (0.75%) (see Figure 29).

PAGE 235

221 Relationship between Percent of Teachers that Regularly Use Technology for Administrative Purposes and Student Absences in Elementary Schools0 2 4 6 8 10 12 14 16 18 20 200320042005 YearPercent of Students with 21+ Days Absent Elementary + 2SD Elementary + 1SD Elementary mean Elementary 1SD Elementary 2SD Figure 29. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Student Absences in Elementary Schools. When the sample was restricted to just elementary and middle schools and percent of gifted students was included in the equation, there was a main effect with gifted but no interactions of percent of gifted students in the school with time or time2. Thus, when all other factors were held equal, schools with highest percentages of gifted students began the study with the lowest percent of students absent more than 21 days and this trend did not chan ge over time (see Figure 30). However, there were differences by school level. Between 2003-04 and 2005-06, first, the trend fo r all elementary schools with gifted students was an increase in percent of students absent more than 21 days, which then leveled out, while the trend for middle schools with gifted students was a decrease in the percent of students absent more than 21 days and then an increase. Both elementary and mi ddle schools ended the study in 2005-06 at approximately the same percent.

PAGE 236

222 Relationship between Percent of Gifted Students on Student Absences by School Level (Gifted Included)0 2 4 6 8 10 12 14 200320042005 YearPercent of Students with 21+ Days Absent Elementary + 2SD Elementary + 1SD Elementary mean Elementary 1SD Elementary 2SD Middle Schools + 2SD Middle Schools + 1SD Middle Schools mean Middle Schools 1SD Middle Schools 2SD Figure 30. Relationship between Percent of Gifted Students on Student Absences by School Level (Gifted Included) When examining the parameter estimates of the technology integration indicators within these data, there was a significant relationship between the in tercept of the percent of teachers who regularly use technology for administrative purposes and percent of students absent more than 21 days. Interactions between time and time2 and the percent of teachers who regularly use technology for administrative purposes were significant predictors of percent of students absent more than 21 days. In order to visualize the significant relationships between the percent that teachers who regularly use technology for administrative purposes and percent of students absent more than 21 days, separate charts were created after controlling for all other factors. Each school level was examined separately. One and two standard deviations above the mean, the mean, and one and two standard devi ations below the mean of percenta ges of teachers who regularly use technology for administrative purposes were compared after controlling for all other factors. In 2003-04, middle schools with the highest percentages of teach ers who regularly use tech nology for administrative

PAGE 237

223 purposes started with the lowest percent of students absent more than 21 days (9.68%), while middle schools with the lowest percentages of teachers who regularly use technology for administrative purposes started with the highest percent of students absent more than 21 days (10.50%) (see Figure 31). The level of percent of students absent more than 21 days decreased in middle schools at one and two standard deviations below the mean in percentages of teach ers who regularly use technology for administrative purposes in 2004-05 (0.14% and 0.47%, respectively); however the middle schools that were at two standard deviations above the mean in percentage of teachers who regularly used technology for administrative purposes had the most increase in percent of students absent more than 21 days (0.85%). In 2005-06, all middle schools experienced increases in their percent of students absent more than 21 days, with middle schools with the highest percentages of teachers who regularly used technology for administrative purposes experiencing the least increase in per cent of students absent more than 21 days (0.07%). At the end of the study in 2005-06, the differences in the percentage of students absent more than 21 days related to the different levels in percentage of teachers who regularly used technology for administrative purposes was 0.01%. Although the curvilin ear trends were significant because there were so many schools in the sample, the difference had modest practical importance.

PAGE 238

224 Relationship between Percent of Teachers that Regularly Use Technology for Administrative Purposes and Student Absences in Middle Schools0 2 4 6 8 10 12 14 200320042005 YearPercent of Students with 21+ Days Absent Middle Schools + 2SD Middle Schools + 1SD Middle Schools mean Middle Schools 1SD Middle Schools 2SD Figure 31. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Student Absences in Middle Schools (Gifted Included). The trends for elementary schools followed a very similar pattern to that of middle schools (see Figure 32). In 2003-04, elementary schools with two standard deviations above th e mean of percentage of teachers who regularly use technology for administrativ e purposes had the lowest percentage of students absent more than 21 days (5.90%), while elementary schools with two standard deviations below the mean of percentage of teachers who re gularly use technology for admini strative purposes had the highest percentage of students absent more than 21 days (6.72%). Although this difference was significant, the difference of 0.82% is very modest. In 2004-05 the percent of students absent more than 21 days increased in all elementary schools. However, elementary schools with the least or two standard deviations below the mean for percentage of teachers who regularly use te chnology for administrative purposes increased the least (0.78%), while elementary schools with the highest percentage of teachers who regularly use technology for administrative purposes increased the mo st (2.10%). The trend for elementary schools with all levels of percentage of teachers who regularly use technology for administrative purposes reversed

PAGE 239

225 again in 2005-06. Elementary schools at two standard deviations ab ove the mean of percentage of teachers who regularly use technology for administrative purpos es had the greatest decreases (0.60%) in percent of students absent more than 21 days, while elementary schools with two standard deviations below the mean of percentage of teachers who re gularly use technology for admini strative purposes experienced the smallest decreases (0.10%) in average school percent of students abse nt more than 21 days (see Figure 32). Relationship between Percent of Teachers that Regularly Use Technology for Administrative Purposes and Student Absences in Elementary Schools0 2 4 6 8 10 12 14 200320042005 YearPercent of Students with 21+ Days Absent Elementary + 2SD Elementary + 1SD Elementary mean Elementary 1SD Elementary 2SD Figure 32. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Absences in Elemen tary Schools (Gifted Included). The significant relationship between the interaction of time2 and the level of technology support – human and the percent of students absent more than 21 days for all school levels without gifted are depicted for each level of school separately. Charts were made for each level of school to visualize the relationship between the level of technology support – human and percent of students absent more than 21 days at one and two standard deviations above the mean, the mean, and one and two standard deviations below the mean. Although the intercept for level of technology support – human were not significant,

PAGE 240

226 interactions between time and time2 and level of technology support – human with percent of students absent more than 21 days were si gnificant. This resulted in curvilinear trends at each grade level. The trends for high school level at one and two standard deviations abov e the mean, the mean, and one and two standard deviations below the mean of level of technology support – human were examined (see Figure 33). When controlling for all other variables, high schools that began the study at two standard deviations above the mean have lower percentage of students absent more than 21 days (13.91%) than high schools at two standard deviations below the mean (14.05%). In the 2004-05 school year high schools at two standard deviations below the mean of level of technology support – human experienced a decline in the percent of students absent more than 21 days (0.03%), while schools with two standard deviations above the mean of level of technology support – huma n experienced the greatest increase in the percent of students absent more than 21 days (0.47%). At the end of the study in 2005-06, all high schools experienced an increase in percent of students absent more than 21 days, with schools at two standard deviations above the mean in level of technology support – human having the least gain in percent of students absent more than 21 days (0.95%), and schools with two standard deviations below the mean having the greatest gains in percent of students absent more than 21 days (2.17%). The intercept for level of technology support – human was not significant. The significant interactions between time and time2 and level of technology support – human resulted in changes in the relationship between level technology support – human and the percent of students absent more than 21 days. More time is needed to examine the directions of these trends.

PAGE 241

227 Relationship between Technical Support Human and Student Absences in High Schools0 2 4 6 8 10 12 14 16 18 20 200320042005 YearPercent of Students with 21+ Days Absent High Schools + 2SD High Schools + 1SD High Schools mean High Schools 1SD High Schools 2SD Figure 33. Relationship between Technology Support – Human and Student Absences in High Schools. When controlling for all other variables, middle schools at two standard deviations above the mean a for level of technical support human started the study in 2003-04 with 9.92% of students absent more than 21 days, while schools at two standard deviations below the mean started the study with 10.06% (see Figure 34). As with high schools, in 2004-05 middle schools at two standard deviations below the mean of level of technical support human experienced decreases (0.03%) in the percent of students absent more than 21 days, while middle schools at two standa rd deviations above the mean a for level of technical support – human experienced the greatest increase (0.4 7%). At the end of the study in 2005-06, middle schools at two standard deviations below the mean experienced the greatest increases (0.84%) in the percent of students absent more than 21 days, while middle schools at two standard deviations above the mean for level of technical support – huma n experienced the greatest decrease (0.37%).

PAGE 242

228 Relationship between Technical Support Human and Student Absences in Middle Schools0 2 4 6 8 10 12 14 16 18 20 200320042005 YearPercent of Students with 21+ Days Absent Middle Schools + 2SD Middle Schools + 1SD Middle Schools mean Middle Schools 1SD Middle Schools 2SD Figure 34. Relationship between Technology Support – Human and Student Absences in Middle Schools. After controlling for all other variables, elementary schools with two standard deviations above the mean for level of technology support human began with 6.38% of students absent more than 21 days in 2003-04, while elementary schools with two standard deviations below the mean began with 6.52% of students absent more than 21 days (see Figure 35). A ll elementary schools experienced increases in percent of students absent more than 21 days in the 2004-05 school year, with elementary schools with two standard deviations above the mean of level of technology support – human having the greatest increase (1.56%) and elementary schools with two standard deviations below the mean with the least increase (1.06%). By the end of the study in 2005-06, the trends for elementary schools were different based on their level of technology support – human. Elementary schools at the two standard deviations above the mean experienced the greatest decrease in the percent of st udents absent more than 21 days (0.96%), while elementary schools at two standard deviations below the mean in level of technology support – human experienced the greatest increase in the percent of stude nts absent more than 21 days (0.25%). More time is needed to examine this relationship and the directions of trend.

PAGE 243

229 Relationship between Technical Support Human and Student Absences in Elementary Schools0 2 4 6 8 10 12 14 16 18 20 200320042005 YearPercent of Students with 21+ Days Absent Elementary + 2SD Elementary + 1SD Elementary mean Elementary 1SD Elementary 2SD Figure 35. Relationship between Technology Support – Human and Absences in Elementary Schools. Hypothesis 2 The second analysis conducted to answer th e second research question used the student misconduct outcome data to test the following hypothesis: H2: After controlling for school level (elementary, middle, and high), school socio-economic status, minority, limited English proficiency, students with disabilities, gifted, teacher qualification, and learning environment quality, mean school student misconduct composite variable will have a negative relationship with indicators of technology integration. The first step was to build the unconditional model. The unconditional model predicted the schools’ student misconduct composite from the average of student misconduct composite for all schools. There were no other predictors. The average student misconduct for all schools was 19.99 points ( t (2311) = 45.84, p <.0001). Model 1: Unconditional Model Level 1: Student Misconduct = 0 + r Level 2: 0 = 00 + u0 Mixed-Effects Model: Student Misconduct = 00 + u0 + r

PAGE 244

230 The intraclass correlation coeffici ent (ICC) was computed to determine the proportion of variance in the student misconduct variable that is accounted for by the schools. The ICC was .86, which is high and supports using multi-level modeling for the analysis. The model fit statistics from this model were used as the baseline for model comparisons. The next step added time to the predictor equation (see Model 2a). The variance components from this analysis showed how much of the variance in the model was accounted for by time. The variance in the slopes between schools was significant. Therefore, tim e was set as a random effect, and the model was estimated. Both the intercept ( t (2311) = 43.89, p <.0001) and time ( t (2311) = -9.63, p <.0001) were significant parameters. Although time added add itional explained variance between schools, time accounted for 36% of the va riance within schools. Model 2a: Unconditional Growth Model Level 1: Student Misconduct = 0 + 1*Time + r Level 2: 0 = 00 + u0 1 = 10 + u1 Mixed-Effects Model: Student Misconduct = 00 + 10*Time + u0 + u1*Time + r To determine if the equation was not linear but curvilinear, time2 was added to the equation so the variance could be compared. Results indicated that time2 was significant ( t (2311) = 3.19, p = 0.0014), but it did not increase the variance explained over the Growth Model (see Model 2b). Time and time2 were retained in the quadratic growth model equation. Model 2b: Quadratic Growth Model Level 1: Student Misconduct = 0 + 1*Time + 2* Time2 + r Level 2: 0 = 00 + u0 1 = 10 + u1 2 = 20 Mixed-Effects Model: Student Misconduct = 00 + 10*Time + 20* Time2 + u0 + u1 + r Next, school level was added to the Quadratic Growth Model to predict misconduct (See Model 3). The significance of the parameter estimates determined if school level was significantly related to the student misconduct and if there was an interaction with time. This model adjusted the mean school student misconduct and the slope of student misconduct growth for school level. The parameter estimates of elementary school relative to middle school and time at the intercept were significant, while time2 was not significant at the intercept. Neith er the interactions between time nor time2 with elementary or high school relative to middle school were significant. All model fit indices indicated improved fit with this model (see

PAGE 245

231 Table 65). This model accounted for 36% of the between school varian ce but no additiona l within school variance from the Quadratic Growth Model. Model 3: School Level as Predictor Level 1: Student Misconduct = 0 + 1*Time + 2*Time2 + r Level 2: 0 = 00 + 01*School Level + u0 1 = 10 + 11*School Type + u1 2 = 20 + 21*School Type Mixed-Effects Model: Student Misconduct = 00 + 01*School Type + 10*Time + 11*School Type*Time + 20*Time2 + 21*School Type*Time2 + u0 + u1 + r Table 54. Model 3: Time, Time Squared, and School Type as Predictors of Misconduct Effect School Level Estimate SE df t p Intercept 43.5557 0.8274 2309 52.64 <.0001 ** Time -3.0542 0.8239 2309 -3.71 0.0002 ** Time2 0.1687 0.3816 2309 0.44 0.6585 School Level Elementary -33.3814 0.9412 2309 -35.47 <.0001 ** School Level High -0.9679 1.2488 2309 -0.78 0.4384 School Level Middle 0 . . Time*School Level Elementary 1.0952 0.9372 2309 1.17 0.2427 Time*School Level High -0.452 1.2435 2309 -0.36 0.7162 Time*School Level Middle 0 . . Time2*School Level Elementary 0.558 0.434 2309 1.29 0.1987 Time2*School Level High 0.00046 0.5759 2309 0 0.9994 Time2*School Level Middle 0 . . Covariance Parameter Estimate SE z p 262.07 8.8328 29.67 <.0001 ** -35.6545 2.7148 -13.13 <.0001 ** 21.3539 1.4158 15.08 <.0001 ** Residual 43.289 1.2732 34 <.0001 ** Note: p < .05; ** p < .01 The next model added student demographic variables to the School Level Model. This model was estimated twice. The first time, the model was run w ith high school as a school level and all of the demographic variables except gifted, because gifted is not a designation at the high school level (see Model 4a). The second time, the data were filtered to exclude high school as a school level and kept the gifted variable with middle and elementary schools (see Mo del 4b). The model fit statistics of the demographic model with all three school levels was compared with the School Level as Predictor Model to determine if

PAGE 246

232 there was a better fit (see Table 65). The significan ce of the parameter estimates determined which of the demographic variables remained in the predictor eq uation (see Table 55). The variance estimates showed the amount of the total variance that was accounted for by each model. When all of the demographics variables except gifted were added to the model (s ee Model 4a), the intercept was significant and the average middle school started with FC AT misconduct score of 43.86 ( t (2259) = 57.82, p <.0001). The parameter estimates for time, elementary school relative to middle school, free or reduced lunch status, minority, LEP, and students with disabilities were significant, while the parameter estimate for time2, and high school relative to middle school were not signif icant. There were no in teractions with time or time2 with any of the demographic variables or school leve l. All model fit indices indicated better fit with the addition of these demographics variables. Adding the demographics variables with school level explained 54% of the between school variance and 36% of the within school variance for a total of 51% of all variance explained. Model 4a: Demographics by School Level (including High School and no Gifted) Level 1: Student Misconduct = 0 + 1*Time + 2*Time2 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + u0 1 = 10 + 11*School Type + 12*SES + 13*Minority + 14* LEP + 15* SWD + u1 2 = 20 + 21*School Type + 22*SES + 23*Minority + 24* LEP + 25* SWD Mixed-Effects Model: Student Misconduct = 00 + 01*School Level + 02*SES + 03*Minority + 04*SWD + 05*LEP + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14*SWD*Time + 15*LEP*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24*SWD*Time2 + 25*LEP*Time2 + u0 + u1 + r Table 55. Model 4a: Misconduct predicted by Time, Scho ol Type, and Demographics Variables (No Gifted) Effect School Level Estimate SE df t p Intercept 43.8553 0.7584 2259 57.82 <.0001 ** Time -3.4066 0.8683 2203 -3.92 <.0001 ** Time2 0.1807 0.3991 2092 0.45 0.6508 School Level Elementary -34.0721 0.8717 2092 -39.09 <.0001 ** School Level High 2.2263 1.1424 2092 1.95 0.0514 School Level Middle 0 . . Free Reduced Lunch 2.6192 0.4238 2092 6.18 <.0001 ** Minority 6.3022 0.4837 2092 13.03 <.0001 ** LEP -2.9498 0.4213 2092 -7 <.0001 ** Students with Disabilities 2.8979 0.3263 2092 8.88 <.0001 ** Time*School Level Elementary 1.4968 1.0041 2092 1.49 0.1362 Time*School Level High -0.9885 1.3128 2092 -0.75 0.4516

PAGE 247

233 Effect School Level Estimate SE df t p Time*School Level Middle 0 . . Time*Free Reduced Lunch 0.0205 0.5952 2092 0.03 0.9725 Time*Minority -0.9399 0.6133 2092 -1.53 0.1255 Time*LEP 0.2539 0.5101 2092 0.5 0.6187 Time*Students with Disabilities -0.00911 0.4336 2092 -0.02 0.9832 Time2*School Level Elementary 0.4451 0.462 2092 0.96 0.3354 Time2*School Level High 0.1533 0.604 2092 0.25 0.7996 Time2*School Level Middle 0 . . Time2*Free Reduced Lunch 0.2171 0.2744 2092 0.79 0.4289 Time2*Minority 0.1091 0.2809 2092 0.39 0.6978 Time2*LEP -0.2954 0.2321 2092 -1.27 0.2033 Time2*Students with Disabilities -0.2511 0.1983 2092 -1.27 0.2056 Covariance Parameter Estimate SE z p 193.62 7.1121 27.22 <.0001 ** -33.592 2.609 -12.88 <.0001 ** 22.3308 1.5287 14.61 <.0001 ** Residual 43.6012 1.3449 32.42 <.0001 ** Note: p < .05; ** p < .01 The results from the analysis in Model 4b indicated that the intercept, time, school level, free or reduced lunch status, minority, LEP, students with di sabilities, and gifted were all significant, while time2 was not significant (see Table 56). Interactions betw een time and elementary school level, free or reduced lunch status, LEP, students with disabilities, and gifted were significant. Interactions between time and elementary relative to middle school and gift ed were significant. Interactions between time2 and gifted were also significant. Because the parame ter for gifted was significant in this model, an unconditional model using the same population with high schools filtered out, predicting FCAT misconduct with average FCAT misconduct was estimated in order to compare the fit of this model. All model fit statistics indicated better model fit (see Table 66). When examining the variance of misconduct in elementary and middle schools, adding demographics variables to the equation explained 56% of the between school variance and 42% more of the within school variance. Tw o sets of analyses were conducted on the rest of the models in order to examine the relationship of gifted with technolo gy integration as one of the predictors of school achievement.

PAGE 248

234 Model 4b: Demographics by School Level (Elementary and Middle School only) Level 1: Student Misconduct = 0 + 1*Time + 2*Time2 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + u0 1 = 10 + 11*School Type + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16*Gifted + u1 2 = 20 + 21*School Type + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26*Gifted Mixed-Effects Model: Student Misconduct = 00 + 01*School Level + 02*SES + 03*Minority + 04*SWD + 05*LEP + 06*Gifted + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP*Time + 15* SWD*Time + 16*Gifted*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Gifted*Time2 + u0 + u1 + r Table 56. Model 4b: Misconduct predicted by Time, School L evel, and Demographics Variab les for Elementary and Middle Schools with Gifted Effect School Level Estimate SE df t p Intercept 43.7317 0.6991 1825 62.56 <.0001 ** Time -3.9217 0.8124 1704 -4.83 <.0001 ** Time2 0.4931 0.3723 1537 1.32 0.1855 School Level Elementary -34.0648 0.8109 1537 -42.01 <.0001 ** School Level Middle 0 . . Free Reduced Lunch 1.8571 0.4521 1537 4.11 <.0001 ** Minority 6.3869 0.5108 1537 12.5 <.0001 ** LEP -3.2139 0.4094 1537 -7.85 <.0001 ** Students with Disabilities 1.8151 0.3301 1537 5.5 <.0001 ** Gifted -2.1716 0.357 1537 -6.08 <.0001 ** Time*School Level Elementary 2.1908 0.9516 1537 2.3 0.0215 Time*School Level Middle 0 . . Time*Free Reduced Lunch 0.3796 0.6517 1537 0.58 0.5604 Time*Minority -0.5206 0.6686 1537 -0.78 0.4363 Time*LEP 0.05544 0.5164 1537 0.11 0.9145 Time*Students with Disabilities 0.473 0.4457 1537 1.06 0.2888 Time*Gifted 1.182 0.4631 1537 2.55 0.0108 Time2*School Level Elementary 0.08988 0.4365 1537 0.21 0.8369 Time2*School Level Middle 0 . . Time2*Free Reduced Lunch 0.2056 0.2992 1537 0.69 0.4921 Time2*Minority -0.1458 0.3047 1537 -0.48 0.6323 Time2*LEP -0.2263 0.2346 1537 -0.96 0.3348 Time2*Students with Disabilities -0.4301 0.2031 1537 -2.12 0.0344 Time2*Gifted -0.5064 0.2108 1537 -2.4 0.0164

PAGE 249

235 Covariance Parameter Estimate SE z p 157.77 6.5899 23.94 <.0001 ** -26.7708 2.5116 -10.66 <.0001 ** 21.4441 1.5534 13.81 <.0001 ** Residual 35.0381 1.2616 27.77 <.0001 ** Note: p < .05; ** p < .01 The next model added the variable that measures the school learning envi ronment factors to the Demographics Model by School Level Model. These in cluded teacher qualifica tions and positive learning environment. This model was estimated twice, first without gifted population but all school levels (see model 5a) and then with elementary and middle school levels and gifted population (see model 5b). When school learning environment factors were added with the demographic and school level variables for all school levels, the parameter estimates for the intercept, elementary relative to middle school, free or reduced lunch status, minority, LEP, students with disa bilities, positive learning environment, and teacher qualifications were significant, while time, time2, and high school relative to middle school, were not significant (see Table 57). There were si gnificant interactions between time and time2 with positive learning environment and time and po sitive teacher qualifications. Time2 also had a significant interaction with students with disabilities. Adding the student learning environment variables explained an additional 3% of the between school variance and explained an additional 1% of the within school variance for a total of 55% of all of the variance explained. All of the model fit indices indicated that this model fit of the data better (see Table 65). Model 5a: School Learning Environment with Demographics by School Level (all school levels without gifted and LEP) Level 1: Student Misconduct = 0 + 1*Time + 2*Time2 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04*SWD + 05* Teacher Qualifications + 06*Positive Learning Environment + u0 1 = 10 + 11*School Type + 12*SES + 13*Minority + 14*LEP + 15*SWD + 16* Teacher Qualifications + 17*Positive Learning Environment + u1 2 = 20 + 21*School Type + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26* Teacher Qualifications + 27*Positive Learning Environment Mixed-Effects Model: Student Misconduct = 00 + 01*School Level + 02*SES + 03*Minority + 04* SWD + 05*Teacher Qualifications + 06* Positive Learning Environment + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14*SWD *Time + 15* Teacher Qualifications*Time + 16* Positive Learning Environment*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* SWD*Time2 + 25* Teacher Qualifications*Time2 + 26* Positive Learning Environment*Time2 + u0 + u1 + r

PAGE 250

236 Table 57. Model 5a: Misconduct Predicted by Demographics and Student Learning Environment by School Level (All School Levels without Gifted) Effect School Level Estimate SE df t p Intercept 41.3088 0.7627 2259 54.16 <.0001 ** Time -1.5258 0.9167 2203 -1.66 0.0962 Time2 -0.3489 0.4139 2086 -0.84 0.3993 School Level Elementary -30.7665 0.8873 2086 -34.67 <.0001 ** School Level High -0.8125 1.1547 2086 -0.7 0.4817 School Level Middle 0 . . Free Reduced Lunch 1.146 0.4414 2086 2.6 0.0095 ** Minority 5.7406 0.4748 2086 12.09 <.0001 ** LEP -2.6873 0.4092 2086 -6.57 <.0001 ** Students with Disabilities 1.9912 0.3273 2086 6.08 <.0001 ** Positive Learning Environment -3.2218 0.2834 2086 -11.37 <.0001 ** Positive Teacher Qualifications -1.32 0.2679 2086 -4.93 <.0001 ** Time*School Level Elementary -1.3693 1.0813 2086 -1.27 0.2055 Time*School Level High 1.2139 1.4189 2086 0.86 0.3923 Time*School Level Middle 0 . . Time*Free Reduced Lunch 1.176 0.6614 2086 1.78 0.0756 Time*Minority -0.8265 0.6158 2086 -1.34 0.1797 Time*LEP 0.2369 0.5097 2086 0.46 0.6422 Time*Students with Disabilities 0.5555 0.4451 2086 1.25 0.2122 Time*Positive Learning Environment 1.7178 0.4827 2086 3.56 0.0004 ** Time*Positive Teacher Qualifications 0.8713 0.414 2086 2.1 0.0354 Time2*School Level Elementary 1.5555 0.4893 2086 3.18 0.0015 ** Time2*School Level High -0.9091 0.6494 2086 -1.4 0.1617 Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.4285 0.3065 2086 -1.4 0.1623 Time2*Minority 0.1947 0.2816 2086 0.69 0.4895 Time2*LEP -0.2305 0.2324 2086 -0.99 0.3213 Time2*Students with Disabilities -0.4876 0.203 2086 -2.4 0.0164 Time2*Positive Learning Environment -0.9132 0.2282 2086 -4 <.0001 ** Time2*Positive Teacher Qualifications -0.3864 0.1909 2086 -2.02 0.0431

PAGE 251

237 Covariance Parameter Estimate SE z p 177.51 6.6148 26.83 <.0001 ** -31.7691 2.5039 -12.69 <.0001 ** 22.1247 1.5067 14.68 <.0001 ** Residual 42.7769 1.3183 32.45 <.0001 ** Note: p < .05; ** p < .01 When the data were filtered to include only elementary and middle schools and gifted was also added to the equation, parameter estimates for the intercept, elementary school relative to middle school, minority, LEP, students with disabilities, gifted, positiv e learning environment, and teacher qualifications were significant, while time, time2, and free or reduced lunch status were not significant. Significant interactions with time included el ementary relative to middle school, minority, teacher qualifications, and positive learning environment. Significant interactio ns with time included free or reduced lunch status, students with disabilities, and positive learning en vironment. Significant interactions with time2 included elementary relative to middle school, students with di sabilities, positive learning environment, and positive teacher qualifications (see Table 58). This model dem onstrated better fit than the previous model by all model fit indices (see Table 66). It explained 3% more of the between school variance and 1% more of the within school variance than the previous model for a total 57% of all the variance. Model 5b: School Learning Environment with Demographics by School Level (Elementary and Middle Schools with Gifted) Level 1: Student Misconduct = 0 + 1*Time + 2*Time2 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + u0 1 = 10 + 11*School Type + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16*Gifted + 17* Teacher Qualifications + 18* Positive Learning Environment + u1 2 = 20 + 21*School Type + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26*Gifted + 27* Teacher Qualifications + 28* Positive Learning Environment Mixed-Effects Model: Student Misconduct = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Gifted*Time + 17* Teacher Qualifications*Time + 18* Positive Learning Environment*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Gifted*Time2 + 27* Teacher Qualifications*Time2 + 28* Positive Learning Environment*Time2 + u0 + u1 + r

PAGE 252

238 Table 58. Model 5b: Misconduct Predicted by Demographics an d Student Learning Environment by School Level for Elementary and Middle School with Gifted Effect School Level Estimate SE df t p Intercept 40.7779 0.7324 1825 55.67 <.0001 ** Time -1.1294 0.9086 1704 -1.24 0.2141 Time2 -0.3959 0.4024 1531 -0.98 0.3253 School Level Elementary -30.1026 0.8711 1531 -34.56 <.0001 ** School Level Middle 0 . . Free Reduced Lunch 0.2592 0.4753 1531 0.55 0.5856 Minority 5.9955 0.5015 1531 11.95 <.0001 ** LEP -2.8548 0.3992 1531 -7.15 <.0001 ** Students with Disabilities 1.0893 0.3302 1531 3.3 0.001 ** Gifted -1.3382 0.3565 1531 -3.75 0.0002 ** Positive Learning Environment -3.724 0.3672 1531 -10.14 <.0001 ** Positive Teacher Qualifications -0.8584 0.2759 1531 -3.11 0.0019 ** Time*School Level Elementary -1.9468 1.1016 1531 -1.77 0.0774 Time*School Level Middle 0 . . Time*Free Reduced Lunch 1.9779 0.73 1531 2.71 0.0068 ** Time*Minority -0.2972 0.6707 1531 -0.44 0.6578 Time*LEP -0.1308 0.5193 1531 -0.25 0.8012 Time*Students with Disabilities 1.1092 0.4582 1531 2.42 0.0156 time*Gifted 0.4707 0.4769 1531 0.99 0.3238 Time*Positive Learning Environment 3.2289 0.6188 1531 5.22 <.0001 ** Time*Positive Teacher Qualifications 0.8082 0.4268 1531 1.89 0.0584 Time2*School Level Elementary 1.6594 0.4877 1531 3.4 0.0007 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.562 0.3367 1531 -1.67 0.0953 Time2*Minority -0.1224 0.306 1531 -0.4 0.6893 Time2*LEP -0.1174 0.2369 1531 -0.5 0.6203 Time2*Students with Disabilities -0.6881 0.2084 1531 -3.3 0.001 ** time2*Gifted -0.2352 0.2163 1531 -1.09 0.2771 Time2*Positive Learning Environment -1.4441 0.2797 1531 -5.16 <.0001 ** Time2*Positive Teacher Qualifications -0.3868 0.1967 1531 -1.97 0.0495

PAGE 253

239 Covariance Parameter Estimate SE z p 145.88 6.1939 23.55 <.0001 ** -24.2363 2.4064 -10.07 <.0001 ** 20.6174 1.5194 13.57 <.0001 ** Residual 34.6398 1.2487 27.74 <.0001 ** Note: p < .05; ** p < .01 The next model added technology integration variables with the demographics, learning environment, and school level variables. These included student access to various types of software, teachers regularly using various types of software, frequency that students use various types of software, and technology support. This model was estimated twi ce, first without gifted population but all school levels (see model 6a) and then with elementary and middle school levels and gifted population (see model 6b). When the model was estimated with all school levels without gifted, the significant technology parameter estimates at the intercept were teachers wh o use technology to deliver instruction and teachers who use technology for administrative purposes. Teachers who use technology for administrative purposes also had a significant interaction with time (see Table 59). Other significant parameter estimates included the intercept, elementary school relative to middle school, free or reduced lunch status, minority, LEP, students with disabilities, positive learning environm ent, and positive teacher qu alifications, while time, time2, and high school relative to middle school were not significant. Significant interactions with time included positive learning environm ent and positive teacher qualifications Significant interactions with time2 included elementary relative to middle school, students with disabilities, and positive learning environment. Only one model fit index indicated that this model had better fit (see Table 65). No additional variance was explained with this model. Two technology integration indicators were retained in the final model for all school levels without gifted, percen t of teachers who use tech nology for administrative purposes and percent of teachers who us e technology to deliver instruction. Model 6a: Technology Integration with Demographics and Student Learning Environment by School Level (All School Levels without Gifted) Level 1: Student Misconduct = 0 + 1*Time + 2*Time2 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06* Teacher Qualifications + 07* Positive Learning Environment + 08*Access Content SW + 09*Access Office SW + 010*Access Ad Prod SW + 011*Teachers Use Deliver Instruction + 012*Teachers use Admin + 013*Frequency Students Use Content + 014*Frequency Students Use Tool + 015*Technical Support Human + 016*Technical Support Hardware + u0

PAGE 254

240 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16* Teacher Qualifications + 17* Positive Learning Environment + 18*Access Content SW + 19*Access Office SW + 110*Access Ad Prod SW + 111*Teachers Use Deliver Instruction + 112*Teachers use Admin + 113*Frequency Students Use Content + 114*Frequency Students Use Tool + 115*Technical Support Human + 116*Technical Support Hardware + u1 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26* Teacher Qualifications + 27* Positive Learning Environment + 28*Access Content SW + 29*Access Office SW + 210*Access Ad Prod SW + 211*Teachers Use Deliver Instruction + 212*Teachers use Admin + 213*Frequency Students Use Content + 214*Frequency Students Use Tool + 215*Technical Support Human + 216*Technical Support Hardware Mixed-Effects Model: Student Misconduct = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06* Teacher Qualifications + 07* Positive Learning Environment + 08*Access Content SW + 09*Access Office SW + 010*Access Ad Prod SW + 011*Teachers Use Deliver Instruction + 012*Teachers use Admin + 013*Frequency Students Use Content + 014*Frequency Students Use Tool + 015*Technical Support Human + 016*Technical Support Hardware + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Teacher Qualifications*Time + 17* Positive Learning Environment*Time + 18*Access Content SW*Time + 19*Access Office SW*Time + 110*Access Ad Prod SW*Time + 111*Teachers Use Deliver Instruction*Time + 112*Teachers use Admin*Time + 113*Frequency Students Use Content*Time + 114*Frequency Students Use Tool*Time + 115*Technical Support Human*Time + 116*Technical Support Hardware*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Teacher Qualifications*Time2 + 27* Positive Learning Environment*Time2 + 28*Access Content SW*Time2 + 29*Access Office SW*Time2 + 210*Access Ad Prod SW*Time2 + 211*Teachers Use Deliver Instruction*Time2 + 212*Teachers use Admin*Time2 + 213*Frequency Students Use Content*Time2 + 214*Frequency Students Use Tool*Time2 + 215*Technical Support Human*Time2 + 216*Technical Support Hardware*Time2 + u0 + u1 + r Table 59. Model 6a: Technology Integration w ith Demographics and Student Learni ng Environment by School Level (All School Levels without Gifted) Effect School Level Estimate SE df t p Intercept 41.5078 0.7731 2259 53.69 <.0001 ** Time -1.0874 0.9687 2203 -1.12 0.2617 Time2 -0.6157 0.4411 2059 -1.4 0.1629 School Level Elementary -31.0568 0.9057 2059 -34.29 <.0001 ** School Level High -0.5642 1.1584 2059 -0.49 0.6263 School Level Middle 0 . . Free Reduced Lunch 1.1219 0.4492 2059 2.5 0.0126 Minority 5.7044 0.4763 2059 11.98 <.0001 ** LEP -2.6893 0.4101 2059 -6.56 <.0001 ** Students with Disabilities 1.9047 0.3273 2059 5.82 <.0001 ** Positive Learning Environment -3.1849 0.2846 2059 -11.19 <.0001 ** Positive Teacher Qualifications -1.2999 0.2685 2059 -4.84 <.0001 ** Access Content Software -0.1843 0.2616 2059 -0.7 0.4812 Access Office Software -0.3068 0.2604 2059 -1.18 0.2388

PAGE 255

241 Effect School Level Estimate SE df t p Access Advanced Production Software -0.2019 0.266 2059 -0.76 0.4478 Teachers Use to Deliver Instruction -0.6685 0.2891 2059 -2.31 0.0209 Teachers Use for Administrative Purposes 0.598 0.2993 2059 2 0.0458 Frequency that Students Use Content Software 0.09566 0.2387 2059 0.4 0.6887 Frequency Students Use Tool-Based Software -0.4775 0.2571 2059 -1.86 0.0634 Technical Support Human -0.08317 0.2404 2059 -0.35 0.7294 Technical Support Hardware 0.1623 0.2285 2059 0.71 0.4777 Time*School Level Elementary -1.9974 1.1636 2059 -1.72 0.0862 Time*School Level High 1.3634 1.4442 2059 0.94 0.3452 Time*School Level Middle 0 . . Time*Free Reduced Lunch 1.2867 0.6779 2059 1.9 0.0578 Time*Minority -1.2016 0.6255 2059 -1.92 0.0549 Time*LEP 0.3761 0.5128 2059 0.73 0.4634 Time*Students with Disabilities 0.5712 0.4463 2059 1.28 0.2007 Time*Positive Learning Environment 1.745 0.485 2059 3.6 0.0003 ** Time*Positive Teacher Qualifications 0.8852 0.4188 2059 2.11 0.0346 Time*Access Content Software 0.6167 0.5054 2059 1.22 0.2226 Time*Access Office Software -0.297 0.5025 2059 -0.59 0.5546 Time*Access Advanced Production Software 0.09899 0.4977 2059 0.2 0.8424 Time*Teachers Use to Deliver Instruction 0.7804 0.5634 2059 1.38 0.1662 Time*Teachers Use for Administrative Purposes -1.3835 0.5736 2059 -2.41 0.016 Time*Frequency that Students Use Content Software -0.4055 0.4878 2059 -0.83 0.4059 Time*Frequency Students Use Tool-Based Software 0.1039 0.518 2059 0.2 0.8411 Time*Technical Support Human -0.5208 0.4334 2059 -1.2 0.2296 Time*Technical Support Hardware -0.05939 0.4501 2059 -0.13 0.895 Time2*School Level Elementary 1.956 0.5346 2059 3.66 0.0003 ** Time2*School Level High -1.1071 0.6622 2059 -1.67 0.0947 Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.5262 0.3152 2059 -1.67 0.0952 Time2*Minority 0.3995 0.2863 2059 1.4 0.1631 Time2*LEP -0.3015 0.234 2059 -1.29 0.1977

PAGE 256

242 Effect School Level Estimate SE df t p Time2*Students with Disabilities -0.4887 0.2038 2059 -2.4 0.0166 Time2*Positive Learning Environment -0.9631 0.2295 2059 -4.2 <.0001 ** Time2*Positive Teacher Qualifications -0.3779 0.1945 2059 -1.94 0.0522 Time2*Access Content Software -0.2741 0.2336 2059 -1.17 0.2407 Time2*Access Office Software 0.241 0.2324 2059 1.04 0.2998 Time2*Access Advanced Production Software -0.2303 0.2303 2059 -1 0.3174 Time2*Teachers Use to Deliver Instruction -0.2895 0.2621 2059 -1.1 0.2695 Time2*Teachers Use for Administrative Purposes 0.5121 0.2641 2059 1.94 0.0526 Time2*Frequency that Students Use Content Software 0.1973 0.2284 2059 0.86 0.3879 Time2*Frequency Students Use Tool-Based Software 0.1043 0.2433 2059 0.43 0.6681 Time2*Technical Support Human 0.3064 0.2003 2059 1.53 0.1262 Time2*Technical Support Hardware 0.02631 0.2081 2059 0.13 0.8994 Covariance Parameter Estimate SE z p 176.24 6.5874 26.75 <.0001 ** -31.9049 2.4956 -12.78 <.0001 ** 21.9224 1.5021 14.59 <.0001 ** Residual 42.5992 1.3169 32.35 <.0001 ** Note: p < .05; ** p < .01 Similar results were found with the elementary and middle school data with gifted. There were no significant technology parameter estimates at the intercept. Interactions of time and time2 were significant with only the percent of teachers who use technology for administrative purposes (see Table 60). Other significant parameter estimates included the intercept, elementary relative to middle school, minority, LEP, students with disabilities, gifted, positive learning en vironment, and positive t eacher qualifications, while time, time2, and free or reduced lunch status were not signi ficant. Significant interactions with time and time2 included elementary relative to middle school, students with disabilities, and positive learning environment. Two of the fit indices indicated that this model had be tter fit (see Table 66), even though adding the technology integration indicators to the model did not explain any additional variance. Teachers

PAGE 257

243 use technology for administrative purposes was the only technology integration indicator retained in the final model for the data with elementary and middle schools and gifted. Model 6b: Technology Integration with Demographics and Student Learning Environment by School Level (Elementary and Middle Schools with Gifted) Level 1: Student Misconduct = 0 + 1*Time + 2*Time2 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 09*Access Content SW + 010*Access Office SW + 011*Access Ad Prod SW + 012*Teachers Use Deliver Instruction + 013*Teachers use Admin + 014*Frequency Students Use Content + 015*Frequency Students Use Tool + 016*Technical Support Human + 017*Technical Support Hardware + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16*Gifted + 17* Teacher Qualifications + 18* Positive Learning Environment + 19*Access Content SW + 110*Access Office SW + 111*Access Ad Prod SW + 112*Teachers Use Deliver Instruction + 113*Teachers use Admin + 114*Frequency Students Use Content + 115*Frequency Students Use Tool + 116*Technical Support Human + 117*Technical Support Hardware + u1 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26*Gifted + 27* Teacher Qualifications + 28* Positive Learning Environment + 29*Access Content SW + 210*Access Office SW + 211*Access Ad Prod SW + 212*Teachers Use Deliver Instruction + 213*Teachers use Admin + 214*Frequency Students Use Content + 215*Frequency Students Use Tool + 216*Technical Support Human + 217*Technical Support Hardware Mixed-Effects Model: Student Misconduct = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 09*Access Content SW + 010*Access Office SW + 011*Access Ad Prod SW + 012*Teachers Use Deliver Instruction + 013*Teachers use Admin + 014*Frequency Students Use Content + 015*Frequency Students Use Tool + 016*Technical Support Human + 017*Technical Support Hardware + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Gifted*Time + 17* Teacher Qualifications*Time + 18* Positive Learning Environment*Time + 19*Access Content SW*Time + 110*Access Office SW*Time + 111*Access Ad Prod SW*Time + 112*Teachers Use Deliver Instruction*Time + 113*Teachers use Admin*Time + 114*Frequency Students Use Content*Time + 115*Frequency Students Use Tool*Time + 116*Technical Support Human*Time + 117*Technical Support Hardware*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Gifted*Time2 + 27* Teacher Qualifications*Time2 + 28* Positive Learning Environment*Time2 + 29*Access Content SW*Time2 + 210*Access Office SW*Time2 + 211*Access Ad Prod SW*Time2 + 212*Teachers Use Deliver Instruction*Time2 + 213*Teachers use Admin*Time2 + 214*Frequency Students Use Content*Time2 + 215*Frequency Students Use Tool*Time2 + 216*Technical Support Human*Time2 + 217*Technical Support Hardware*Time2 + u0 + u1 + r Table 60. Model 6b: Technology Integration w ith Demographics and Student Learni ng Environment by School Level for Elementary and Middle Schools with Gifted Effect School Level Estimate SE df t p Intercept 40.9119 0.7467 1825 54.79 <.0001 ** Time -0.5019 0.966 1704 -0.52 0.6034 Time2 -0.7451 0.4322 1504 -1.72 0.0849 School Level Elementary -30.3046 0.8936 1504 -33.91 <.0001 **

PAGE 258

244 Effect School Level Estimate SE df t p School Level Middle 0 . . Free Reduced Lunch 0.1894 0.4828 1504 0.39 0.6949 Minority 5.8885 0.5038 1504 11.69 <.0001 ** LEP -2.8378 0.4002 1504 -7.09 <.0001 ** Students with Disabilities 1.0123 0.3301 1504 3.07 0.0022 ** Gifted -1.2689 0.3582 1504 -3.54 0.0004 ** Positive Learning Environment -3.744 0.3695 1504 -10.13 <.0001 ** Positive Teacher Qualifications -0.8579 0.2759 1504 -3.11 0.0019 ** Access Content Software -0.02455 0.2726 1504 -0.09 0.9283 Access Office Software -0.5042 0.2608 1504 -1.93 0.0534 Access Advanced Production Software -0.2166 0.2743 1504 -0.79 0.4298 Teachers Use to Deliver Instruction -0.4688 0.2998 1504 -1.56 0.1181 Teachers Use for Administrative Purposes 0.4445 0.3157 1504 1.41 0.1594 Frequency that Students Use Content Software -0.02405 0.2496 1504 -0.1 0.9233 Frequency Students Use Tool-Based Software -0.4634 0.2624 1504 -1.77 0.0776 Technical Support Human 0.1686 0.2509 1504 0.67 0.5019 Technical Support Hardware 0.3047 0.2328 1504 1.31 0.1908 Time*School Level Elementary -2.7863 1.1882 1504 -2.34 0.0192 Time*School Level Middle 0 . . Time*Free Reduced Lunch 2.0013 0.7483 1504 2.67 0.0076 ** Time*Minority -0.6287 0.6795 1504 -0.93 0.355 Time*LEP 0.01751 0.5206 1504 0.03 0.9732 Time*Students with Disabilities 1.1344 0.4585 1504 2.47 0.0135 Time*Gifted 0.3969 0.483 1504 0.82 0.4114 Time*Positive Learning Environment 3.3401 0.6213 1504 5.38 <.0001 ** Time*Positive Teacher Qualifications 0.7822 0.4303 1504 1.82 0.0693 Time*Access Content Software 0.5017 0.5223 1504 0.96 0.3369 Time*Access Office Software 0.3643 0.507 1504 0.72 0.4726 Time*Access Advanced Production Software 0.296 0.5038 1504 0.59 0.5569 Time*Teachers Use to Deliver Instruction 0.421 0.5866 1504 0.72 0.473 Time*Teachers Use for Administrative Purposes -1.7077 0.6079 1504 -2.81 0.005 ** Time*Frequency that Students Use Content Software -0.07635 0.5173 1504 -0.15 0.8827 Time*Frequency Students -0.1351 0.5323 1504 -0.25 0.7996

PAGE 259

245 Effect School Level Estimate SE df t p Use Tool-Based Software Time*Technical Support Human -0.5604 0.4531 1504 -1.24 0.2164 Time*Technical Support Hardware -0.03409 0.4586 1504 -0.07 0.9407 Time2*School Level Elementary 2.135 0.5358 1504 3.98 <.0001 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.592 0.3463 1504 -1.71 0.0876 Time2*Minority 0.08395 0.3102 1504 0.27 0.7867 Time2*LEP -0.2018 0.2376 1504 -0.85 0.396 Time2*Students with Disabilities -0.6962 0.2089 1504 -3.33 0.0009 ** Time2*Gifted -0.2185 0.2191 1504 -1 0.3187 Time2*Positive Learning Environment -1.5289 0.2814 1504 -5.43 <.0001 ** Time2*Positive Teacher Qualifications -0.3421 0.2 1504 -1.71 0.0874 Time2*Access Content Software -0.1763 0.2416 1504 -0.73 0.4657 Time2*Access Office Software -0.07092 0.2348 1504 -0.3 0.7627 Time2*Access Advanced Production Software -0.3421 0.2327 1504 -1.47 0.1417 Time2*Teachers Use to Deliver Instruction -0.181 0.2736 1504 -0.66 0.5083 Time2*Teachers Use for Administrative Purposes 0.7687 0.2791 1504 2.75 0.006 ** Time2*Frequency that students use content software -0.01618 0.2433 1504 -0.07 0.947 Time2*Frequency Students Use Tool-Based Software 0.2094 0.2513 1504 0.83 0.4047 Time2*Technical Support Human 0.2312 0.2092 1504 1.11 0.2692 Time2*Technical Support Hardware 0.03214 0.2126 1504 0.15 0.8799 Covariance Parameter Estimate SE z p 145.11 6.1622 23.55 <.0001 ** -24.4972 2.3968 -10.22 <.0001 ** 20.4403 1.5125 13.51 <.0001 ** Residual 34.3042 1.2416 27.63 <.0001 ** Note: p < .05; ** p < .01 The last models estimated in order to answer the second hypothesis included all school levels, demographic, student learning environment, and significant technology integration variables. These models were different because the model fit to the data fo r all schools levels without gifted included two

PAGE 260

246 technology integration variables percent of teachers who regularly use technology for administrative purposes and the percent of teachers who regularly us e technology to deliver instruction (see model 7a); while the model fitted to the data with elementary and middle school levels and gifted included only one technology integration variable – th e percent of teachers who regularly use technology for administrative purposes (see model 7b). For the model with all schools levels and no gifted, the same parameter estimates and interactions identified in the previous models as significant were significant again (see Table 61). Although, there was no diff erence in the percentage of variance expl ained in this model than was in the Demographic Model with Student Learning Environment by school level or the Technology Integration with Demographic and Student Learning Environment Model by school level, three of the model fit indices indicated better model fit (see Tabl e 65). The level-1 residuals for the final model for predicting student misconduct using all school levels without gifted ranged between -38.39 and 55.67 with a standard deviation of 4.74. There were outliers, skewness was 0.83, and kurtosis was 13.89, which would indicate that the residuals were not normally distributed. Di stribution of the empirical bayes intercepts ranged between -49.69 and 107.64 with standard deviation of 12.26. Skewness was 1.69, and kurtosis was 8.31, which indicated that the intercept residuals at leve l-2 were not normally distributed. Distribution of the empirical bayes slopes ranged between -31.19 and 18.15 with standard deviation of 3.31. Skewness was -1.91, and kurtosis was 13.02, which indicated that most of the slope residuals at level-2 were not normally distributed. Because the residuals for student conduct outcome were not normally distributed, the results of the analysis may be biased. Final Model 7a: Significant Technology Integration Indicators with Demographics and Student Learning Environment by School Level (All School Levels without Gifted) Level 1: Student Misconduct = 0 + 1*Time + 2*Time2 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Teacher Qualifications + 07* Positive Learning Environment + 08*Teachers use Admin + 09*Technical Support Hardware + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16* Teacher Qualifications + 17* Positive Learning Environment + 18* Teachers use Admin + 19* Technical Support Hardware + u1 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26* Teacher Qualifications + 27* Positive Learning Environment + 28*Teachers use Admin + 29* Technical Support Hardware

PAGE 261

247 Mixed-Effects Model: Student Misconduct = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06* Teacher Qualifications + 07* Positive Learning Environment + 08*Teachers use Admin + 09*Technical Support Hardware + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD*Time + 16*Teacher Qualifications*Time + 17* Positive Learning Environment*Time + 18* Teachers use Admin*Time + 19* Technical Support Hardware*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Teacher Qualifications*Time2 + 27* Positive Learning Environment*Time2 + 28*Teachers use Admin*Time2 + 29* Technical Support Hardware*Time2 + u0 + u1 + r Table 61. Final Model 7a: Significant Technology Integration with Demographics and Student Learning Environment by School Level (A ll School Levels without Gifted) Effect School Level Estimate SE df t p Intercept 41.4262 0.7646 2259 54.18 <.0001 ** Time -1.3915 0.9284 2203 -1.5 0.1341 Time2 -0.4009 0.4201 2080 -0.95 0.3401 School Level Elementary -30.9249 0.8914 2080 -34.69 <.0001 ** School Level High -0.6513 1.1539 2080 -0.56 0.5725 School Level Middle 0 . . Free Reduced Lunch 1.2054 0.4448 2080 2.71 0.0068 ** Minority 5.7442 0.4752 2080 12.09 <.0001 ** LEP -2.7024 0.4084 2080 -6.62 <.0001 ** Students with Disabilities 1.9303 0.3272 2080 5.9 <.0001 ** Positive Learning Environment -3.1923 0.2843 2080 -11.23 <.0001 ** Positive Teacher Qualifications -1.2948 0.2679 2080 -4.83 <.0001 ** Teachers Use to Deliver Instruction -0.828 0.2735 2080 -3.03 0.0025 ** Teachers Use for Administrative Purposes 0.4156 0.2829 2080 1.47 0.142 Time*School Level Elementary -1.5555 1.1001 2080 -1.41 0.1575 Time*School Level High 0.964 1.423 2080 0.68 0.4982 Time*School Level Middle 0 . . Time*Free Reduced Lunch 1.1075 0.6656 2080 1.66 0.0963 Time*Minority -1.1147 0.6224 2080 -1.79 0.0735 Time*LEP 0.3164 0.5105 2080 0.62 0.5355 Time*Students with Disabilities 0.5528 0.4459 2080 1.24 0.2152 Time*Positive Learning Environment 1.666 0.484 2080 3.44 0.0006 ** Time*Positive Teacher Qualifications 0.8877 0.4146 2080 2.14 0.0324 Time*Teachers Use to Deliver Instruction 0.8693 0.5255 2080 1.65 0.0982 Time*Teachers Use for Administrative Purposes -1.3314 0.5408 2080 -2.46 0.0139

PAGE 262

248 Effect School Level Estimate SE df t p Time2*School Level Elementary 1.6227 0.4995 2080 3.25 0.0012 ** Time2*School Level High -0.8353 0.6513 2080 -1.28 0.1998 Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.4171 0.3085 2080 -1.35 0.1764 Time2*Minority 0.3299 0.2852 2080 1.16 0.2476 Time2*LEP -0.263 0.2329 2080 -1.13 0.2591 Time2*Students with Disabilities -0.4782 0.2034 2080 -2.35 0.0188 Time2*Positive Learning Environment -0.8918 0.2286 2080 -3.9 <.0001 ** Time2*Positive Teacher Qualifications -0.3973 0.1913 2080 -2.08 0.0379 Time2*Teachers Use to Deliver Instruction -0.332 0.2443 2080 -1.36 0.1743 Time2*Teachers Use for Administrative Purposes 0.5328 0.2502 2080 2.13 0.0333 Covariance Parameter Estimate SE z p 176.45 6.5917 26.77 <.0001 ** -31.5923 2.4955 -12.66 <.0001 ** 21.9185 1.5046 14.57 <.0001 ** Residual 42.8402 1.3212 32.43 <.0001 ** Note: p < .05; ** p < .01 For the model with elementary and middle school levels and gifted, the same significant parameter estimate, percent of teachers who regularly use techno logy for administrative purposes, was identified as in the previous model (see Table 62). Interactions between time and time2 with percent of teachers who regularly use technology for administ rative purposes were significant. Although there was no difference in the percentage of variance explained in this model than was in the Demographic Model with Student Learning Environment by school level or the Technology Integration with Demographic and Student Learning Environment Model by school level, the AIC, AICC, and BIC indices all indicated better model fit (see Table 66). The level-1 residuals for the final model for predicting student misconduct using elementary and middle schools with gifted ranged between -37.91 and 55.27 with a standard deviation of 4.16. There were outliers. Skewness was 0.87 and kurtosis was 17.78, which would indicate that the residuals were not normally distri buted. Distribution of the empirical bayes intercepts ranged between 41.84 and 86.87 with standard deviation of 11.02. Skewness was 1.78, and kurtosis was 8.67, which indicated that the intercept residuals at level-2 were not normally distributed. Distribution of the empirical

PAGE 263

249 bayes slopes ranged between -33.53 and 16.67 with standard deviation of 3.21. Skewness was -2.13, and kurtosis was 18.24, which indicated that the slope residuals at level-2 were not normally distributed. Because the residuals for student misconduct outcome were not normally distributed, the results of the analysis may be biased. Final Model 7b: Technology Integration with Demographics and Student Learning Environment by School Level (Elementary and Middle Schools with Gifted) Level 1: Student Misconduct = 0 + 1*Time + 2*Time2 + r Level 2: 0 = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 09*Teachers use Admin + u0 1 = 10 + 11*School Level + 12*SES + 13*Minority + 14* LEP + 15* SWD + 16*Gifted + 17* Teacher Qualifications + 18* Positive Learning Environment + 19* Teachers use Admin + u1 2 = 20 + 21*School Level + 22*SES + 23*Minority + 24* LEP + 25* SWD + 26*Gifted + 27* Teacher Qualifications + 28* Positive Learning Environment + 29*Teachers use Admin Mixed-Effects Model: Student Misconduct = 00 + 01*School Level + 02*SES + 03*Minority + 04* LEP + 05* SWD + 06*Gifted + 07* Teacher Qualifications + 08* Positive Learning Environment + 09*Teachers use Admin + 10*Time + 11*School Level*Time + 12*SES*Time + 13*Minority*Time + 14* LEP *Time + 15* SWD *Time + 16*Gifted*Time + 17* Teacher Qualifications*Time + 18* Positive Learning Environment*Time + 19*Teachers use Admin*Time + 20*Time2 + 21*School Level*Time2 + 22*SES*Time2 + 23*Minority*Time2 + 24* LEP*Time2 + 25* SWD*Time2 + 26*Gifted*Time2 + 27* Teacher Qualifications*Time2 + 28* Positive Learning Environment*Time2 + 29*Teachers use Admin*Time2 + u0 + u1 + r Table 62. Final Model 7b: Significant Technology Integration with Demographics and Student Learning Environment by School Level for Elemen tary and Middle Schools with Gifted Effect School Level Estimate SE df t p Intercept 40.748 0.7365 1825 55.33 <.0001 ** Time -0.6991 0.9234 1704 -0.76 0.4491 Time2 -0.594 0.4088 1528 -1.45 0.1464 School Level Elementary -30.0672 0.8776 1528 -34.26 <.0001 ** School Level Middle 0 . . Free Reduced Lunch 0.2898 0.4779 1528 0.61 0.5443 Minority 5.9326 0.502 1528 11.82 <.0001 ** LEP -2.8511 0.3987 1528 -7.15 <.0001 ** Students with Disabilities 1.0678 0.33 1528 3.24 0.0012 ** Gifted -1.2998 0.3566 1528 -3.65 0.0003 ** Positive Learning Environment -3.7888 0.3689 1528 -10.27 <.0001 ** Positive Teacher Qualifications -0.8545 0.2759 1528 -3.1 0.002 ** Teachers Use for -0.07352 0.2539 1528 -0.29 0.7722

PAGE 264

250 Effect School Level Estimate SE df t p Administrative Purposes Time*School Level Elementary -2.5202 1.1235 1528 -2.24 0.025 Time*School Level Middle 0 . . Time*Free Reduced Lunch 1.8076 0.7328 1528 2.47 0.0137 Time*Minority -0.5006 0.6744 1528 -0.74 0.458 Time*LEP -0.05263 0.5192 1528 -0.1 0.9193 Time*Students with Disabilities 1.0508 0.458 1528 2.29 0.0219 Time*Gifted 0.3391 0.4781 1528 0.71 0.4783 Time*Positive Learning Environment 3.3045 0.6199 1528 5.33 <.0001 ** Time*Positive Teacher Qualifications 0.8379 0.4268 1528 1.96 0.0498 Time*Teachers Use for Administrative Purposes -1.1022 0.4716 1528 -2.34 0.0196 Time2*School Level Elementary 1.9232 0.4974 1528 3.87 0.0001 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.4843 0.338 1528 -1.43 0.1521 Time2*Minority -0.01672 0.3078 1528 -0.05 0.9567 Time2*LEP -0.1572 0.2368 1528 -0.66 0.507 Time2*Students with Disabilities -0.6556 0.2084 1528 -3.15 0.0017 ** Time2*Gifted -0.1762 0.2168 1528 -0.81 0.4164 Time2*Positive Learning Environment -1.47 0.2797 1528 -5.25 <.0001 ** Time2*Positive Teacher Qualifications -0.4056 0.1967 1528 -2.06 0.0394 Time2*Teachers Use for Administrative Purposes 0.5246 0.2167 1528 2.42 0.0156 Covariance Parameter Estimate SE z p 145.38 6.1696 23.56 <.0001 ** -24.2891 2.4025 -10.11 <.0001 ** 20.7305 1.5226 13.61 <.0001 ** Residual 34.5163 1.2447 27.73 <.0001 ** Note: p < .05; ** p < .01 The last step was to add in USDOE funded Magnet Schools and USDOE Technology Magnet Schools as variables in the model. Results of the m odels for both sets of data, all school levels without gifted and elementary and middle schools with gifted, indicated that having USDOE funded Magnet Schools status or U.S. technology magnet school status was not significantly related to student misconduct.

PAGE 265

251 Table 63. Model 8a: Magnet Schools with Significant Technology Integration Demographics and Student Learning Environment by School Level (A ll School Levels without Gifted) Effect School Level Estimate SE df t p Intercept 45.4078 4.0277 2257 11.27 <.0001 ** Time -11.0247 4.8535 2201 -2.27 0.0232 Time2 3.2192 2.2135 2078 1.45 0.146 School Level Elementary -30.9619 0.8909 2078 -34.75 <.0001 ** School Level High -0.7423 1.1529 2078 -0.64 0.5197 School Level Middle 0 . . Free Reduced Lunch 1.2207 0.4443 2078 2.75 0.0061 ** Minority 5.7775 0.4771 2078 12.11 <.0001 ** LEP -2.7278 0.4086 2078 -6.68 <.0001 ** Students with Disabilities 1.8823 0.3274 2078 5.75 <.0001 ** Positive Learning Environment -3.2248 0.2841 2078 -11.35 <.0001 ** Positive Teacher Qualifications -1.293 0.2675 2078 -4.83 <.0001 ** Teachers Use for Administrative Purposes -0.8418 0.2734 2078 -3.08 0.0021 ** Technical Support Human 0.413 0.2827 2078 1.46 0.1441 Not a Technology Magnet School US -8.4497 4.6607 2078 -1.81 0.07 Technology Magnet School US 0 . . Not a US Magnet School 4.5509 2.4956 2078 1.82 0.0684 US Magnet School 0 . . Time*School Level Elementary -1.5888 1.0993 2078 -1.45 0.1485 Time*School Level High 0.9782 1.4233 2078 0.69 0.492 Time*School Level Middle 0 . . Time*Free Reduced Lunch 1.1353 0.6653 2078 1.71 0.0881 Time*Minority -1.0224 0.6245 2078 -1.64 0.1018 Time*LEP 0.2697 0.5111 2078 0.53 0.5977 Time*Students with Disabilities 0.6017 0.4467 2078 1.35 0.1782 Time*Positive Learning Environment 1.6511 0.4851 2078 3.4 0.0007 ** Time*Positive Teacher Qualifications 0.8878 0.4142 2078 2.14 0.0322 Time*Teachers Use for Administrative Purposes 0.9135 0.5255 2078 1.74 0.0823 Time*Technical Support Human -1.2903 0.5409 2078 -2.39 0.0172 Time*Not a Technology Magnet School US 9.0991 5.5901 2078 1.63 0.1037 Time*Technology Magnet School US 0 . . Time*Not a US Magnet 0.6297 2.9788 2078 0.21 0.8326

PAGE 266

252 Effect School Level Estimate SE df t p School Time*US Magnet School 0 . . Time2*School Level Elementary 1.6428 0.4992 2078 3.29 0.001 ** Time2*School Level High -0.8056 0.6518 2078 -1.24 0.2166 Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.4318 0.3084 2078 -1.4 0.1616 Time2*Minority 0.2771 0.2861 2078 0.97 0.3329 Time2*LEP -0.2352 0.2332 2078 -1.01 0.3132 Time2*Students with Disabilities -0.4867 0.2038 2078 -2.39 0.017 Time2*Positive Learning Environment -0.8632 0.2295 2078 -3.76 0.0002 ** Time2*Positive Teacher Qualifications -0.3997 0.1911 2078 -2.09 0.0366 Time2*Teachers Use for Administrative Purposes -0.352 0.2444 2078 -1.44 0.1499 Time2*Technical Support Human 0.511 0.2502 2078 2.04 0.0413 Time2*Not a Technology Magnet School US -2.0751 2.5546 2078 -0.81 0.4167 Time2*Technology Magnet School US 0 . . Time2*Not a US Magnet School -1.6128 1.3705 2078 -1.18 0.2394 Time2*US Magnet School 0 . . Covariance Parameter Estimate SE z p 175.88 6.5727 26.76 <.0001 ** -31.3306 2.4864 -12.6 <.0001 ** 21.8162 1.4997 14.55 <.0001 ** Residual 42.7692 1.3186 32.43 <.0001 ** Note: p < .05; ** p < .01 Table 64. Model 8b: Magnet Schools with Significant Technology Integration Demographics and Student Learning Environment by School Level for Elemen tary and Middle Schools with Gifted Effect School Level Estimate SE df t p Intercept 44.2523 4.1917 1823 10.56 <.0001 ** Time -0.2663 5.2147 1702 -0.05 0.9593 Time2 -1.962 2.4094 1526 -0.81 0.4156 School Level Elementary -30.0632 0.8773 1526 -34.27 <.0001 ** School Level Middle 0 . . Free Reduced Lunch 0.2925 0.4776 1526 0.61 0.5403 Minority 5.9523 0.5037 1526 11.82 <.0001 ** LEP -2.8631 0.3992 1526 -7.17 <.0001 **

PAGE 267

253 Effect School Level Estimate SE df t p Students with Disabilities 1.0379 0.3304 1526 3.14 0.0017 ** Gifted -1.2995 0.3561 1526 -3.65 0.0003 ** Positive Learning Environment -3.7998 0.3687 1526 -10.31 <.0001 ** Positive Teacher Qualifications -0.851 0.2755 1526 -3.09 0.002 ** Teachers Use for Administrative Purposes -0.07138 0.2539 1526 -0.28 0.7786 Not a Technology Magnet School US -5.8317 4.8241 1526 -1.21 0.2269 Technology Magnet School US 0 . . Not a US Magnet School 2.3413 2.5132 1526 0.93 0.3517 US Magnet School 0 . . Time*School Level Elementary -2.4999 1.1226 1526 -2.23 0.0261 Time*School Level Middle 0 . . Time*Free Reduced Lunch 1.8066 0.7329 1526 2.46 0.0138 Time*Minority -0.4694 0.676 1526 -0.69 0.4875 Time*LEP -0.06392 0.5197 1526 -0.12 0.9021 Time*Students with Disabilities 1.0381 0.4585 1526 2.26 0.0237 Time*Gifted 0.3536 0.4778 1526 0.74 0.4594 Time*Positive Learning Environment 3.2449 0.6204 1526 5.23 <.0001 ** Time*Positive Teacher Qualifications 0.8449 0.4263 1526 1.98 0.0477 Time*Teachers Use for Administrative Purposes -1.0906 0.4728 1526 -2.31 0.0212 Time*Not a Technology Magnet School US -2.2918 5.9365 1526 -0.39 0.6995 Time*Technology Magnet School US 0 . . Time*Not a US Magnet School 1.8868 3.0198 1526 0.62 0.5322 Time*US Magnet School 0 . . Time2*School Level Elementary 1.907 0.497 1526 3.84 0.0001 ** Time2*School Level Middle 0 . . Time2*Free Reduced Lunch -0.4846 0.3381 1526 -1.43 0.152 Time2*Minority -0.04735 0.3083 1526 -0.15 0.878 Time2*LEP -0.1441 0.237 1526 -0.61 0.5432 Time2*Students with Disabilities -0.6333 0.2086 1526 -3.04 0.0024 ** Time2*Gifted -0.192 0.2167 1526 -0.89 0.3758 Time2*Positive Learning Environment -1.4064 0.2805 1526 -5.01 <.0001 ** Time2*Positive Teacher Qualifications -0.411 0.1965 1526 -2.09 0.0366 Time2*Teachers Use for Administrative Purposes 0.5119 0.2173 1526 2.36 0.0186 Time2*Not a Technology Magnet School US 3.9068 2.7387 1526 1.43 0.1539 Time2*Technology Magnet 0 . .

PAGE 268

254 Effect School Level Estimate SE df t p School US Time2*Not a US Magnet School -2.5746 1.3857 1526 -1.86 0.0634 Time2*US Magnet School 0 . . Covariance Parameter Estimate SE z p 144.98 6.1517 23.57 <.0001 ** -24.0726 2.3924 -10.06 <.0001 ** 20.5932 1.5153 13.59 <.0001 ** Residual 34.4372 1.2415 27.74 <.0001 ** Note: p < .05; ** p < .01 Table 65. Model Fit Indices for Models Predicting Student Misconduct Scores for All School Levels (without Gifted) Model -2 Log Likelihood AIC (smaller is better) AICC (smaller is better) BIC (smaller is better) Model 1: Misconduct Predicted by Average Misconduct of All Schools in Florida 55797.2 55803.2 55803.2 55820.4 Model 2a: Time as a Predictor of Misconduct 55271.3 55283.3 55283.3 55317.8 Quadratic Model 2b: Time2 as a Predictor of Misconduct 55261.2 55275.2 55275.2 55315.4 Model 3: Time, Time2, and School Level as Predictors of Misconduct 53722 53748 53748.1 53822.7 Model 4a: Misconduct predicted by Time, School Level, and Demographics Variables 50341.3 50391.3 50391.5 50534.4 Model 5a: Demographics and Student Learning Environment by School Level 50091.9 50153.9 50154.2 50331.4 Model 6a: Technology Integration with Demographics and Student Learning Environment by School Level 50038.4 50154.4 50155.4 50486.4 Final Model 7a: Significant Technology Integration with Demographics and Student Learning Environment by School Level 50072.6 50146.6 50147 50358.4 Model 8a: Magnet Schools with Significant Technology Integration Demographics and Student Learning Environment by School Level (All School Levels without Gifted) 50060.5 50146.5 50147.1 50392.6

PAGE 269

255 Table 66. Model Fit Indices for Models Predicting Student Misconduct Scores for Elementary and Middle School Levels (with Gifted) Model -2 Log Likelihood AIC (smaller is better) AICC (smaller is better) BIC (smaller is better) Model 1: Misconduct Predicted by Average Misconduct of All Elementary and Middle Schools in Florida 46610.5 46616.5 46616.5 46633.2 Model 4b: Misconduct predicted by Time, School Level, and Demographics Variables No High School includes gifted 38168.6 38218.6 38218.8 38356.3 Model 5b: Demographics and Teacher Qualifications by School Level 38005.9 38067.9 38068.3 38238.8 Model 6b: Technology Integration with Demographics and Teacher Qualifications by School Level 37951.5 38067.5 38068.9 38387.1 Final Model 7b: Significant Technology Integration with Demographics and Student Learning Environment by School Level 37995.1 38063.1 38063.6 38250.5 Model 8b: Magnet Schools with Significant Technology Integration Demographics and Student Learning Environment by School Level for Elementary and Middle Schools with Gifted 37982 38062 38062.7 38282.4 The result of the analysis for all the models indicated that Hypothesis 2 for Research Question 2 was partially correct. When the sample included schools at all three school levels and all other school level, demographic, and school learning environment factors were controlled, there was only a significant negative relationship between the percent of teachers wh o regularly use technology to deliver instruction and the level of student misconduct at the intercept. With the dataset with elementary and middle school with gifted, the percent of teachers who regularly use technology to deliver instruction was not significant. Also with both datasets, there were si gnificant interactions between time and time2 and the percent of teachers who regularly use technology for administrativ e purposes with school level student misconduct resulting in a curvilinear trend. After controlling so that all other variables were held at the mean, the trend for each school level could be examined separately by comparing schools with different levels of the technology indicator. The relationship between the percent of teachers who regul arly use technology to deliver instruction and the

PAGE 270

256 level of student misconduct was only significant at the in tercept. Charts were made for each level of school to visualize the relationship between the percent of teachers who regularly use technology to deliver instruction and the level of student misconduct at one and two standard deviations above the mean, the mean, and one and two standard deviations below the mean were examined. When controlling for all other variables, high schools that began the study at two standard deviations above the mean had a student misconduct score (39.12) that was 3.31 points lower than high schools at two standard deviations below the mean (42.43). Because there were no interactions with time these tr ends were parallel (see Figure 36). Relationship between Percent of Teachers that Regularly Use Technology to Deliver Instruction and Student Misconduct in High Schools5 10 15 20 25 30 35 40 45 50 200320042005 YearStudent Misconduct High Schools + 2SD High Schools + 1SD High Schools mean High Schools 1SD High Schools 2SD Figure 36. Relationship between Percent of Teachers Who Regularly Use Technology to Deliver Instruction and Student Misconduct in High Schools. When controlling for all other variables, middle schools at two standard deviations above the mean a for percent of teachers who regularly use technology to deliver instruction started with student misconduct scores at 39.77, while schools at two standard deviations below the mean started with scores at 43.08 or 3.31 points lower in 2003-04 (See Figure 37). Because time was not significant, the trends remained parallel.

PAGE 271

257 Relationship between Percent of Teachers that Regularly Use Technology to Deliver Instruction and Student Misconduct in Middle Schools5 10 15 20 25 30 35 40 45 50 200320042005 YearStudent Misconduct Middle Schools + 2SD Middle Schools + 1SD Middle Schools mean Middle Schools 1SD Middle Schools 2SD Figure 37. Relationship between Percent of Teachers Who Regularly Use Technology to Deliver Instruction and Student Misconduct in Middle Schools. When controlling for all other variables, elementary schools at two standard deviations above the mean for percent of teachers who regularly use tech nology to deliver instruction started with student misconduct scores at 8.85, while schools at two standard deviations below the mean started with scores at 12.16 or 3.31 points lower in 2003-04 (See Figure 38). Because time was not significant, the trends remained parallel.

PAGE 272

258 Relationship between Percent of Teachers that Regularly Use Technology To Deliver Instruction and Student Misconduct in Elementary Schools5 10 15 20 25 30 35 40 45 50 200320042005 YearStudent Misconduct Elementary + 2SD Elementary + 1SD Elementary mean Elementary 1SD Elementary 2SD Figure 38. Relationship between Percent of Teachers Who Regularly Use Technology to Deliver Instruction in Elementary Schools. Figure 39 illustrates the relations hip between the average school pe rcent of teachers who regularly use technology for administrative purposes and average school student misconduct score for high schools. Percent of teachers who regularly use technology for administrative purposes was compared at one and two standard deviations below the mean, the mean, and one and two standard deviations above the mean. This allowed the extreme cases of schools that had the high est percent of teachers who regularly use technology for administrative purposes, +2 standard deviations above the mean, and schools that had lowest percent of teachers who regularly use technology for administrativ e purposes, -2 standard deviations below the mean to be compared. Between 2003-04 and 2005-06, for all schools levels at all levels of teachers who regularly use technology for administrative purposes, the level of student misconduct decreased. Schools that had the highest percent of teachers who regu larly use technology for administra tive purposes started the study in 2003-04 with the highest student misconduct scores (41.61) and schools that had the lowest percent of teachers who regularly use technology for administra tive purposes had started with the lowest student

PAGE 273

259 misconduct scores (39.94). The interactions between the percent of teachers who regularly use technology for administrative purposes and time and time2 with student misconduct scores were significant, so the slopes of the trends were curvilinear. By 2005-06, high schools with the two standard deviations above the mean in percent of teachers who regularly use tech nology for administrative purposes had greatest decreases that resulted in the lowest levels of student misconduct (see Figure 39). Relationship between Percent of Teachers Who Use Technology for Administrative Purposes and Student Misconduct in High Schools5 10 15 20 25 30 35 40 45 50 200320042005 YearStudent Misconduct High Schools + 2SD High Schools + 1SD High Schools mean High Schools 1SD High Schools 2SD Figure 39. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Student Misconduct in High Schools. Middle schools had a similar beginning pattern to hi gh school, that is after controlling for all other factors, schools that were two stan dard deviations above the mean in the percent of teachers who regularly use technology for administrative purposes had the highest student misconduct scores in 2003-04 (42.26) while those with two standard deviations below the mean had the lowest levels (40.60). Schools at all levels of percent of teachers who regularly use technology for administrative purposes experienced decreases in the level of student misconduct over the course of the study. However, middle schools with two standard

PAGE 274

260 deviations above the mean of percent of teachers who regularly use technology for administrative purposes had the greatest decreases and ended the study with the lowest level of student misconduct (see Figure 40). Relationship between Percent of Teachers Who Use Technology for Administrative Purposes and Student Misconduct in Middle Schools5 10 15 20 25 30 35 40 45 50 200320042005 YearStudent Misconduct Middle Schools + 2SD Middle Schools + 1SD Middle Schools mean Middle Schools 1SD Middle Schools 2SD Figure 40. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Student Misconduct in Middle Schools. Elementary schools experienced a similar pattern to middle schools. Schools at two standard deviations above the mean in percent of teachers who regularly use technology for administrative purposes began the study with the highest student misconduct score (11.33), while schools at two standard deviations below the mean in percent of teachers who regularly use technology for administrative purposes began the study with the lowest student misconduct score (9.67). Between 2003-04 and 2004-05 elementary schools at all levels of percent of teachers who regularly use technology for administrative purposes experienced decreases in level of student misconduct; however, schools at two standard deviations above the mean experienced the greatest decrease (3.32). Between 2004-05 and 2005-06 level of student misconduct increased. Although elementa ry schools at two standard deviations above the mean in level of percent of

PAGE 275

261 teachers who regularly use technology for administra tive purposes experienced the greatest increase, the resulting level of student misconduct was still lower than at the beginning of the study (see Figure 41). Relationship between Percent of Teachers Who Use Technology for Administrative Purposes and Student Misconduct in Elementary Schools5 10 15 20 25 30 35 40 45 50 200320042005 YearStudent Misconduct Elementary + 2SD Elementary + 1SD Elementary mean Elementary 1SD Elementary 2SD Figure 41. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Student Misconduct in Elementary Schools. When the sample was restricted to just elementary and middle schools and percent of gifted students was included in the equation, there was a main effect with gifted but no interactions of percent of gifted students in the school with time or time2 and level of student misconduct. Thus, when all other factors were held equal, schools at two standard deviations above the mean in percentages of gifted students began the study with the lowest student misconduct scores and this trend did not change over time (see Figure 42). However, overall elementary schools had lower levels of student misconduct than middle schools.

PAGE 276

262 Relationship between Per cent of Gifted Students on Student Misconduct by School Level (Gifted Included)5 10 15 20 25 30 35 40 45 50200320042005 YearStudent Misconduct Elementary + 2SD Elementary + 1SD Elementary mean Elementary 1SD Elementary 2SD Middle Schools + 2SD Middle Schools + 1SD Middle Schools mean Middle Schools 1SD Middle Schools 2SD Figure 42. Relationship between Percent of Gifted Students and Level of Student Misconduct by School Level (Gifted Included). When examining the parameter estimates of the technology integration indicators within the data for elementary and middle schools with gifted, there was no significant relationship between the intercept of the percent of teachers who re gularly use technology for admi nistrative purposes and student misconduct. However, there were sign ificant interactions between time and time2 with the percent of teachers who regularly use technology for administrativ e purposes. In order to visualize the significant relationships of the percent of t eachers who regularly use technology for administrative purposes with student misconduct, after controlling fo r all other factors the trends are depicted in separate charts. Each school level was examined separately. One and two standard deviations above the mean, the mean, and one and two standard deviations below the mean of levels of percentages of teachers who regularly use technology for administ rative purposes were compared afte r controlling for all other factors. In 2003-04 when the study began, there was no significant difference between where middle schools with the highest percentages and the lowest percentages of teachers who regularly use technology for

PAGE 277

263 administrative purposes with the highest student misconduct scores (see Figure 43). The level of student misconduct decreased in all middle schools 2004-05; ho wever the middle schools that were two standard deviations above the mean in percentage of teacher s who regularly used tech nology for administrative purposes had the most decline student misconduct (2.45), while middle schools with two standard deviations below the mean in percentage of teacher s who regularly used tech nology for administrative purposes had the least decline in stude nt misconduct (0.14). In 2005-06, this rate of decline in the trend had reversed, with middle schools with the highest percen tages of teachers who regularly used technology for administrative purposes experiencing the lowest decreases in student misconduct (1.54 vs. 3.42). Over the course of the study all student misconduct at all le vels of percentages of teachers who regularly used technology for administra tive purposes decreased. Relationship between Percent of Teachers that Regularly Use Technology for Administrative Purposes and Student Misconduct in Middle Schools5 10 15 20 25 30 35 40 45 50 200320042005 YearStudent Misconduct Middle Schools + 2SD Middle Schools + 1SD Middle Schools mean Middle Schools 1SD Middle Schools 2SD Figure 43. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Student Misconduct in Middle Schools (Gifted Included).

PAGE 278

264 The trends for elementary schools followed a very similar pattern to that of middle schools (see Figure 44). In 2003-04 when the study began, there was no significant difference in level of student misconduct between elementary schools with two standard deviations above the mean and two standard deviations below the mean of percentage of teach ers who regularly use tech nology for administrative purposes. In 2004-05 the average sc hool student misconduct score decl ined in all elementary schools. However, elementary schools with the least or two standard deviations above the mean for percentage of teachers who regularly use technology for administ rative purposes declined the most (3.05), while elementary schools with the least percentage of t eachers who regularly use tech nology for administrative purposes declined the leas t (0.73). The trend for elementary schools at two standard deviations below the mean of percentage of teachers who regularly use te chnology for administrativ e purposes continued to decline at a slower rate in level of student misconduct (0.17) in 2005-06; however schools with all other levels of percentage of teachers who regularly us e technology for administrative purposes reversed directions and experienced increases in the level of student misconduct. Schools with two standard deviations above the mean of percentage of teach ers who regularly use technology for administrative purposes had the greatest increase in student misconduct (1.71).

PAGE 279

265 Relationship between Percent of Teachers that Regularly Use Technology for Administrative Purposes and Student Misconduct in Elementary Schools5 10 15 20 25 30 35 40 45 50 200320042005 YearStudent Misconduct Elementary + 2SD Elementary + 1SD Elementary mean Elementary 1SD Elementary 2SD Figure 44. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Misconduct in Elementary Schools (Gifted Included).

PAGE 280

266 Chapter 5: Discussion This chapter reviews the general limitations of this study and the difficulties encountered during all phases. Results are interpreted and compared with those found in previous studies. Each research question is addressed separately and then common themes are discussed. Finally, recommendations for future research to further investig ate these questions are proposed. Limitations It is important to review the limitations of this study before drawing conclusions. Most important to framing any recommendation from the results obtained is that this study was correlational; therefore, causality cannot be inferred. The study was conducted with data from publ ic elementary, middle, and high schools in Florida over a four year period from 2003-04 to 2006-07. The results may be specific to this set of schools, and may not be generalized to public schools in other states or to other schools in Florida. The data used were at the school level, not the student level. Percentages of groups of students within the school (e.g., free or reduced lunch, minority LEP, students with disabilities, and gifted) were used in the analysis. Student level data were not available to the public due to FERPA laws and confidentiality. Because the data used were at the school level, inferences cannot be made at the student level. This study does not inform about the relationshi p between the integration of technology and specific groups of students; it only provides information about the relationship between technology integration and the outcome variables in schools with different percentages of these different groups of students. Measurement issues are another major area of concern, which is common when using secondary data. In order to conduct this longitudinal study, items from different surveys were used. These surveys were created by the Florida Department of Education for purposes different from this study, so the locations and wordings of some items changed from year to year. The items chosen for this study may not have accurately measured the constructs th at this study was designed to examine. Availability and public release of data planned for use in this study was another issue that resulted in challenges. Some of the moderating predictor variables from the Florida Indicators Report were not

PAGE 281

267 available for the last year of the st udy (2006-07) so the values from the previous year were used to impute the missing data in order to examine the relationship between the fourth year of FCAT and technology integration data. When these indicators were used as outcome variables, the duration of the trends analyzed were shortened to three years. Percen t of students with gifted status is only reported for elementary and middle schools. Therefore, the analysis was conducted in two stages. The first stage used all public schools at elementary, middle, and high levels without the gi fted demographic variable, and the second stage used elementary and middle schools with the gifted demogr aphic variable included. In addition, the Florida Department of Education reported that they were reanalyzing the third grade FCAT Reading scores for the 2005-06 school year due to irregularities. These sc ores were deleted from the database, and the FCAT Reading scores from the other grade levels were used to obtain the mean school FCAT reading score for 2005-06. Additional measurement issues that occurred were specific to this study and may limit the validity of the findings. First, the location of computers for st udent use in either the classroom or computer lab has been found to impact the amount of time that they are used by students and students’ achievement (Adelman et al., 2002; Becker, 2001; Mann, 1999; O’Dwyer et al., 2004, 2005; Smerdon et al., 2006). Obtaining an accurate count of computers located in di fferent areas of schools was one of the first issues encountered during this study. This composite variable was going to be used to provide a measure for student access to computers, both in the regular clas sroom and in computer labs. The items, which asked for locations of desktop and laptop computers that were used to form this composite variable were interpreted by the people completing the survey so differently that, on close examination of the data, it appeared that over 100 schools used the same computers in counts for multiple locations and others may have done so also. This composite variable was not used in the analysis. In the second issue, many of the variables that were used to measure the composite variable related to support for technology were missing. Analysis of the missing data revealed that the items that measured the level of instructional and technical support in 2003-04 and 2004-05, which also had the most missing data, did not have an option to indicate there was no support. It seemed reasonable to assume that items in the STAR survey that were unanswered we re skipped because the answer was none or zero;

PAGE 282

268 therefore, all missing responses fo r technology integration indicators were changed to zero in order to retain the items in the analysis. In the third issue, several additional variables th at were theorized to have relationships with student achievement were dropped from the study. During factor analysis, several variables did not have the variability required to have adequate loadings on f actors (greater than .3) to include them in the study. As reported in previous studies (Anderson & Becker 2001; Fulton et al., 2004), the variable proportion of the technology budget spent on professional development was expected to be an important indicator for the level of support for technology integration; however, this variable did not load on the factor, so it was dropped from the study. Another variable expected to help measure the student learning environment, and have a relationship with student achievement, was class size or the ratio of students per instructional staff (Marzano, 2003; National Center for Education Statistics, 2005). This variable did not load on this factor, so it was not included in the analysis. Achievement Outcomes Research Question 1 What is the relationship between indicators of technology integration and changes in mean student achievement when controlling for school level, school socio-economic status, minority, limited English proficiency, students with disabilities, gifted, teacher qualification, and learning environment quality? The analyses conducted to answer the first res earch question examined the relationship between technology integration variables and reading, math, and writing achievement, while controlling for moderating variables. The first research question was answered by conducting multi-level models with the FCAT achievement data for reading, mathematics, and writing. In general, reading achievement as measured by FCAT reading Norm Referenced Test sc ores, math achievement as measured by FCAT math Norm Referenced Test scores, and writing achievement as measured by FCAT Writing rubric scores for all public schools in Florida had significant variability in the intercept. Over time, the slopes for reading and math achievement were significantly curvilinear and S-shaped because time, time2, and time3 were significant. However, for writing achievemen t time was not significant, although time2 and time3 were significant. First, the technology indicators that were not signi ficantly related to mean school FCAT Reading, Math, and Writing achievement were examined. The fi rst set of indicators measured student access to

PAGE 283

269 software. Student access to all types of software (i.e., software for delivery of content, office suite software, and advanced production software) was not significantly related to mean school FCAT Reading, Math, nor Writing scores. This finding counters the positive find ing of Mann et al. (1999). Other researchers have found positive relationships between increased student access to hardware and software and the frequency that students use computer s (e.g., Adelman et al. 2002; Bebell, 20 05; Becker, 2001; Lowther et al., 2003; O’Dwyer et al., 2004, 2005; Shapley et al., 2006; Silvernail & Lane, 2004). The frequency that students use software may be positively related to improved achieve ment. However, the findings from this study suggest that there is not a direct relationship or path between access and school achievement. Providing support is an essential condition for technology integration recommended by ISTE (ISTE NETS Project, 2007). Anderson and Becker (2001) and Fulton et al. (2004) also found technical support crucial to successful techno logy integration. However, two te chnology indicators that had no significant relationship with mean school FCAT Reading, Math, or Writing scores were both composite variables used to measure support when all school levels (without gifted) were examined. Neither human support from the technology integration specialist and the technical specialist nor the reliability of the hardware and Internet connections were significantly related to mean school reading, math, or writing achievement. This was an unexpected finding. Finding no relationship would suggest that technical support does not have a direct relationship or path with student achievement. Support for technology may be a mediating variable that is related to how often t eachers use technology. Anot her unexpected finding from the dataset with middle and elementary schools with gi fted was that one standard deviation increase in the level of dependability of the hardware and Internet was associated with a significant decrease of 0.28 point in the intercept of mean school FCAT Reading score. Although this had a modest effect, it beckons the researcher to investigate the negative association furthe r. It also may be that the items used to measure support did not adequately measure this construct. Another relationship that has been found to be positively related to the frequency that students use computers is the frequency that teach ers use technology (Adelman et al. 2002; Becker, 2001 ; Becker et al., 1999; Knezek et al., 2003; O’Dwyer et al., 2004, 2005) The findings from this study indicated that the proportion of teachers who regularly us e technology for delivery of instru ction was not significantly related to mean school FCAT Reading, Math, nor Writing ac hievement. One explanation for this result is the

PAGE 284

270 method used to derive the composite variable did not adequately meas ure the frequency that teachers use technology to deliver instruction. The items in the surv ey that composed the factor included more advanced production software programs used for creating instruction such as video editors and web editing programs. Although teachers may not have used advanced pro duction programs, they frequently may have used technology in less advanced ways to deliver instruction such as using technology to make presentations. Frequent use of technology in this mode only would not have been captured by the composite variable, because the composite was measured by the sum of time the teacher spent using technology in each instructional delivery method, and it was not measured by the total amount of time the teacher used technology in any method. Alternatively, these results suggest that the path between teacher use of technology for delivery of instruction and mean school achievement in Reading, Math, and Writing is not direct. Teacher use of technology for the delivery of instruction may be related to how often students use software, which may be related to achievement. Perhaps, the connection between teacher use and school mean achievement takes more time to be manifest. Nevertheless, in this study when data at all le vels of school without gifted were analyzed, the frequency that students used content delivery software was not significantly related to mean school FCAT Math achievement or mean school FCAT Reading achi evement at the intercept. This finding counters previous research findings reported about reading and achievement by Borman (2003), Kulick (2003), and Mann (1999); and math achievement reported by Borman (2003), Kulick (2003), Mann (1999), Penuel et al. (2002), and Wenglinsky (1998, 2005). However, the findings coincide with newer research results for reading reported by Dynski et al. (2007), Russell et al. (2004), and Shapley et al. (2006) and research results for math reported by Dynski et al. (2007), O’ Dwyer et al. (2005), and Shapley et al. (2006). These more recent research studies used multi-level modelin g statistical techniques an d found no significant relationship between frequency of student use of technology and reading or math achievement. Although these four studies examined relationships over a short period of time (less than one year) the current study analyzed data from four years using multi-level modelin g and found similar results. It is possible that more than four years is needed to show significant relati onships between increases in school level reading and math achievement and frequency that students use cont ent software. The positive results reported by Mann

PAGE 285

271 (1999) occurred after seven years, which would support that time may be needed for agents to adapt to change, as proposed by complexity theory. Noteworthy, with the dataset that included all sc hool levels without gifted, there were significant interactions between the frequency that students used content software and the slope of FCAT Reading with time, time2, and time3 (see Figure 45). This relationship was repeated with the dataset that included elementary and middle schools with gifted; there were significant inter actions between the frequency that students used content software and FCAT Reading with time. These interactions demonstrate that the relationship of the frequency that students used content software with FCAT Reading achievement changed over time. Although all schools began the study at the same mean school FCAT Reading score, when measured in 2004, schools at all levels that were at two standard deviations above the mean in frequency that students used content software had mean FCAT Reading scores one point lower than schools at two standard deviations below the mean. This sign ificant difference disappeared in later years. Relationship between Frequency Students Use Content Software and FCAT Reading at All School Levels without Gifted673 674 675 676 677 678 679 680 681 682 2003200420052006 YearFCAT Reading all schools + 2SD all schools + 1SD all schools mean all schools 1SD all schools 2SD Figure 45. Relationship between Frequency that Students Use Content Software and FCAT Reading at All School Levels without Gifted

PAGE 286

272 When the data with all school levels without gifted were examined, the only technology indicator that had a relationship with mean school FCAT Wr iting was the frequency that students use content software. However, this relationship was negative at the intercept, which was similar to FCAT Reading in 2004. That is, when all other factors were controlled, one standard deviation higher percentage of frequency that students used content software was associated with a decrease of 0.01 point in mean school FCAT Writing score in 2003-04. The interactions of time, time2, and time3 with the frequency that students use content software were not significant, so the tr ends were parallel over time. Although this relationship was significant, the practical importance was modest. When the dataset with elementary and middle sc hools and gifted was an alyzed, there were no significant relationships between FCAT Writing and any of the technology indicators. The ceiling effect of the test experienced by gifted students, who score at th e top range of a test, may have resulted in this lack of significant relationships. Improvement can not be detected when students score at the highest range of the test and the follow-up test is given at the same level. Furthermore, in this study when data at all levels of school without gifted were analyzed, the frequency that students used office software or advanced production software was not significantly related to mean school FCAT Writing achievement. This finding counters previous research findings reported by Goldberg et al. (2003), Kulick (2003), Lowther et al. (2003), Mann (1999), O’Dwyer et al. (2005), and Penuel et al. (2002). This lack of significant results may be due to the decreased variability in this dataset. Wr iting in this study was measured by the writing rubric with scores that ranged between 0 and 6. The FCAT Wr iting test has been updated to include other types of measurement. The data from the new FCAT Writing assessment for future studies may have greater variability and ability to measure writing achievement. In addition, the lack of significant results found with this study may be due to how the computer was used. Although O’Dwyer et al. (2005) found a significant relationship between using a computer at school to edit papers; they also found a negative relationship when using the computer at school to create presentations. In the current study, using the computer for creating presentations was included in the category for to ol-based use. Therefore, it is important for future studies to examine the specific use of technology, when examining the relationship of the integration of technology with student writing achievement.

PAGE 287

273 On a positive note, when the data with all school levels without gifted were examined, the only technology indicator that had a positive relationship at the intercept with mean school FCAT reading was the frequency that students used tool-based software. This was an encouraging relationship to find, because it supports research findings reported by Kulik (2003) O’Dwyer et al. (2005), P earson et al. (2005), and Wenglinsky (1998, 2005). In all of these studies, positive relationships were found between reading achievement and using word processors for writing. Ho wever, in this study, the effect size was modest. One standard deviation increase in frequency of student use of tool-based software was related to an increase of only 0.22 point in the mean school FCAT Re ading scale score. In addition, when the data with elementary and middle schools with gifted were exam ined, the frequency that students used tool-based software was not significant. Conversely, in the dataset with elementary and middle schools with gifted, the frequency that students used content software was significantly related to FCAT Reading achievement. One standard deviation increase in the frequency that students used content software was associated with an increase of 0.18 point in the mean school FCAT Reading score at the intercept, a modest effect size. This finding supports the research reported by Kulick (2003) and counters the no significant difference reported by Dynanski et al. (2007). These different significant results in frequency that students use content software and tool-based software may be associated with differenc es in school level. Other researchers have reported differences in how students use t echnology at the elementary and seco ndary levels (Barron et al., 2003; Hart et al., 2002; and Wenglinsky, 1998, 2005). Students at elementary levels may be using computers to learn how to read; therefore, they may be using content delivery software. Students at the secondary level may be focusing on reading to learn; therefore, they may be using technology as tools to extend and deepen their learning. Also, these differences in how st udents use technology as a tool and for content delivery may be confounding the research results. If these software uses had been combined, significant changes in the frequency that students used all types of software ma y have been detected. As a result, no information would be available to relate how students used soft ware with achievement. Thus, these conflicting results mean that the specific ways techno logy is used to support learning and deliver content are important components to examine. The statewide data used in this study provided br oad categories of software in the

PAGE 288

274 items used to measure how often students use software. It did not include specific information about which programs students used to learn content or how technology was integrated into the curriculum. The best methods in which to collect this kind of information would be through surveys of teachers and direct observations. Therefore, future research is needed to look at how technology is integrated, along with frequency of student use, in order to clarify these findings. In addition, how students are assessed may make a difference in measuring the relationship of technology use with achiev ement. Kim and Reeves (200 7) suggest that the relationship between the tool and learning is dynamic, complex, and intertwined. They recommended that students use the technology tool while being assessed. During the time of this study, the FCAT Writing has not been administered on a computer in Florida schools. Moreove r, a variety of assessment measures administered at multiple points in time are needed to measure methods of technology integration in instruction, ways that students use the technology, and achievement. Therefore, in the future, specific methods of integration of technology in each academic curriculum along with compatible and multiple forms of as sessment need to be examined in order to determine which specific methods support achievement and which methods interfere with achievement. The most interesting finding was the significant relationship between both school FCAT Reading and FCAT Math achievement and the percentage of teachers who regularly use technology for administrative purposes. With the dataset that included a ll levels of school without gifted, the percentage of teachers who regularly use technology for administ rative purposes was significantly related to school FCAT Reading at the slope, but not at the intercept, wh ich meant schools at different levels of percentage of teachers who regularly used techno logy for administrative purposes started with the same mean FCAT Reading scores (see Figure 46). However with FCAT Math achievement, this relationship was negative at the intercept (see Figure 47). That is, when all other factors were controlled, one standard deviation higher percentage of teachers who regularly use technology for administrative purpose s was associated with a decrease of 0.23 points in mean school FCAT Math score in 2003-04. Nevertheless, because time, time2, and time3 were significant with both FCAT Reading and Math, this relationship changed over time, so that by the end of the study, schools at two standard deviations above the mean in percentage of teachers who regula rly use technology for admi nistrative purposes were

PAGE 289

275 predicted to have mean FCAT Reading and FCAT Math scores two points higher than schools that were two standard deviations below the mean. With the dataset that included elementary and middle schools with gifted, there were similar sign ificant interactions between the perc entage of teachers who regularly use technology for administrative purposes and FCAT Reading at the intercept, time, and time2 and with FCAT Math at the intercept, time, time2, and time3. By the end of the study schools with two standard deviations above the mean of percentage of t eachers who regularly use technology for administrative purposes were predicted to have mean FCAT Reading scores one point higher and mean FCAT Math scores two points higher than schools with two standard deviations below the mean. Although these relationships were significant, the practical importance was modest. This suggests that the path between teacher use of technology for administrative pur poses and mean school achievement in FCAT Reading and FCAT Math may be direct. In addition, this demonstrates that time may be an important variable when examining the relationship of technology integration indicators with achievement. The changes th at occurred over time in the relationship between percent of teachers who regularly use technology for administrative purposes and student achievement suggests that more time may be needed before the relationship is fully developed or becomes established. The administrative uses that were used to create this composite variable included administrative activities such as maintaining electronic grade books, analyzing student assessment information, and communicating with parents and students by e-mail. Perhaps, teachers were able to glean more quickly important information about their students’ progress through maintaining electronic grade books and analyzing student assessment information with technology. As a result, they were able to share this valuable feedback with their students and parents. Afterward, students responded to the constructive feedback by improving their performance. As stude nts improved thei r performance teachers may have decreased their use of technology for monitoring and communicating progress. Thus, the relationship between teachers’ use of technology for administrative purposes and student achievement may be dynamic as both teachers and students adjust their strategies to match current conditions. Accordingly, it is important to examine specifically how technology is being used by both teachers and students during the dynamic process of instruction and learning.

PAGE 290

276 Relationship between Percent of Teachers Regularly Use Technology for Administrative Purposes and FCAT Reading at All School Levels without Gifted673 674 675 676 677 678 679 680 681 682 2003200420052006 YearFCAT Reading all schools + 2SD all schools + 1SD all schools mean all schools 1SD all schools 2SD Figure 46. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and FCAT Reading at All School Levels without Gifted The overall relationship during the course of the study for each of the significant technology indicator variables was an increase in mean school FC AT Reading scores, although the effect size of this increase was modest. Noteworthy, all of the figures th at illustrate the relationship between each of these variables and mean school FCAT Reading score had similar decrease between 2003-04 and 2004-05. A major event occurred in this same time period. The Florida’s A+ law provided consequences to schools for not meeting adequate yearly progress, and disaggregate d results for each school’s mean FCAT scores were made public for the first time. Schools that did not make adequate progress would have been in danger of losing funding and ultimately control of their school. In response to this event, teachers may have changed their methods of instruction and the way they integrated technology. All trends in all charts made a sharp upward turn in direction after 2004-05. After the fi rst disaggregated results were published and educators could evaluate the results from the changes in their methods and adjust their instructional practices, school mean FCAT Reading scores increased between 2004-05 and 2006-07.

PAGE 291

277 Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and FCAT Math at All School Levels without Gifted678 680 682 684 686 688 690 692 2003200420052006 YearFCAT Math all schools + 2SD all schools + 1SD all schools mean all schools 1SD all schools 2SD Figure 47. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and FCAT Math at All School Levels without Gifted The trends for FCAT Math achievement may demonstrate similar responses to the consequences of Florida’s A+ Law. Initially, the trend between 200 3-04 and 2004-05 was shallow. However, the steep incline in the mean school FCAT Math scores between 2004-05 and 2005-06, similar to the trend for Reading FCAT scores, is prominent in most of the figures that illustrate the relationship between the technology integration variables and mean school FCAT Math score. The s-shape of the graphs depicting the shallower gains made from 2005-0 6 to 2006-07 may indicat e that there is a ceiling to the amount that schools can increase their mean FCAT Math scores. More time will be needed to see if this is a ceiling or if the upward trend continues; as suggested by the rese arch reported by Borman (2003), positive results of school reform may take five years or more to become established.

PAGE 292

278 Interestingly, this study found a significant relationship between U.S. magnet schools and FCAT Writing achievement. Although being a U.S. magnet school had no significant relationship with the intercept, the interactions of U.S. magnet school with time and time2 were significant. This resulted in two changes of direction over time with approximately 0.03 point difference between magnet and non-magnet schools. Schools designated as U.S. magnet schools began the study in 2003-04 with lower mean school FCAT Writing scores and then in 2004-05 U.S. magnet schools had higher mean school FCAT Writing scores. This trend reversed in 2005-06 and remained reversed in 2006-07 (see Figure 48). Although this is a significant trend over time, the difference is modest. Th e flatter pattern may be due to the specific focus of the U.S. magnet school on an interest area, so that le ss attention is devoted to the development of writing skills. Relationship between US Magnet School and FCAT Writing Scores at All School Levels 3.50 3.60 3.70 3.80 3.90 4.00 4.10 4.20 4.30 4.40 4.50 200320042005 YearFCAT Writing Not a US Magnet Schools All Schools US Magnet Schools All Schools Figure 48. Relationship between U.S. Magnet School St atus and Mean FCAT Writing Scores in All Schools

PAGE 293

279 Overall, the inconsistent findings counter the prev ious positive reports about the relationship of technology integration with student achievement in reading, math, and writing (Borman, 2003; Goldberg et al., 2003; Kulick, 2003; Lowther et al., 2003; Mann, 1999; O’Dwyer et al., 2005; Penuel et al., 2002; Wenglinsky, 1998, 2005) and support newer research results reported by Dynski et al. (2007), Russell et al. (2004), and Shapley et al. (2006). More time is need ed to examine this relationship in Florida. These inconsistent trends may also be due to the differen t ways that technology can be integrated into the curriculum for students. Future research about the re lationship between technology integration and student achievement must examine how the technology is used. Another factor that may interfere with finding positive relationships between technology integration and student achievement is the measurement of achievement. Skills are best assessed using the same methods and activities in which the student customarily uses the skills (Berliner, 1990; Kim & R eeves, 2007; Russell & Higgins, 2003; Wenglinsky, 2005). Positive relationships may be found when stud ent achievement is assessed through the technology that the students used to learn and practice those skills. In addition, only using the results of one standardized assessment each year may not adequately measure students’ achievement, especially when the students are gifted and already are performing at the top of the measurement scale. Multiple assessments conducted in a variety of formats over many points in time would better represent students’ growth in skills and knowledge. Another explanation for these inconsistent findings is that the relationship between the integration of technology and achievement is dynamic and the path occurs in both directions. That is, teachers influence what students do and learn, and in turn, the responses of the students influence how teachers modify and adapt or change the instructional methods that they use. Th e integration of technology is only one of the many factors in the complex lear ning phenomenon that occu rs within the classroom. Mediating Outcomes Research Question 2 What is the relationship between indicators of technology integration and changes in mediating outcomes of absence rate and student misconduct, when controlling for school level, school socioeconomic status, minority, limited E nglish proficiency, students with disabilities, gifted, teacher qualification, and learning environment quality? The analysis conducted to answer the second research question ex amined the relationship between technology integration variables and mediating variables (percent of students absent more than 21 days and mean school student misconduct score), while controllin g for moderating variables. This research question used multi-level models with percent of students abse nt more than 21 days and mean student misconduct

PAGE 294

280 score as outcomes. Misconduct was measured with a composite variable created from the sum of the mean percent of students with in-school suspensions, mean percent of students with out-of-school suspensions, and the mean number of crime incidents per student. In general, both the percentage of students absent 21+ days for all public schools in Florida and the mean school student misconduct score had significant variability and significant slopes that were curvilinear because time and time2 were significant. First, after controlling for school level and scho ol level demographic and learning environment variables, the technology indicators that were not significantly related to percent of students with more than 21 days absent or mean school student misconduct scores were examined. The first set of indicators measured student access to software. Student access to a ll types of software (i.e., software for delivery of content, office suite software, and advanced production software) was not significantly related to the percent of students with more than 21 days absent nor significantly related to mean school student misconduct scores. Furthermore, the results with both datasets indicated that the frequency that students used content delivery software or tool-based software was not significantly related to percent of students with more than 21 days absent or mean school student misconduct. These results counter the findings about the relationship between student use of software and attendance found by Barron et al. (1999), but support the findings of Muir-Herzig (2004) and Shapely et al. (2006). Likewise the frequency that students used content delivery software or tool-based software was not significantly related to mean school student misconduct either, which supports the findings of Waxman (2003) and counter the findings of Barron et al. (1999), Kmitta and Davis (2004), and Shapely et al.(2006). However, the measures in this study may not adequately measure the frequency of specific methods that students use technology. Accordingly, it is important in the future to continue to examine the relationship between how and how often students use technology and student absences and student misconduct. Future studies should include information about how technology has been integrated into the curriculum and the met hods that students and teachers are using with technology that is collected from a variety of sources. Triangulation of findings would need results from teacher and student surveys an d interviews and classroom observations. Another technology indicator that had no significant relationship with percent of students with over 21 days absent or with mean school student misconduct scores in both datasets was the composite variable used to measure the level of technical support for hardware and access to the Internet. However,

PAGE 295

281 with the dataset that included all schools without gifted, there was a significant interaction between technical support – human and percent of students with over 21 days absent with time and time2. Although all schools started at the same level of student absen ces, through the course of study the relative level of student absences in schools with two standard deviations above the mean and two standard deviations below the mean of technical support – human changed. That is, in 2004-05, schools with two standard deviations above the mean of technical support – human had the greatest absences, and in 2005-06 schools at two standard deviations above the mean had the l east absences. This may be a result of the changing demands on the types and modes of support provided by the specialists. The students’ response may be different to a tech specialist working in the classroom with students, as opposed to the tech specialist providing the teachers with training outside of the classroom. Closer ex amination of the specific supports provided by the technical support and technology inte gration support specialists and their relationship with student attendance is needed to understand the dynamics of this relationship. The findings from this study i ndicated that the percent of teach ers who regularly use technology for delivery of instruction was not significantly relate d to percent of students with more than 21 days absent, which supports the findings of Muir-Herzig (2004). This suggests that the path between teacher use of technology for delivery of instruction and percent of students with more than 21 days absent is not direct. However the percent of teacher s who regularly use technology to deliver instruction was significantly related to mean school student misconduct at the intercept with the dataset that included all schools without gifted. This means that schools with the greatest per cent of teachers who regularly use technology to deliver instruction were predicted to begin the study with the lowest levels of school misconduct. Because the interactions with time were no t significant, once established, this trend remained parallel (see Figure 49). One explanation is that when teachers use technology to deliver instruction, students find the lessons more engaging and are spending more time on-task learning, which results in decreased off-task and disruptive behavior. Another explanation is that schools that have less disruptive behavior allow more teachers to use t echnology to deliver instruction. Futu re research is needed to examine how teachers are using the technology to deliver instruction and the relationship of different instructional methods with student misconduct.

PAGE 296

282 Relationship between Teachers Who Use Technology to Deliver Instruction and Student Misconduct at All School L evels wit hout Gifted5 10 15 20 25 30 35 40 45 50 200320042005 YearStudent Misconduct all schools + 2SD all schools + 1SD all schools mean all schools 1SD all schools 2SD Figure 49. Relationship between Percent of Teachers Who Regularly Use Technology to Deliver Instruction and Student Misconduct at All School Levels (without Gifted) When the data with all school levels without gi fted and the data at el ementary and middle school levels with gifted were examined, the technology indicator that was significantly related to percent of students with more than 21 days absent and mean school student misconduct score was the relationship with the percent of teachers who regularly use technology for administrative purposes. The relationship of the percent of teachers who regularly use technology for administrative purposes and percent of students with more than 21 days absent was significant at the intercept, time, and time2 (see Figure 50). This resulted in curvilinear trends in both datasets with the relati onship between the highest and lowest levels of the percent of teachers who regularly use technology for administrative purposes and percent of students with more than 21 days absent reversing through time. When examining the relationship with school mean student misconduct scores in both the dataset for all schools without gifted and for elementary and middle schools with gifted, the interactions of the

PAGE 297

283 percent of teachers who regularly use technology for administrative purposes with time and time2 were significant, but the intercept was not si gnificant (see Figure 51). This meant that schools at all levels of the percent of teachers who regularly use technology for administrative purposes began the study with approximately the same level of student misconduct. The trends in both datasets were curvilinear with the relationship between the highest and lowest levels of the percent of teachers who regularly use technology for administrative purposes and mean school student misconduct reversing through time. Schools that began with two standard deviations above the mean in percent of teachers who regularly use technology for administrative purposes had the greatest decline in level of student misconduct in the first year and the least decline in the second year. This trend was reversed for schools that began the study with two standard deviations below the mean in percent of teachers who regularly use technology for administrative purposes. At the end of the study, the level of student misconduct for all schools had decreased. This study cannot determine causality. It could be that the level of student attendance or the level of student misconduct impacts th e degree that teachers regularly use technology for administrative purposes. Datasets of longer duration are needed for examining this relationship in order to understand these associations. Also, examination of the trends of the relationship between the percent of teachers who regularly use technology for administrative purposes and level of student attendance and level of student misconduct should be continued to better understand their shape and to determine if the equation for the model is cubic.

PAGE 298

284 Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Student Absences at All School Levels without Gifted8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 12.5 13.0 13.5 14.0 200320042005 YearPercent of Students with 21+ Days Absent all schools + 2SD all schools + 1SD all schools mean all schools 1SD all schools 2SD Figure 50. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Percent of Students Absent More than 21 Days in All Schools without Gifted.

PAGE 299

285 Relationship between Percent of Teachers Who Use Technology for Administrative Purposes and Student Misconduct at All School Levels without Gifted5 10 15 20 25 30 35 40 45 50 200320042005 YearStudent Misconduct all schools + 2SD all schools + 1SD all schools mean all schools 1SD all schools 2SD Figure 51. Relationship between Percent of Teachers Who Regularly Use Technology for Administrative Purposes and Student Misconduct at All School Levels (without Gifted) Interestingly, this study found a significant pos itive relationship between U.S. magnet schools and percent of students with more than 21 days absent (see Figure 52). Being a U.S. magnet school had a significant relationship with the percent of students absent more than 21 days at the intercept and with the interaction with time and time2. This resulted in U-shaped trends over time. Schools designated as a U.S. magnet school began the study in 2003-04 with 8.64% of students with more than 21 days absent, while those schools that were not U.S. magnet schools began with 6.71% of students with more than 21 days absent. The percent of students with more than 21 days absent in U.S. magnet schools at all school levels increased each year, while the percent of students with more than 21 days absent in schools that were not designated as a U.S. magnet school decreased each y ear. With the dataset that included elementary and middle schools with gifted, the relationship between U. S. magnet status with the intercept for percent of students with more than 21 days absent was not significant, but the relationships with time and time2 were

PAGE 300

286 significant. This resulted in similar trends with the schools designated as U.S. magnet schools having a higher percent of students with more than 21 days absent. Relationship between US Magnet School and Student Absences at All School Levels 0 2 4 6 8 10 12 14 16 18 20 200320042005 YearPercent of Students with 21+ Days Absent Not a US Magnet Schools All Schools US Magnet Schools All Schools Figure 52. Relationship between U.S. Magnet School Status and Student Absences at All School Levels However, when the relationship of percent of students with more than 21 days absent and U.S. technology magnet school was examined, only the intercept was significant. This meant that U.S. technology magnet schools were predicted to have begun the study with 3.24% less students with more than 21 days absent than schools without this designa tion. Over time the trends of schools with and without technology magnet status were parallel (see Figure 53).When the dataset that included elementary and middle schools with gifted was examined, again only the intercept was significant. This suggests that there may be a negative relationship between advanced levels of technology integration and student absences, or it may indicate that technology magnet schools attract more students who attend school. The sample set of U.S. technology magnet schools included only 8 schools over all three points in time.

PAGE 301

287 These inconsistent patterns may be due to the dynamic nature of the interactions within the classroom. Students respond to lear ning activities and interactions that they have with their teachers by changing their behavior, and in turn, their teachers respond to these changes of their students by modifying and adapting their methods. The path between student behavioral outcomes and the methods teachers use to integrate technology may occur in both directions. Datasets of longer duration are needed to investigate the trends in these relationships to determine if meaningful and consistent relationships evolve. Relationship between US Technology Magnet School and Student Absences at All School Levels 0 2 4 6 8 10 12 14 16 18 20 200320042005 YearPercent of Students with 21+ Days Absent Not a US Technology Magnet School All Schools US Technology Magnet School All Schools Figure 53. Relationship between US Technology Magnet School and Student Absences at All School Levels without Gifted Variance Explained A common result obtained from analyses across bo th achievement and behavioral outcomes was the change in variance explained. The variances for each achievement model estimated for all school levels without gifted and estimated for elementary and middle schools with gifted are depicted in Table 67, and each behavioral model are depicted in

PAGE 302

288 Table 68. A large proportion of variance was explained when school level was added to the model as a predictor variable for each achievement outcome. This supports previous re search that reports the importance of school level for influencing both instructional methods and student achievement (Barron et al., 2003; Benner et al., 2002; Hart et al., 2002; Wenglinsky, 1998, 2005). Future research must continue to look more closely at the specific methods of how techno logy is integrated into the instructional routines at each level of school. The next model, which included the addition of demographics variables, resulted in a large, additional proportion of variance explained. This supports previous research that reports the importance of school level demographics (in particular economically disadvantaged status) in having relationships with the methods of instruction and use of technology within schools (Aldeman et al., 2002; Becker, 2001; DeBell & Chapman, 2006; Lubienski, 2006; National Center for Education Statistics, 2005; Parsad & Jones, 2005; Wenglinsky, 1998, 2004, 2005). Ongoing research is needed to monitor the equity of educational opportunities afforded to students in Florida’s K-12 institutions. The addition of positive learning environment variables into the models resulted in the next largest reduction in variance in the models. These models ex plained most of the available variance, supporting recommendations by Barron et al. (1999), Bloom (1968, 1976, 1984), Carroll (1963, 1989), Marzano, 2003, and Slavin (1987, 1994) that the learning environment is a critical component that impacts students’ learning and achievement. These reductions in variance by the moderating variables demonstrate the importance of including them in the model. However, because these variable s and the technology indicat ors were correlated, the variance explained cannot be used as a measure of th eir relative importance (Pedhazur, 1997). The order for adding variables into the model determines the amount of variance explained by each model. If they had been added in a different order, the variance explained attributed to each model would have been different. Technology integration indicators were added last in order to determine which indicators were significant after controlling for all other variables. It is importa nt that significant technology integration indicators were found.

PAGE 303

289 Table 67. Variance for Each Model for Ac hievement Outcomes by Dataset Achievement Outcome and Variance Reading Math Writing Dataset and Model 2 2 2 All Schools without Gifted Model 1: Unconditional Model 496.31 41.79 956.72 45.22 0.0678 0.0355 Model 2: Growth Model Time as a Predictor 497.98 35.10 961.92 24.42 0.0771 0.0199 Model 2: Quadratic Model Time2 as a Predictor 498.62 32.55 962.73 21.19 0.0772 0.0199 Model 2: Polynomial Model -Time3 as a Predictor 510.34 15.62 963.86 16.67 0.0777 0.0192 Model 3: School Level Model 175.28 10.93 229.53 14.21 0.0734 0.0184 Model 4: Demographics Model 40.55 10.17 70.90 13.74 0.0432 0.0183 Model 5: Learning Environment Model 32.14 10.25 57.03 13.88 0.0401 0.0183 Model 6: Technology Integration Model 32.00 10.19 56.24 13.80 0.0399 0.0182 Final Model 7: Significant Technology Integration Model 31.99 10.22 56.49 13.84 0.0399 0.0183 Elementary and Middle Schools with Gifted Model 1: Outcome Predicted by Average Outcome of All Schools in Florida 291.17 42.86 531.44 48.58 0.0669 0.0388 Model 4: Demographics Model 26.33 9.56 47.53 13.03 0.0429 0.0191 Model 5: Learning Environment Model 22.74 9.53 42.01 12.98 0.0416 0.0190 Model 6: Technology Integration Model 22.59 9.46 41.43 12.86 0.0414 0.0189 Final Model 7: Significant Technology Integration Model 22.68 9.48 41.59 12.92 none Table 68. Variance by Each Model for Each Mediating Outcome by Dataset Mediating Outcome and Variance Absences Misconduct Dataset and Model 2 2 All Schools without Gifted Model 1: Unconditional Model 29.34 9.08 417.19 67.87 Model 2: Growth Model Time as a Predictor 23.94 7.42 509.77 43.53 Model 2: Quadratic Model Time2 as a Predictor 24.05 7.29 509.93 43.34 Model 3: School Level Model 16.22 7.18 262.07 43.29 Model 4: Demographics Model 10.57 7.46 193.62 43.60

PAGE 304

290 Mediating Outcome and Variance Absences Misconduct Dataset and Model 2 2 Model 5: Learning Environment Model 8.66 5.94 11.49 6.99 Model 6: Technology Integration Model 8.73 5.89 11.53 6.98 Final Model 7: Significant Technology Integration Model 8.71 5.88 11.52 6.96 Elementary and Middle Schools with Gifted Model 1: Outcome Predicted by Average Outcome of All Schools in Florida 19.11 8.03 357.88 60.26 Model 4: Demographics Model 4.30 6.24 157.77 35.04 Model 5: Learning Environment Model 3.74 5.03 6.09 6.07 Model 6: Technology Integration Model 3.76 5.02 6.05 6.09 Final Model 7: Significant Technology Integration Model 3.78 4.98 6.11 6.03 Instrumentation At various stages of this study, issues related to the measurement of variables were a concern. The first issue was the ability to accurately count the number of computers that students have available and the locations for these computers. This is important information for policy makers, educators, and researchers to have in order to track chan ges in technology access, as well as for planning future initiatives. Another issue was the reliability of the items used to measure factors of interest. The items in the Florida Innovates Survey need to be continuously evaluated and revi sed in order to provide accurate measures. This is especially needed for items that measure how students are using technology and how frequently they use the technology. The Florida Innovates Survey is one method used to measure technology integration. Additional measures are needed in order to collect more detailed information about the integration of technology in the instructional routines within the schools. Teachers an d students should provide information about how and how often they are using technology in order to make meaningful decisions about how technology integration is related to achievement. Finally, in or der to connect technology integration with student achievement and student behavioral outcomes, student level data is needed.

PAGE 305

291 Conclusions According to complexity theory, information must be exchanged among the elements in an organization in order to maintain organization health (Caldwell, 2005; McElroy, 2000; Morrison, 2002). The exchange and availability of this information is essential for the organization to be able to adapt and survive (O’Day, 2002; Wheatley, 1999). Educatio n is a complex phenomenon; therefore, multiple measurements over extended periods of time are required (Slavin, 1997, 1994). Spanning a four-year period, this study examined the relationship of the inte gration of technology with mean school achievement when controlling for moderating variables in Florida public elementary, middle, and high schools. Noteworthy, the variance explained by the addition of moderating variables demonstrate how important school level, demographics, and learning environment variables are in the complex model of school achievement. When applied to the learning or adaptation of the school organization, the Carroll Model of School Learning (1963, 1989) and Bloom’s Theory of School Learning (1968; 1976) explain why the amount of time required for observable change or opportunity to learn is critical. In order to change or increase achievement levels, the school organization must have an opportunity to change that matches the time it needs to change. During the change pr ocess the individual agents must ad just to feedback and improve their performance to support the school improvement plan (O’Day, 2002; Wheatley, 1999). Observable improvements in achievement at the school level are ex pected to be small because they are measured by the mean of all the changes in an outcome for all the st udents in the school. The small changes observed in mean school achievement over the span of this study, conf irm that time is a critical factor for both learning of individuals (Berliner, 1990; Bloom, 1984; Carroll, 1963, 1989; Marzano, 2003) and school change (Borman, 2003; Weick & Quinn, 1999). Bloom (1968; 1976) attempted to delineate the amount of variance for achievement that is explained by each of the five factors in the Carroll Model of School Learning by using correlations derived from research. He reported the follo wing correlations between achievem ent and factors in his Theory of School Learning when the time allotted to learning is optimal: cognitive entry behaviors, which include Carroll’s aptitude for learning and ability to understand (+.70 or one half of the variance); quality of instruction (+.50 or one fourth of the variance); and pe rseverance or affective en try behaviors (+.50 or one

PAGE 306

292 fourth of the variance). The factor that is under the co ntrol of the school is the quality of instruction or one fourth of the variance of school level achievement. Quality of instruction is complex that includes many variables such as the expertise of the teacher, the positive learning en vironment, and the many variables involved in the implementation of the curriculum and daily lesson plans. The integration of technology is only one of these many factors. Becau se all of these factors or variable s that impact the student learning within the classroom envir onment are correlated, it is difficult to extract their unique contribution to the equation. As a result, the expected variance explained by the integra tion of technology, as one of the variables within the quality of instruction component of the school learning equation should be very small. Indeed, after controlling for all other variables the small variance explained by the addition of technology integration variables, demonstrate this phenomenon. It is important that after controlling for all the other moderating variables, technology integration did have a significant relationship with mean school achievement. Moreover, this study demonstrates that the schools’ responses to technology integration as an agent of change have been episodic and non-linear (Caldwell, 2005; Jacobson & Wilensky, 2006). The impact of change agents, such as new educational programs or technology initiatives, takes time to become apparent. The positive results from comprehensive school reform occurred after the fifth year (Borman et al., 2002). In addition, the inconsistencies in the significant findings about the relationship between technology integration variables and the outcomes studied support the need for more time to establish trends and patterns. Therefore, continued analyses of the longitudinal trends for Florida schools in the relationship between technology integration variables and school achievement, while controlling for moderating variables, are recommended.

PAGE 307

293 References Adelman, N., Donnelly, M. B., Dove, T., Tiffany -Morales, J., Wayne, A., & Zucker, A. (2002). The integrated studies of educatio nal technology: Professional development and teachers’ use of technology. Arlington, VA: SRI International. Retrieved February 17, 2008, from http://policyweb.sri.com/cep /research_areas/displayRAPa stProjects.jsp?Nick=tealearn Anderson, R. E. & Becker, H. J. (2001). School investments in instructional technology. Report #8 Irvine, CA: University of California, Center for Research on Information Technology and Organizations. Retrieved February 17, 2008, from http://www.crito.uci.edu /tlc/html/findings.html Anderson, R. E. & Dexter S. L. (2001). School technology leadership: Incidence and impact. Report #6. Irvine, CA: University of California, Center for Research on Information Technology and Organizations. Retrieved February 17, 2008, from http://www.crito.uci.edu /tlc/html/findings.html Apple Computer, Inc. (1995). Changing the conversation about teaching, learning, & technology: A report on 10 years of ACOT research Cupertino, CA: Apple Computer, Inc. Retrieved February 17, 2008, from http://www.apple.com/education/k12/leadership/acot/library.html Barrnett, H. (2003). Investing in technology: The payoff in student learning. Eric Digest (ERIC Document Reproduction Service No. ED479843) Barron, A.E., Hogarty, K.Y., Kromrey, J.D., & Lenkway P. (1999). An examination of the relationships between student conduct and the number of computers per student in Florida schools. Journal of Research on Computing in Education, 32 (1), 98-107. Barron, A. E., Kemker, K., Harmes, C., & Kalaydjian, K. (2003). Large-scale resear ch study on technology in K-12 schools: Technology integration as it relates to the National Technology Standards. Journal of Research on Technology in Education. 35 (4), 489-507. Bebell, D. (2005).Technology Promoting Student Excellence: An investigation of the first year of 1:1 Computing in New Hampshire middle schools. Technology and Assessment Study Collaborative, Boston College. Retrieved February 17, 2008, from http://escholarship.bc.edu/intasc/32/

PAGE 308

294 Becker, H. J. (2001). How are teachers using computers in instruction? Paper presented at the annual meeting of the American Educational Research As sociation, Seattle, WA, April, 2001. Retrieved February 17, 2008, from http://www.crito.uci.edu/tlc/html/findings.html Becker, H. J., Ravitz, J. R. & Wong Y. (1999). Teacher and teacher-directed student use of computers and software. Report #3 Irvine, CA: University of California, Center for Research on Information Technology and Organizations. Retrieved February 17, 2008, from http://www.crito.uci.edu/tlc/html/findings.html Benner, A. D., Shapley, K. S., Heikes, E. J., & Pieper, A. M. (2002). Technology integration in education (TIE) initiative: Stat ewide survey report Austin, TX: Texas Center for Educational Research. Retrieved February 17, 2008, http://www.tcer.org/research/tie/index.aspx Berends, M., Kirby, S. N., Naftel, S., & McKelvey, C. (2001). Implementation and performance in New American Schools: Three years into scale-up. Santa Monica, CA: Rand. Berliner, D. C. (1990). What’s all the fuss about instructional time? The Nature of Time in school theoretical concepts, practitioner perceptions New York: Teachers College Press. Retrieved February 17, 2008, from http://courses.ed.asu.edu/berliner/readings/fuss/fuss.htm Bloom, B. S. (1968). Learning for mastery. Evaluation Comment 1 (2), 1–11. Bloom, B. S. (1976). Human characteristics and school learning. New York: McGraw-Hill. Bloom, B. S. (1984). The 2 sigma problem: The search for methods of group instruction as effective as one-to-one tutoring. Educational Researcher 13(6), 4-16. Borman, G. D., Hewes, G. M., Overman, L. V., & Brown, S. (2003). Comprehensive school reform and achievement: A meta-analysis. Review of Educational Research, 73 (2), 125-230. Bruning, R. H., Schraw, G. J., Norby, M. M., & Ronning, R. R. (2004). Cognitive psychology and instruction (4th ed.). Upper Saddle River: Merrill/Prentice Hall. Bureau of Education Information and Accountability Se rvices, Florida Department of Education. (2007). Technical assistance paper: Master School Identification File – 2006-07 Available from http://doeweb-prd.doe.state.fl.us/EDS/MasterSchoolID/index.cfm

PAGE 309

295 Bureau of Instruction and Innovation, Florida Department of Education (2007a). Florida innovates school survey results. Retrieved February 17, 2008, from http://www.flinnovates.org/survey/PDF/FL_Innovates_2006-07_State_Report.pdf Bureau of Instruction and Innovation, Florida Department of Education (2007b). Florida innovates school survey [Data files]. Available from http://www.flinnovates.org/survey/index.php Bryk, A. S. & Hermonson, K. L. (1993). Educational indicator systems: Observations on their structure, interpretations, and use. Review of Research in Education 19, 451-483. Caldwell, R. (2005). Things fall apart? Discou rses on agency and change in organizations. Human Relations 58 (1) 83-114. Carroll, J. B. (1963). A model of school learning. Teachers College Record, 64 723-733. Carroll, J. B. (1989). The Carroll model: A 25-year retrospective and prospective view. Educational Researcher, 18 (1) 26-31. Christensen, R., Gr iffin D., & Knezek, G. (2001). Measures of teacher stages of technology integration and their correlates with student achi evement. Paper presented at the annual meeting of the American Association of Colleges for Teacher Ed ucation, March 1-4, 2001, Dallas TX. Coughlin, E., & Lemke, C. (1999). Professional compet ency continuum: Professional skills for the digital age classroom (Dimension 3). Santa Monica, Ca lifornia: Milken Exchange on Education Technology. Retrieved February 17, 2008, from http://www.mff.org/edtech/projects.taf?_function=detail&Content_uid1=104 Cuban, L. (1986). Teachers and machines: The classroom use of technology since 1920. New York: Teachers College Press. Cuban, L. (1998). How schools change refo rms: redefining reform success and failure. Teachers College Record, 99, 453-77. Cuban, L. (2001). Oversold and underused: Co mputers in the classroom Cambridge: Harvard University Press. Culp, K. M., Honey, M., & Maninach, E. (2005). A retrospective on twenty years of education technology policy. Journal of Educational Computing Research, 32 (3) 279-307.

PAGE 310

296 Dede, C., Korte, S., Nelson, R., Valdez, G., Ward, D. J. (2005). Transforming learning for the 21st century: An economic imperative Naperville, IL: Learning Point Associ ates. Retrieved February 17, 2008, from http://www.learningpt.org/page.php?pageID=138 DeBell, M., & Chapman, C. (2006). Computer and internet use by students in 2003 (NCES 2006–065). Washington, DC: U.S. Department of Education, National Center for Education Statistics. Dennen, V. P. (2004). Cognitive apprenticeship in edu cational practice: Research on scaffolding, modeling, mentoring, and coaching as instructional practices. In Jonassen, D. (Ed.). Handbook of research on educational communications and technology (2nd ed., pp. 813-828). Association for Educational Communications and Technology. Dexter, S. L., Anderson, R. E., & Ronnkvist, A. M. (2002). Quality technology support: What is it? Who has it? And what difference does it make? Journal of Educational Computing Research, 26 (3) 265-285. Division of Accountability, Research and Measurement, Florida Department of Education. (2007a). Master School Identification [Data files] Available from http://doewebprd.doe.state.fl.us/EDS/MasterSchoolID/index.cfm Division of Accountability, Research and Measurement, Florida Department of Education. (2007b). Florida Comprehensive Assessment Test Norm Referenced Test Scores [Data files]. Available from Assessment and School Performance Florida Comprehensive Assessment Test Web site, http://fcat.fldoe.org/ Division of Accountability, Research and Measurement, Florida Department of Education. (2007c). Florida School Indicators Report (FSIR) [Data files]. Available from http://data.fldoe.org/fsir/default.cfm Division of Accountability, Research and Measurement, Florida Department of Education. (2007d). Measuring Adequate Yearly Progress (AYP) [Data files]. Available from http://schoolgrades.fldoe.org/reports/index.asp Donnelly, M. B., Dove, T., & Tiffany-Morales, J. (2 002). Technology-related professional development in the context of educational reform: A literature re view. Arlington, VA: SRI International. Retrieved February 17, 2008, from http://policyweb.sri.com/cep /research_areas/displayRAPa stProjects.jsp?Nick=tealearn

PAGE 311

297 Dwyer, D. C., Ringstaff, C., & Sandholtz, J. H. (1990). Teacher beliefs and practices part I: Patterns of change. Cupertino, CA: Apple Computer, Inc. Retrieved February 17, 2008, from http://www.apple.com/education/k12/leadership/acot/library.html Dynarski, M., Agodini, R., Heaviside, S., Novak, T., Carey, N., Campuzano, L., Means, B., Murphy, R., Penuel, W., Javitz, H., Emery, D., & Sussex, W. (2007). Effectiveness of reading and mathematics software products: Findings from the first student cohort (NCEE 2007-4005). Washington, D.C.: U.S. Department of Education, Institute of Education Sciences. Florida Department of Education (2007a). Change, and response to change, in Florida’s public schools Tallahassee, FL: Education Information and Accountability Services. Retrieved February 17, 2008, from http://www.fldoe.org/eias/ei aspubs/pdf/changes0207.pdf Florida Department of Education (2007b). Florida school indicators report: Indicator descriptions Bureau of Education Information and Accountability Services. Retrieved February 17, 2008, from http://data.fldoe.org/fsir/indicator_desc.cfm Florida Department of Education (2007c). 2007 Guide to Calculating Adequate Yearly Progress (AYP) Technical Assistance Paper. Office of Evaluation and Reporting, Division of Accountability, Research, and Measurement. Retrieved February 17, 2008, from http://schoolgrades.fldoe.org/pdf/0607/2007AYPTAP.pdf Florida Department of Education (2006). FCAT Update. Assessment & School Performance Division of Accountability, Research, and Measurement. Retrieved February 17, 2008, from http://fcat.fldoe.org/pdf/FCAT_Update_Oct06.pdf Florida Department of Education (2005a). FCAT Handbook—A Resource for Educators Retrieved February 17, 2008, from http://fcat.fldoe.org/handbk/fcathandbook.asp Florida Department of Education (2005b). The new FCAT NRT. Retrieved February 17, 2008, from http://fcat.fldoe.org/pdf/fcat-nrt-sat10.pdf Fulton, K., Glenn, A. D., & Valdez, G. (2004). Teacher Education and Technology Planning Guide Naperville, IL: Learning Point Associates. Retrieved February 17, 2008, from http://www.ncrel.org/tech/planguide/guide.pdf

PAGE 312

298 Gredler, M. E. (2004). Games and simulations and their relationship to learning. In Jonassen, D. (Ed.). Handbook of research on educational communications and technology (2nd ed., pp. 571-581). Association for Educational Communications and Technology. Goldberg, A., Russell, M., & Cook, A. (2003). The effect of computers on student writing: A meta-analysis of studies from 1992 to 2002. Journal of Technology, Learning, and Assessment, 2 (1). Retrieved February 17, 2008, from http://escholarship.bc.edu/jtla/vol2/1/ Guskey, T. R. (2001). Benjamin S. Bloom’s contributions to curriculum, instruction, and school learning Paper presented at the annual meeting of the Amer ican Educational Research Association, Seattle, WA, April 2001. Harcourt Assessment, Inc. (2002, 2004). Stanford Achievement Test Series, tenth edition: Technical data report San Antonio, TX: Harcourt Assessment, Inc. Hart, H. M., Allensworth, E., Lauen, D. L., & Gladden, R. M. (2002). Educational technology: Availability and use in Chicago’s public schools. Chicago: Consortium on Chicago School Research. Hill, J. R., Wiley, D., Nelson, L. M. & Han, S. (2004) Exploring research on internet-based learning: From infrastructure to interactions. In Jonassen, D. (Ed.). Handbook of research on educational communications and technology (2nd ed., pp. 433-460). Association for Educational Communications and Technology. Holland, J. H. (2006). Studying complex adaptive systems. Journal System Science & Complexity (19) 1-8 Human Resources Research Organization & Harcourt Educational Measurement (2003). Florida Comprehensive Assessment Test for reading an d mathematics: Technical report for test administrations of FCAT 2003. San Antonio, TX: Harcourt Educa tional Measurement. Retrieved February 17, 2008, from http://fcat.fldoe.org/pdf/fcatechrpt2003.pdf ISTE NETS Project (2005a). National Educational Technology Standards for Administrators (NETS•A) Project Retrieved February 17, 2008, from http://www.iste.org/inhouse/nets/c nets/administrators/index.html ISTE NETS Project (2005b). Preparing teachers to use technology (NETS-T) project Retrieved September 13, 2007, from http://www.iste.org/inhouse/ne ts/cnets/teachers/index.html

PAGE 313

299 ISTE NETS Project (2005c). National Educational Technology Standards (NETS) and the states, National Educational Technology Standards Project Retrieved February 17, 2008, from http://cnets.iste.org/docs/States_using_NETS.pdf ISTE NETS Project (2007). National Educational Technology Standards for Students (NETS-S) Project Retrieved February 17, 2008, from http://www.iste.org/inhouse/nets/cnets/index.html Jacobson M. J. & Wilensky, U. (2006). Complex systems in education: Scientific and educational importance and implications for the learning sciences. The Journal of the Learning Sciences, 15 (1), 11-34. Jonassen, D. H., & Reeves, T. C. (1996). Learning w ith technology: Using computers as cognitive tools. In Jonassen, D. (Ed.). Handbook of research on educational communications and technology (1st ed., pp. 693-719). Association for Educational Communications and Technology. Kim, B. & Reeves, T. C. (2007). Reframing research on learning with technology: In search of the meaning of cognitive tools. Instructional Science, 35 207-256. Kmitta, D. & Davis, J. (2004). Why PT3? An anal ysis of the impact of educational technology. Contemporary Issues in Technology and Teacher Education, 4 (3), 323-344. Knezek, G., Christensen, R., & Fluke, R. (2003). Testing a will, skill, tool model of technology integration Paper presented at the annual meeting of the American Educational Research Association, Chicago IL, April, 2003. Kulik, J. (2003). Effects of using instructional technology in elementary and secondary schools: What controlled evaluation studies say ? Arlington, VA: SRI International. Retrieved February 17,2008, from http://www.sri.eu/policy/csted/reports/ sandt/it/Kulik_ITinK12_Main_Report.pdf Lemke, C., Wainer, A., & Haning, N. (2006). National trends: Enhancing education through technology, No Child Left Behind Title IID – year three in review State Educational Technology Directors Association. Retrieved February 17, 2008, from http://www.setda.org/web/ guest/nationaltrendsreport Lockee, B., Moore, D. M., Burton, J. (2003) 20: Foundations of programmed instruction. Handbook of research on educational communications and technology (pp. 545-569). Association for Educational Communications and Technology.

PAGE 314

300 Lowther, D. L., Ross, S. M., & Morrison, G. M. (2003). When each one has one: The influences on teaching strategies and student achieveme nt of using laptops in the classroom. Educational Technology Research and Development 51 (3) 23-44. Luke, D. A. (2004). Multilevel modeling Newbury Park: Sage Publications. Lubienski, S. T. (2006). Examining instruction, ach ievement, and equity with NAEP mathematics data. Education Policy Analysis Archives, 14 (14) 1-30. Retrieved February 17, 2008, from http://epaa.asu.edu/epaa/v14n14/ Luppicini, R. (2007). Review of computer mediated communicatio n research for education. Instructional Science, 35 141-185. Mann, D., Shakeshaft, C., Becker, J., & Kottkamp, R. (1999). West Virginia story: Achievement gains from a statewide comprehensive instructional technology program. Retrieved February 17, 2008, from http://www.mff.org/publications/publications.taf?page=155 Marzano, R. (2003). What works in schools: Translating research into action Alexandria, VA: Association for Supervision and Curriculum. McElroy, M. W. (2000). Integrating complexity theory, knowledge management and organizational learning. Journal of Knowledge Management 4, 195 – 203. McLellan, H. (2004). Virtual Realities. Handbook of research on educational communications and technology (2nd ed., pp.461-490). Association for Educational Communications and Technology. Metri Group (2006). Technology in schools: What the research says. Cisco Systems, Inc. Retrieved February 17, 2008, from http://www.cisco.com/web/strategy/docs/ed ucation/TechnologyinSchoolsReport.pdf Morrison, K. (2002). School leadership and complexity theory New York: Routledge Falmer Mory, E. H. (2004). Feedback research revisited. In Jonassen, D. (Ed.). Handbook of research on educational communications and technology. (2nd ed., pp. 745-783). Association for Educational Communications and Technology. Muir-Herzig, R. G. (2004). Technology and its impact in the classroom. Comput ers & Education, 42 111131.

PAGE 315

301 National Association for the Education of Young Children (1998). Technology and young children—ages three through eight A position statement of the National Association for the Education of Young Children. Washington, DC: National Association fo r the Education of Young Children. Retrieved February 17, 2008, from http://www.naeyc.org/about/positions.asp National Center for Education Statistics (2005). The condition of education 2005 (NCES 2005–094). Washington, DC: U.S. Government Printing Office. National Commission on Excellence in Education (1983). A nation at risk Washington, DC: U.S. Department of Education. Retrieved February 17, 2008, from http://www.ed.gov/pubs/NatAtRisk/risk.html National Council for Social Studies (1994). Expectations of excellence: Cu rriculum standards for social studies. Silver Spring, MD: National Council for the Social Studies. Retrieved February 17, 2008, from http://www.socialstudies.org/standards/strands/ National Council of Teachers of English & International Reading Association (2006). Standards for the English language arts. Retrieved February 17, 2008, from http://www.ncte.org/about/ over/standards/110846.htm National Council of Teachers of Mathematics (1989). Curriculum and evaluation standards for school mathematics Retrieved February, 17, 2008, from http://standards.nctm.org/ National Committee on Science Education Standards, and Assessment & National Research Council (1996). National science education standards Washington, DC: National Academies Press. Retrieved February, 17, 2008, from http://www.nap.edu/catalog/4962.html O'Day, J. (2002). Complexity, accountability, and school improvement. Harvard Educational Review, (72) 3, 293-329. O’Dwyer, L. M., Russell, M., & Bebell, D., (2005). Id entifying teacher, school, an d district characteristics associated with middle and high school teache rs’ use of technology: A multilevel perspective. Journal of Educational Computing Research, 33 (4) 369-393.

PAGE 316

302 O’Dwyer, L. M., Russell, M. & Bebell, D. J. (2004, September 14). Identifying t eacher, school and district characteristics associated with el ementary teachers’ use of tech nology: A multilevel perspective. Education Policy Analysis Archives, 12 (48). February, 17, 2008, from http://epaa.asu.edu/epaa/v12n48/ O’Dwyer, L. M., Russell, M. Bebell, D. J., & Tuck er-Seeley, K. L. (2005). Examining the relationship between home and school computer use and students’ English/ language arts test scores. The Journal of Technology, Learning, and Assessment, 3 (3) Retrieved February 17, 2008, from http://escholarship.bc.edu/jtla/ Oppenheimer, T. (2003). The flickering mind: The false promise of technology in the classroom and how learning can be saved. New York: Random House. Park, O. & Lee, J. (2004). Adaptive instru ctional systems. In Jonassen, D. (Ed.). Handbook of research on educational communications and technology (2nd ed., pp. 651-684). Association for Educational Communications and Technology. Parsad, B. & Jones, J. (2005). Internet access in U. S. public schools and classrooms: 1994–2003 (NCES 2005–015). Washington, DC: U.S. Department of Education National Center for Education Statistics. Pearson, P. D., Ferdig, R. E., Blomeyer, Jr., R. L., & Moran, J. (2005). The effects of technology on reading performance in the middle-school grades: A meta-analysis with recommendations for policy. Naperville, IL: Learning Point Associates. Retrieved February 17, 2008, from http://www.ncrel.org/tech/reading/index.html Pedhazur, E. J. (1997). Multiple regression in behavioral research: Explanation and prediction (3rd ed.). United States: Wadsworth: Thomson Learning. Penuel, W. R., Kim, D. Y., Michalchik, V., Lewis. S., Means, B., Murphy, R., Korbak. C., & Whaley, A., (2002). Using technology to enhance connections between home and school: A research synthesis. Arlington, VA: SRI International. Re trieved February 17, 2008, from http://ctl.sri.com/publications/d isplayPublication.jsp?ID=83

PAGE 317

303 pjmathison (2006). 200 6-07 Florida state government t echnology investment forecast. Market Research Retrieved October 9, 2006, from http://www.marketresearch.com/product/displ ay.asp?productid=1217274&SID=81902631359075401-294040174#pagetop Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis methods Newbury Park: Sage Publications. Rieber, L. P. (2004). Microwor lds. In Jonassen, D. (Ed.). Handbook of research on educational communications and technology (2nd ed., pp.583-603). Association for Educational Communications and Technology. Ringstaff, C. & Kelley, L. (2002). The learning return on our educational technology investment: A review of findings from research San Francisco: WestEd RTEC. Retrieved February 17, 2008, from http://www.eric.ed.gov/ERICWebPortal/custom/por tlets/recordDetails/detailmini.jsp?_nfpb=true &_&ERICExtSearch_SearchValu e_0=ED462924&ERICExtSearch_SearchType_0=no&accno=E D462924 Romiszowski, A. & Mason, R. (2004). Computer-mediated communication. In Jonassen, D. (Ed.). Handbook of research on educational communications and technology (2nd ed., pp. 387-431). Association for Educational Communications and Technology. Ronnkvist, A. M., Dexter, S. L ., & Anderson, R. E. (2000). Technology support: Its depth, breadth, and impact in America’s schools. Report #5 Irvine, CA: University of Ca lifornia, Center for Research on Information Technology and Organizations. Retrieved February 17, 2008, from http://www.crito.uci.edu/tlc/html/findings.html Roschelle, J., Tatar, D., Shechtman, N., Hegedus, S., Hopkins, B., Knudsen, J., & Stroter, A. (2007). Can a technology-enhanced curriculum improve student learning of important mathematics? (SimCalc Technical Report 1). Menlo Park, CA: SRI International. Russell M. & Higgins, J. (2003). A ssessing effects of technology on learning: Limitations of today’s standardized tests. Boston College: Technolo gy and Assessment Study Collaborative. Retrieved February 17, 2008, from http://escholarship.bc.edu/intasc/3/

PAGE 318

304 Russell, M., O’Dwyer, L., Bebell, D., & Tucker-Seeley, K. (2004). Examining the relationship between students’ mathematics test scores and computer use at home and at school. Boston, MA: Technology and Assessment Study Collaborative, Boston College. Retrieved February, 17, 2008, from http://escholarship.bc.edu/intasc/28/ SAS Institute Inc. (2005). SAS/STAT (version 9.13) [Computer Program]. Cary, NC: SAS. Shapiro, A. & Niederhauser, D. (2004). Learning from hypertext: Research issues and findings. In Jonassen, D. (Ed.). Handbook of research on educational communications and technology (2nd ed., pp. 605-620). Association for Educational Communications and Technology. Shapley, K., Sheehan, D., Caranikas-Walker, F., Huntsberger, B., & Maloney, C. (2006). Evaluation of the Texas technology immersion pilot: First-year results Austin, TX: Texas Center for Educational Research. Retrieved February 17, 2008, from http://www.tcer.org/research/etxtip/index.aspx Silvernail, D. L., & Lane, D. M. M. (2004). The impact of Maine’s one-to-one laptop program on middle school teachers and students. Maine Education Po licy Research Institute, University of Southern Maine Office. Retrieved February 17, 2008, from http://mainegovimages.informe.org/mlte/articles/resear ch/MLTIPhaseOneEvaluationReport2004.pdf Slavin, R.E. (1987). A theory of school and classroom organization. Educational Psychologist, 22 89-108. Slavin, R.E. (1994). Quality, appropri ateness, incentive, and time: A mode l of instructional effectiveness. International Journal of Educational Research, 21 141-157. Slavin, R.E. (2005). Evidence-based reform: Advancing the education of students at risk. Retrieved from February, 17, 2008, from http://www.americanprogress.org/issues/2005/03/b492641.html Smerdon, B., Cronen, S., Lanahan, L., Anderson J., Iannotti, N., & Angeles, J. (2000). Teachers' tools for the 21st century: A report on teachers' use of technology (NCES 2000-102). U.S. Department of Education. Washington, DC: National Center for Education Statistics. Retrieved February 17, 2008, from http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2000102 Taylor, L. M., Casto, D. J., & Walls, R. T. (2007). Le arning with versus without technology in elementary and secondary school. Computers in Human Behavior, 23 798-811.

PAGE 319

305 Technology Counts (2006). Florida the information edge: Usin g data to accelerate achievement Bethesda MD: Editorial Projects in Education Research Center. Retrieved February 17, 2008, from http://www.edweek.org/ew/toc/2006/05/04/index.html TSSA Collaborative (2001). Technology standards for school administrators. Retrieved February 17, 2008, from http://osx.latech.edu/tssa/pdf/tssa.pdf Tyack, D. & Cuban, L. (1995). Tinkering toward utopia: A century of public school reform Cambridge: Harvard University Press. Waxman, H. C., Lin, M. & Michko, G. (2003). A meta-analysis of the effectiveness of teaching and learning with technology on student outcomes Naperville, IL: Learning Point Associates. Retrieved February 17, 2008, from http://www.ncrel.org/tech/effects2/ Weick, K. E. (1976). Educational organizations as loosely coupled systems. Administrative Science Quarterly, 21 (1) 1-19. Weick, K. E. & Quinn, R. E. (1999). Organizational change and development. Annual Review of Psychology, 50 361-386. Wenglinsky H. (1998). Does it computer? The relationship between educational technology and student achievement in mathematics. Princeton, NJ: Educational Testing Service. Retrieved February 17, 2008, from ftp://ftp.ets.org/pub/res/technolog.pdf Wenglinsky, H. (2004). Closing the r acial achievement gap: The role of reforming instructional practices. Education Policy Analysis Archives, 12 (64) 1-22. Retrieved February 17, 2008, from http://epaa.asu.edu/epaa/v12n64/ Wenglinsky, H. (2005). Using technology wisely: The keys to success in schools. New York: Teachers College Press. Wenglinsky, H. (2006). On ideology, causal inference and the reification of statistical methods: Reflections on “Examining instruction, achievement and equity with NAEP mathematics data.” Education Policy Analysis Archives, 14 (17) 1-5. Retrieved February 17, 2008, from http://epaa.asu.edu/epaa/v14n17/ Wheatley, M. J. (1999). Leadership and the new sc ience: Discovering order in a chaotic world. San Francisco: Berret-Koehler Publishers.

PAGE 320

306 Wilensky, U. & Resnick, M. (1999). Thinking in levels: A dynamic systems approach to making sense of the world. Journal of Science Education and Technology, 8 (1) 3-19.

PAGE 321

307 Appendix A: IRB Application for Exempt Certification

PAGE 322

Appendix A: IRB Application for Exempt Certification 308

PAGE 323

Appendix A: IRB Application for Exempt Certification (Continued) 309

PAGE 324

Appendix A: IRB Application for Exempt Certification (Continued) 310

PAGE 325

311 Appendix B: Items from STAR Survey Side-by-side Comparison 2003-04, 2004-05, 2005-06, 2006-07

PAGE 326

Appendix B: Items from STAR Survey 312

PAGE 327

Appendix B: Items from STAR Survey (Continued) 313

PAGE 328

Appendix B: Items from STAR Survey (Continued) 314

PAGE 329

Appendix B: Items from STAR Survey (Continued) 315

PAGE 330

Appendix B: Items from STAR Survey (Continued) 316

PAGE 331

Appendix B: Items from STAR Survey (Continued) 317

PAGE 332

318 Appendix C: Data Preparation Procedures

PAGE 333

Appendix C: Data Preparation Procedures (Continued) 319 Merging Data Files First, the Master School Identification Files (MSI D) for each academic year were entered into the dataset. They were downloaded from the FLDOE website in MS Excel format or text format. The fields that were not of interest to the study were deleted w ith MS Excel. The data were searched for blank cells or cells with only a space in them, and these were replaced with a dot designating missing data. There were no missing schools, county and school identification codes or designated school levels. All school levels that were not elementary, middle, high, or combination schools were changed to the designation “other” school level. These cleaned files were saved in comma delimited text format so they could be imported into SAS 9.1. The number of schools and the primary assignment for accountability by the FLDOE for their school level for each year are listed in Table C 1. In 2005-06, the FLDOE changed the way that schools were classified. Thus, the number of high schools almost doubled, while the number of elementa ry, middle, and combination schools also increased. At the same time the proportion of the other types of schools decreased by more than half. Since the focus of the study was on public elementary, middle, and high schools that primarily served regular education students, additional criteria were used to classify the schools. Therefore, the schools’ service setting, primary function, and designation as a charter school were examined for the 2005-06 and 2006-07 school years. All schools that were not classified as havi ng regular education as their primary service function (e.g., vocational/technical education, adult education, or alternative education) were changed to other types, and all schools that were designated as charter schools were changed to other types. Schools that had a service setting designated as virtual or university lab school were changed to other types. Magnet school information was not available in 2003-04. In 2004-05, magnet programs and schools were listed, but their specialty was not. In 2005-06 and 2006-07, magnet schools with a specialty in technology were listed (see Table C 1). In order to fill in missing information about magnet schools, a request for a list of magnet schools in Florida that had receive d grants in FY 2001 and FY 2004 was ma de to Steve Brock house, the contact person for magnet school assistance at the U.S. Depa rtment of Education (USDOE). Abstracts of these grant applications were obtained, and the list of fund ed magnet schools extracted. The school code for each

PAGE 334

Appendix C: Data Preparation Procedures (Continued) 320 of these schools was obtained from the MSID files. Magnet schools are only funded once by the USDOE; the district must support the continuation of these programs, because the purpose of the magnet grants is to provide seed money for school district s to equip a school with the update d infrastructure necessary to make it an attractive alternative for school choice selection in order to expedite diversifying the student population. It was assumed that technology infrastructure would be an important component for all magnet schools with all themes and foci, and after the building had been equipped with the latest technology, the school would continue to integrate technology into the curriculum. An additional field was created to designate schools that had received one of these grants for the year it was received and for every year after that date. A separate file was created for each year, so that these magnet schools were merged with the MSID files for each year. Table C 1 includes the nu mber of USDOE funded magne t schools in each school year. Table C 1. Master School Identification Files: Schools by Type for each School Year 2003-04 2004-05 2005-06 2006-07 School Level N % N % N % N % Combination 70 1.8 79 1.93 102 2.43 113 2.71 Elementary 1683 43.2 1688 41.23 1712 40.83 1719 41.29 High 429 11.01 431 10.53 460 10.97 461 11.07 Middle 478 12.27 487 11.9 499 11.9 510 12.25 Other Types 1236 31.72 1409 34.42 1420 33.87 1360 32.67 Magnet Program --87 2.13 142 3.39 161 3.87 Magnet School --71 1.73 127 3.03 135 3.24 Technology Magnet ----21 0.5 22 0.53 Total Magnet Schools 158 3.86 269 6.42 296 7.11 Magnet Schools funded by USDOE 28 0.72 58 1.42 60 1.43 60 1.44 Technology Magnets (USDOE) 8 0.21 17 0.42 17 0.41 17 0.41 Total EL, MS, & HS 2590 66.48 1409 34.42 3343 79.73 2690 64.62 Total All Schools 3896 100 4094 100 4193 100 4163 100

PAGE 335

Appendix C: Data Preparation Procedures (Continued) 321 The next set of files brought into the dataset we re the school level mean FCAT NRT scores for reading, mathematics, and writing. These scores were used to measure the outcome variables in the study. The files were obtained for each of the four school years from the FLDOE at the Assessment and School Performance: Florida Comprehensive Assessment Test website and then merged with the MSID files to determine the number of schools at each school level with FCAT scores. The number of schools by school level that participated in the FCAT for Mathematics, Reading, and Writing each y ear are depicted in Table C 2. Over 3000 schools pa rticipated in the FCAT Mathematics, Reading, and Wr iting assessment each year. Less than one percent of the schools were mi ssing a school level, except in 2006-07 for Writing FCAT that had just over one percent of the schools with no designated school level. When the FCAT reading scores for the 2006-07 school year were released, Florida announced that there had been a problem with the score results reported for the Reading FCAT for 3rd grade for the 200506 school year. Since this year and grade had been included in the analysis, a new sample was created excluding the mean FCAT scores fo r third grade of elementary school s. Thus, the mean FCAT reading scores for elementary schools were based on the m ean school scores from fourth and fifth grades. The number of schools did not change, just the number of grades used to create the mean score. Table C 2. Schools Participating in the FCAT each Year 2003-04 2004-05 2005-06 2006-07 School Levels N % N % N % N % Mathematics Missing 12 0.39 8 0.26 4 0.12 5 0.15 Combination 67 2.18 76 2.43 89 2.77 103 3.15 Elementary 1612 52.44 1622 51.92 1638 51.04 1649 50.46 High 406 13.21 406 13 421 13.12 423 12.94 Middle 477 15.52 483 15.46 496 15.46 508 15.54 Other Types 500 16.27 529 16.93 561 17.48 580 17.75 Total Schools 3074 100 3124 100 3209 100 3268 100 Reading Missing 9 0.29 8 0.26 4 0.13 5 0.15 Combination 67 2.2 76 2.45 89 2.79 103 3.17 Elementary 1601 52.46 1609 51.89 1628 51.07 1638 50.48 High 406 13.3 406 13.09 421 13.21 423 13.04 Middle 477 15.63 483 15.58 496 15.56 508 15.65

PAGE 336

Appendix C: Data Preparation Procedures (Continued) 322 2003-04 2004-05 2005-06 2006-07 School Levels N % N % N % N % Other Types 492 16.12 519 16.74 550 17.25 568 17.5 Total Schools 3052 100 3101 100 3188 100 3245 100 Writing Missing 9 0.3 6 0.19 8 0.25 47 1.45 Combination 67 2.21 75 2.42 90 2.83 104 3.2 Elementary 1600 52.81 1615 52.16 1628 51.23 1619 49.86 High 403 13.3 405 13.08 419 13.18 423 13.03 Middle 474 15.64 484 15.63 495 15.58 496 15.28 Other Types 477 15.74 511 16.51 538 16.93 558 17.19 Total Schools 3030 100 3096 100 3178 100 3247 100 The number of missing reported mean school level FCAT test scores for each school level is depicted in Table C 3. For all years, just over four percent of all schools were missing their mean school level FCAT Mathematics and Reading test scores. Th e proportion of missing scores for FCAT Writing was approximately five percent for 2003-04 to 2005-06 school years and just over five percent in 2006-07. Among elementary, middle, and high schools, high schools had the highest proportion of missing FCAT scores, with the highest proportion of missing school level FCAT scores at just over 2.5% in 2006-07. Table C 3. Missing Mean School Level FCAT Test Scores for Mathematics by School Level and Year 2003-04 2004-05 2005-06 2006-07 School Levels N % N % N % N % Mathematics Missing 8 66.67 2 25 0 0 0 0 Combination 2 2.99 4 5.26 10 11.24 10 9.71 Elementary 2 0.12 2 0.12 3 0.18 2 0.12 High 6 1.48 6 1.48 8 1.9 11 2.6 Middle 1 0.21 1 0.21 0 0 0 0 Other Types 111 22.2 117 22.12 110 19.61 115 19.83 Total Schools 130 4.23 132 4.23 131 4.08 138 4.22 Reading Missing 5 55.56 2 25 0 0 0 0 Combination 2 2.99 4 5.26 10 11.24 10 9.71 Elementary 3 0.19 2 0.12 3 0.18 1 0.06 High 6 1.48 6 1.48 8 1.9 11 2.6 Middle 1 0.21 1 0.21 0 0 0 0

PAGE 337

Appendix C: Data Preparation Procedures (Continued) 323 2003-04 2004-05 2005-06 2006-07 School Levels N % N % N % N % Other Types 111 22.56 123 23.7 112 20.36 114 20.07 Total Schools 128 4.19 138 4.45 133 4.17 136 4.19 Writing Missing 6 66.67 2 33.33 2 25 3 6.38 Combination 2 2.99 3 4 12 13.33 13 12.5 Elementary 5 0.31 4 0.25 4 0.25 1 0.06 High 6 1.49 7 1.73 11 2.63 11 2.6 Middle 1 0.21 0 0 0 0 1 0.2 Other Types 131 27.46 126 24.66 129 23.98 142 25.45 Total Schools 151 4.98 142 4.59 158 4.97 171 5.27 Note: Percents are given for the proportion of schools w ith missing scores relative to all schools in the same category Next, the Florida School Indicators Reports (FSIR) files for each school for each year, which were obtained on-line from the FLDOE, were merged. The FSIR contained multiple records for some schools (158 in 2003-04, 194 in 2004-05, and 291 in 2005-06). All of these schools contained combinations of grade levels (e.g., elementary and middle or elementary, middle and high school). The numbers of students and the proportions of students in the various categories were different; however, the total number of staff and the proportion of instructional st aff were the same. To condense these records so there would be only one per school, the means of all variables with proportions were obtained, except for variables with counts that were different for each entry (i.e., the number of students and the number of crimes) the sum was used. The school level designated in the MSID was used to reclassify each school for analysis. Thus some schools were not reclassified as co mbination schools, but as the specific school level used in the MSID. The changes in the number of school s included in these data sets for each year as the records for each school were condensed and then merged with the MSID are delineated in Table C 4. The 2005-06 data were released as this study was being conducted; however, the 2006-07 report was not available. In 2003-04 the original 3037 entries condensed to 2864 schools. After the MSID file was merged and the school level identification was used to classify schools the number of schools in each classification changed. For example th e 1825 original entries for elementary schools in 2003-04 dropped to 1713 schools after being condensed to one record per school, and then after being classified by the school

PAGE 338

Appendix C: Data Preparation Procedures (Continued) 324 level information in the MSID file, the number dr opped to 1610 schools. The reclassification to Combination or Other Types involved 1013 schools. In another example, High Schools in 2003-04 dropped from 521 entries to 460 schools and then dropped to 412 schools when 48 schools’ classification was changed to Combination or Other Types. The change designated as Other Types included the following classifications: Adult, Charter Schools, Department of Juvenile Justice Division, Other Types, Special Education Schools, and Vocational/ Technical School s. Combination Schools provided instruction to students in grades included in both elementary and secondary levels (Bureau of Education Information and Accountability Services, 2007). Table C 4. Number of Schools by School Level included in the FSIR and Merged with the MSID by Year All Entries By School ID With MSID School Level N % N % N % 2003-04 Elementary 1825 60.09 1713 59.81 1610 56.22 High 521 17.16 460 16.06 412 14.39 Middle 691 22.75 533 18.61 474 16.55 Combination 58 2.03 Other Types 310 10.82 Double Entry 143 4.99 Triple Entry 15 0.52 Total Multiple Entries 158 5.51 Total Entries 3037 100 Total Schools 2864 100 2864 100 2004-05 Elementary 1883 60.66 1767 60.72 1640 56.36 High 520 16.75 442 15.19 425 14.6 Middle 701 22.58 529 18.18 483 16.6 Combination 74 2.54 Other Types 288 9.9 Double Entry 150 5.15 Triple Entry 22 0.76 Total Multiple Entries 194 6.39 Total Entries 3037 100 Total Schools 2910 100 2910 100

PAGE 339

Appendix C: Data Preparation Procedures (Continued) 325 All Entries By School ID With MSID School Level N % N % N % 2005-06 Elementary 1728 52.99 1728 58.18 1660 55.89 High 438 13.43 438 14.75 436 14.68 Middle 545 16.71 545 18.35 496 16.7 Combination 88 2.96 Other Types 550 16.87 290 9.76 Double Entry 227 7.64 Triple Entry 32 1.08 Total Multiple Entries 291 8.92 Total Entries 3261 100 Total Schools 2970 100 2970 100 The FSIR did not include any information about minority status or information at the high school level about the proportion of students on Free or Reduced Price Lunch Programs. Therefore, data also were obtained from the AYP Reports on the FLDOE Evaluation and Reporting website for each school for each year. These files were merged so that missing demograp hic information in the FSIR was filled in with data from the AYP. In addition, neither the FSIR nor th e AYP reported the proportion of students who were gifted at the high school level. Consequently, all high schools in the merged FSIR and AYP dataset have missing data for the proportion of gifted students. There are no gifted programs for students at the high school level in Florida, so this information is not tracked. More missing data resulted from variables that were not reported when the population of a school had less than 10 students in order to preserve the privacy of individuals who may be identifiable. For this reason, higher proportions of schools had missing data about proportions of LEP students, students with disabilities, and gifted students. Finally, to determine the number of variables that were missing by school level for each year, this dataset was merged with the MSID files. Thus, any gifted information at the high school level included in the final dataset came from schools that were reclassifi ed to high school level in the MSID file. The number and proportion of missing variables are depicted in Table C 5 for 2003-04, in Table C 6 for 2004-05, and in Table C 7 for 2005-06. High schools had the highest proportion of missing indicator variables when compared to elementary or mi ddle school for all three years.

PAGE 340

Appendix C: Data Preparation Procedures (Continued) 326 Table C 5. Missing Demographic Indicators in the FS IR and AYP by School Level for 2003-04 FSIR, AYP, and MSID FSIR (N=2864) FSIR and AYP (N=2864) EL (N=1620) HS (N=412) MS (N=474) Missing Variables N % N % N % N % N % School Code 0 0 0 0 0 0 0 0 0 0 School Name 0 0 0 0 0 0 0 0 0 0 School Level 0 0 0 0 0 0 0 0 0 0 Total number of instructional staff 0 0 0 0 0 0 0 0 0 0 Total number of students 0 0 0 0 0 0 0 0 0 0 Percent students with disabilities 2864 100 66 2.3 2 0.12 16 3.88 2 0.42 Percent students eligible for free or reduced price lunch program 34 1.19 33 1.15 0 0 14 3.4 2 0.42 Percent LEP students 460 16.06 57 1.99 0 0 19 4.61 0 0 Percent gifted students 341 11.91 340 11.87 147 9.13 46 11.17 28 5.91 Percent of students absent more than 21 days 766 26.75 766 26.75 158 9.81 376 91.26 18 3.8 Total crime incidents 0 0 0 0 0 0 0 0 0 0 Stability percent of students that remain for the year 0 0 54 1.89 2 0.12 16 3.88 2 0.42 Percent of students with in-house suspensions 21 0.73 21 0.73 0 0 12 2.91 2 0.42 Percent of students with out-of-school suspensions 0 0 0 0 0 0 0 0 0 0 Percent of teachers with an advanced degree 0 0 0 0 0 0 0 0 0 0 Average number of years experience 1 0.03 1 0.03 0 0 1 0.24 0 0 Percent of core academic classes taught by out-of-field teachers 265 9.25 265 9.25 14 0.87 18 4.37 5 1.05 Total number of instructional staff 0 0 0 0 0 0 0 0 0 0

PAGE 341

Appendix C: Data Preparation Procedures (Continued) 327 Table C 6. Missing Demographic Indicators in the FS IR and AYP by School Level for 2004-05 FSIR, AYP, and MSID FSIR (N=2910) FSIR and AYP (N=2910) EL (N=1640) HS (N=425) MS (N=483) Missing Variables N % N % N % N % N % School Code 0 0 0 0 0 0 0 0 0 0 School Name 0 0 0 0 0 0 0 0 0 0 School Level 0 0 0 0 0 0 0 0 0 0 Total number of instructional staff 0 0 0 0 0 0 0 0 0 0 Total number of students 0 0 0 0 0 0 0 0 0 0 Percent students with disabilities 2910 100 79 2.71 10 0.61 28 6.59 1 0.21 Percent students eligible for free or reduced price lunch program 30 1.03 30 1.03 5 0.3 18 4.24 0 0 Percent LEP students 442 15.19 37 1.27 0 0 24 5.65 0 0 Percent gifted students 325 11.17 248 8.52 97 5.91 46 10.82 14 2.9 Percent of students absent more than 21 days 810 27.84 810 27.84 204 12.44 394 92.71 19 3.93 Total crime incidents 0 0 0 0 0 0 0 0 0 0 Stability percent of students that remain for the year 0 0 79 2.71 10 0.61 28 6.59 1 0.21 Percent of students with in-house suspensions 24 0.82 24 0.82 6 0.37 14 3.29 0 0 Percent of students with out-of-school suspensions 0 0 0 0 0 0 0 0 0 0 Percent of teachers with an advanced degree 0 0 0 0 0 0 0 0 0 0 Average number of years experience 0 0 0 0 0 0 0 0 0 0 Percent of core academic classes taught by out-offield teachers 295 10.14 295 10.14 6 0.37 20 4.71 2 0.41 Total number of instructional staff 0 0 0 0 0 0 0 0 0 0

PAGE 342

Appendix C: Data Preparation Procedures (Continued) 328 Table C 7. Missing Demographic Indicators in the FS IR and AYP by School Level for 2005-06 FSIR, AYP, and MSID FSIR (N=2970) FSIR and AYP (N=2970) EL (N=1660) HS (N=436) MS (N=496) Missing Variables N % N % N % N % N % School Code 0 0 0 0 0 0 0 0 0 0 School Name 0 0 0 0 0 0 0 0 0 0 School Level 0 0 0 0 0 0 0 0 0 0 Total number of instructional staff 0 0 0 0 0 0 0 0 0 0 Total number of students 0 0 0 0 0 0 0 0 0 0 Percent students with disabilities 2970 100 98 3.3 13 0.78 32 7.34 2 0.4 Percent students eligible for free or reduced price lunch program 46 1.55 44 1.48 5 0.3 17 3.9 1 0.2 Percent LEP students 438 14.75 40 1.35 0 0 22 5.05 0 0 Percent gifted students 328 11.04 234 7.88 80 4.82 42 9.63 10 2.02 Percent of students absent more than 21 days 478 16.09 478 16.09 189 11.39 103 23.62 19 3.83 Total crime incidents 0 0 0 0 0 0 0 0 0 0 Stability percent of students that remain for the year 0 0 98 3.3 13 0.78 32 7.34 2 0.4 Percent of students with in-house suspensions 34 1.14 34 1.14 5 0.3 11 2.52 1 0.2 Percent of students with out-of-school suspensions 0 0 0 0 0 0 0 0 0 0 Percent of teachers with an advanced degree 0 0 0 0 0 0 0 0 0 0 Average number of years experience 2 0.07 2 0.07 0 0 1 0.23 0 0 Percent of core academic classes taught by out-offield teachers 305 10.27 305 10.27 5 0.3 20 4.59 2 0.4 Total number of instructional staff 0 0 0 0 0 0 0 0 0 0

PAGE 343

Appendix C: Data Preparation Procedures (Continued) 329 Last, the technology indicator variables from the responses to the STAR surveys were brought into the data set. When merging together the three data files that comprise the STAR data for 2006-07, it was found that one school had 3 additional duplicates. These were removed before merging the technology data with the MSID file. For the school year 2003-04, there were 251 schools (7.63%) with a school code and some data, but they were missing names and school levels (see Table C 8). After merging the STAR data with the MSID files, the missing school level vari able was decreased to seven schools in 2003-04, no schools in 2004-05 and 2005-06, and no schools in 2006-07. Table C 8. Schools in Original STAR Data and Merged with MSID by School Level and by School Year 2003-04 2004-05 Original STAR STAR & MSID Original STAR STAR & MSID School Level N % N % N % N % Missing 251 7.63 7 0.21 3 0.1 0 0 Combination 66 2.01 67 2.04 59 1.94 64 2.11 Elementary 1640 49.88 1647 50.09 1622 53.43 1605 52.87 High 428 13.02 428 13.02 397 13.08 391 12.88 Middle 478 14.54 478 14.54 476 15.68 470 15.48 Other Types 425 12.93 661 20.1 479 15.78 506 16.67 Total Schools 3288 100 3288 100 3036 100 3036 100 2005-06 2006-07 Missing 1 0.03 0 0 1 0.03 0 0 Combination 77 2.46 81 2.58 96 2.96 97 2.99 Elementary 1665 53.13 1650 52.65 1692 52.21 1667 51.43 High 415 13.24 409 13.05 418 12.9 415 12.8 Middle 501 15.99 492 15.7 517 15.95 502 15.49 Other Types 475 15.16 502 16.02 517 15.95 560 17.28 Total Schools 3134 100 3134 100 3244 100 3244 100 The variables that were going to be used to create the composites for measuring the technology integration indicators were examined for missing responses to items (see Table C 1). The first concern was why there were 272 schools with missing responses for so many items in 2003-04. The next concern was the high level of no response for two items ( Level of school-based technical support and Level of schoolbased instructional technology ) in both the 2003-04 and the 2004-05 school years. Additional items in the

PAGE 344

Appendix C: Data Preparation Procedures (Continued) 330 2003-04 STAR survey with high levels of no response were Level of dependability of the Internet connection and Degree of delays when using the Internet Table C 9. Missing Response for Items in the STARS Survey for each School Year 2003-04 2004-05 2005-06 2006-07 Item N % N % N % N % Modern multi-media computers in media center (desktops) 272 8.27 0 0 2 0.06 31 0.96 Modern multi-media computers in classrooms (desktops) 272 8.27 0 0 2 0.06 31 0.96 Modern multi-media computers in computer labs primarily serving general education (desktops) 272 8.27 0 0 2 0.06 31 0.96 Modern multi-media computers in mobile computer labs (desktops) 0 0 31 0.96 Older computer or not multi-media in media center (desktops) 272 8.27 0 0 2 0.06 31 0.96 Older computer or not multi-media in classrooms (desktops) 272 8.27 0 0 2 0.06 31 0.96 Older computer or not multi-media in computer labs primarily serving general education (desktops) 272 8.27 0 0 2 0.06 31 0.96 Older computer or not multi-media in mobile computer labs (desktops) 272 8.27 0 0 31 0.96 Modern multi-media computers in media center (laptops) 1 0.03 2 0.06 31 0.96 Modern multi-media computers in classrooms (laptops) 0 0 2 0.06 31 0.96 Modern multi-media computers in computer labs primarily serving general education (laptops) 1 0.03 2 0.06 31 0.96 Modern multi-media computers in mobile computer labs (laptops) 272 8.27 0 0 2 0.06 31 0.96

PAGE 345

Appendix C: Data Preparation Procedures (Continued) 331 2003-04 2004-05 2005-06 2006-07 Item N % N % N % N % Older computer or not multi-media in media center (laptops) 0 0 2 0.06 31 0.96 Older computer or not multi-media in classrooms (laptops) 0 0 3 0.1 31 0.96 Older computer or not multi-media in computer labs primarily serving general education (laptops) 0 0 2 0.06 31 0.96 Older computer or not multi-media in mobile computer labs (laptops) 0 0 2 0.06 31 0.96 Percent student computers with concept mapping software 272 8.27 0 0 2 0.06 33 1.02 Percent student computers with graphics software 272 8.27 0 0 2 0.06 33 1.02 Percent student computers with multimedia authoring software 272 8.27 0 0 2 0.06 33 1.02 Percent student computers with presentation software 272 8.27 0 0 2 0.06 33 1.02 Percent student computers with spreadsheet software 272 8.27 0 0 2 0.06 33 1.02 Percent student computers with video editing software 272 8.27 0 0 2 0.06 33 1.02 Percent student computers with web authoring software 272 8.27 0 0 2 0.06 33 1.02 Percent student computers with basic word processing software 272 8.27 0 0 2 0.06 33 1.02 Percent student computers with robust word processing software 272 8.27 0 0 2 0.06 33 1.02 Percent student computers with FCAT Explorer software 272 8.27 0 0 2 0.06 33 1.02 Percent student computers with other test prep tools software 272 8.27 0 0 2 0.06 33 1.02 Percent student computers with Integrated Learning Systems software 272 8.27 0 0 2 0.06 33 1.02 Percent student computers with content-specific skills practice/tutorials software 272 8.27 0 0 2 0.06 33 1.02

PAGE 346

Appendix C: Data Preparation Procedures (Continued) 332 2003-04 2004-05 2005-06 2006-07 Item N % N % N % N % Percent student computers with content-specific simulation software 272 8.27 0 0 2 0.06 33 1.02 Percent student computers with other contentspecific resources software 272 8.27 0 0 2 0.06 33 1.02 Percent student computers with general reference tools software 272 8.27 0 0 2 0.06 33 1.02 Administrative tasks 272 8.27 0 0 3 0.1 32 0.99 Delivery of lessons 272 8.27 0 0 3 0.1 32 0.99 Desktop video production 272 8.27 0 0 3 0.1 32 0.99 Email to other school or district staff 272 8.27 0 0 3 0.1 32 0.99 Email to students or parents 272 8.27 0 0 3 0.1 32 0.99 Presentations 272 8.27 0 0 3 0.1 32 0.99 Research 272 8.27 0 0 3 0.1 32 0.99 Analysis of student assessment information 272 8.27 0 0 3 0.1 32 0.99 Video conferencing 272 8.27 0 0 3 0.1 32 0.99 Webpage publishing 272 8.27 0 0 3 0.1 32 0.99 Frequency students use drill and practice software 272 8.27 0 0 3 0.1 32 0.99 Frequency students use Integrated Learning Systems 272 8.27 0 0 3 0.1 32 0.99 Frequency students use multimedia software 272 8.27 0 0 3 0.1 32 0.99 Frequency students use presentation software 3 0.1 32 0.99 Frequency students use simulation software 272 8.27 0 0 3 0.1 32 0.99 Frequency students use research software 3 0.1 33 1.02 Frequency students use tool-based software 272 8.27 0 0 3 0.1 32 0.99 Percent of technology budget devoted to professional development 272 8.27 0 0 1 0.03 26 0.80 Level of school-based technical support 545 16.58 243 8.00 1 0.03 25 0.77 Level of school-based instructional technology specialist support 972 29.56 666 21.94 1 0.03 25 0.77 Level of dependability of the Internet connection 310 9.43 0 0 2 0.06 33 1.02 Degree of delays when using the Internet 310 9.43 0 0 2 0.06 32 0.99

PAGE 347

Appendix C: Data Preparation Procedures (Continued) 333 2003-04 2004-05 2005-06 2006-07 Item N % N % N % N % Time at your school for a technical issue to be resolved 272 8.27 0 0 2 0.06 32 0.99 Total Schools 3288 100 3036 100 3134 100 3241 100 To examine the 272 missing responses for each of the items, the schools with missing responses for the first item were filtered into a separate data base. The same 272 schools were missing responses for all of the items. Next the types of schools that had no responses were examined. The numbers and proportion of each school level with the non-responses are included in Table C 10. Although these schools accounted for 7.83% of all of the sch ools in the STAR dataset, they we re only 2.65% of the schools that would be used in the research study. To further examine if there were relationships between the schools that had no responses on the survey items, the 272 schools were filtered to include only elementary, middle, and high schools, and then the frequency count of the schools in each county was conducted. Seventeen of the schools were not in any of the 67 counties that woul d be included in the study. The rest of the schools came from 42 different counties. The two counties with the most schools that did not respond to the survey were Hillsborough and Palm Beach, each with 29 sc hools, which was 12.61% of Hillsborough schools and 14.15% of Palm Beach schools. The proportion of schools out of all schoo ls in the districts ranged from a minimum of 1.75% (one school in the county) to a maximum of 100% (15 out of 15 schools). A total of five counties were excluded from the analysis because they had no schoo ls (N=35) respond to any of the technology indicators. Thirteen counties had more than 15% of their schools not respond. Forty-six out of 67 counties (69%) had greater than 90% response rate from their schools.

PAGE 348

Appendix C: Data Preparation Procedures (Continued) 334 Table C 10. Schools with Missing Responses for All STAR Items in 2003-04 School Level N % of Missing Schools % of All Schools Combination 9 3.31 0.27 Elementary 38 13.97 1.16 High 36 13.24 1.09 Middle 13 4.78 0.40 Other Types 176 64.71 5.35 Total EL, HS, & MS 87 31.99 2.65 Total Missing Response Schools 272 100 8.27 Total All Schools 3288 -100 The original wording of the items and response options were examined for additional items with high levels of no response. The two items ( Level of school-based technical support and Level of schoolbased instructional technology ) in both the 2003-04 and the 2004-05 school years did not have any option for schools to designate that they did not have any technical support or instructional technology support. It would seem reasonable that the lack of response was r eally meant to be a response for no level of support. An option for no level of support was added to these two items for the 2005-06 school year, and the lack of response dropped to only one school. To verify that the schools with the lack of response in 2003-04 and 2004-05 meant no level of support, the 2005-06 responses of these schools were examined to see if the same schools chose the option for no support. For the item measuring the level of school-based technical support 61 schools were missing responses in both 2003-04 and 2004-05, but only 13 schools were missing responses in years 2003-04 and 2004-05 and indicated no support in 2005-06, while 78 schools selected the no technical support option for the 2005-06 school year. This suggests that different schools report having no technical support each year. The change in number of schools with missing information for the 2003-04 and 2004-05 school years (61) to only one school not reporting information in 2005-06 suggested that not having the option to select no support impacted the way the schools responded. The responses of the

PAGE 349

Appendix C: Data Preparation Procedures (Continued) 335 schools to the level of school-based instructional technology support suggested that missing responses in 2003-04 and 2004-05 were meant to indicate no level of sup port, because there we re 220 schools with missing responses for both 2003-04 and 2004-05 school year but only one missing response in 2005-06. Seventy-three schools had missing information for 2003-04 and 2004-05 and indicated no level of instructional technology support in 2005-06, which suggests that not having an option for selecting no support impacted how schools responded in 2003-04 and 2004-05. However, there were 271 schools that selected no instructional technology support in 2005-06, which suggests that different schools also report different levels of support each year (see Table C 11). For this study, missing responses to the technolo gy indicators were set to zero, after the 272 schools in 2003-04 with no responses to any of the technology indicators had been deleted from the 200304 dataset. Table C 11. Number of schools only Missing Responses for Level of Support Items for 2003-04 to 2005-06 Item N % Level of school-based technical support No Missing Responses 2432 86.58 Missing 2005-06 1 0.04 No Support 2005-06 61 2.17 Missing one year 283 10.07 Missing 2003-04 and 2004-05 61 2.17 Missing 2004-05 and No Support 2005-06 15 0.53 Missing 2003-04 and No Support 2005-06 4 0.14 Missing 2004-05, 2004-05, and NS 2005-06 13 0.46 Level of school-based instructional technology specialist support No Missing Responses 1912 68.07 Missing 2005-06 1 0.04 No Support 2005-06 271 9.65 Missing one year 649 23.10 Missing 2003-04 and 2004-05 220 7.83 Missing 2004-05 and No Support 2005-06 43 1.53 Missing 2003-04 and No Support 2005-06 33 1.17 Missing 2004-05, 2004-05, and NS 2005-06 73 2.60

PAGE 350

Appendix C: Data Preparation Procedures (Continued) 336 The 272 schools that were missing technology integration indicator responses for all the survey items for the school year 2003-04 were deleted from the dataset for 2003-04. The combined dataset for each school year was filtered to in clude only elementary, middle, and high public schools with primary function of serving regular education students from the 67 counties in Florida. This included 2464 schools in 2003-04 and 2004-05, 2549 schools in 2005-06, and 2586 schools in 2006-07. These files were merged into one dataset that was filtered to include only schools that participated in the STAR survey for all four school years. This resulted in the sample of 2345 schools for the research study. The 2006-07 dataset was sorted by 3 missing variables that seemed to be commonly missing to visually inspect the file for schools that did not resp ond to any of the items of the survey. There were 12 schools that had all missing data. This missing indicators dataset was filtered by having missing information for the three variables. Eight schools were other types, two were elementary, one was middle, and one was high schools. One county had three schools, but two were in the other category. It seemed reasonable to delete these schools fr om the dataset, leaving at total of 3232 schools to merge into the total dataset. Descriptive Statistics To assure that the dataset did not contain any unusual responses, the responses were analyzed using descriptive statistics. Variables with absolute sk ewness values over one and absolute kurtosis values over three were flagged. Results are in the Table C 12. The FCAT Mathematics scores appeared to be approximately normal. The mean score for all schools ranged between 655.53 with standard deviation of 32.45 in 2004-05 to 667.39 with standard deviation of 29.94 in 2006-07. Skewness for all four years ranged between 0.51 and 0.75 and kurtosis ranged between -0.23 and -0.62. However, when examining the data at each school level, high school had high kurtosis (1.65 to 3.65) for all four years. A similar pattern was found with FCAT Reading scores. The mean score for all schools ranged between 657.90 with standard deviation of 25.13 in 2004-05 to 668.81 with standard deviation of 21.86 in 2005-06. Skewness for all four years ranged between 0.28 and 0.72, and kurtosis ranged between -0.50 to 0.02. Again at the high school level, kurtosis ranged between 1.05 and 1.61. The m ean FCAT writing scores for all four years for all

PAGE 351

Appendix C: Data Preparation Procedures (Continued) 337 schools ranged between 3.70 with standard deviation of 0.31 in 2003-04 to 3.91 with standard deviation of 0.32 in 2006-07. The FCAT outcome data were not transformed for any of the analyses. Table C 12. Descriptive Statistics for FCAT Outcome Scores Level and Year N Mean Std. Dev. Min Max Skew Kurt ** FCAT Mathematics All School Levels 2003-04 2313 658.42 30.31 595 792 0.75 -0.23 2004-05 2313 655.53 32.45 592 785 0.56 -0.62 2005-06 2313 664.03 32.39 594 794 0.60 -0.47 2006-07 2313 667.39 29.94 604 781 0.51 -0.48 Elementary 2003-04 1520 640.20 13.91 595 689 0.10 -0.28 2004-05 1520 635.84 16.36 592 697 0.22 -0.16 2005-06 1520 644.78 16.69 594 705 0.21 -0.27 2006-07 1520 650.14 17.05 604 712 0.16 -0.23 High 2003-04 347 710.58 15.03 672 792 0.61 2.39 2004-05 347 708.86 13.25 676 785 0.82 3.25 ** 2005-06 347 717.56 14.79 675 794 0.71 2.28 2006-07 347 716.15 13.79 677 781 0.64 1.65 Middle 2003-04 446 679.90 14.81 641 719 0.11 -0.21 2004-05 446 681.16 14.84 647 723 0.31 -0.18 2005-06 446 687.95 16.02 652 736 0.30 -0.24 2006-07 446 688.22 14.55 657 731 0.33 -0.36 FCAT Reading All School Levels 2003-04 2298 664.61 21.69 613 754 0.28 -0.50 2004-05 2298 657.90 25.13 607 768 0.72 -0.06 2005-06 2298 668.81 21.86 622 767 0.58 0.02 2006-07 2298 667.66 22.33 619 763 0.63 -0.03 Elementary 2003-04 1505 652.97 14.60 613 701 0.05 -0.45 2004-05 1505 643.58 13.24 607 693 0.18 -0.24 2005-06 1505 657.34 13.72 622 704 0.17 -0.35 2006-07 1505 655.57 13.29 619 705 0.19 -0.26 High 2003-04 347 693.19 13.07 650 754 0.25 1.60 2004-05 347 699.84 14.65 658 768 0.42 1.61 2005-06 347 703.93 13.82 668 767 0.48 1.57 2006-07 347 703.48 14.10 669 763 0.42 1.05 Middle

PAGE 352

Appendix C: Data Preparation Procedures (Continued) 338 Level and Year N Mean Std. Dev. Min Max Skew Kurt ** 2003-04 446 681.62 13.81 643 716 -0.03 -0.16 2004-05 446 673.58 13.70 640 711 0.11 -0.10 2005-06 446 680.20 12.08 650 712 0.06 -0.21 2006-07 446 680.57 12.69 650 716 0.13 -0.33 FCAT Writing All School Levels 2003-04 2276 3.70 0.31 3 5 0.20 0.53 2004-05 2276 3.75 0.30 3 5 0.27 0.58 2005-06 2276 3.88 0.31 3 5 0.22 0.43 2006-07 2276 3.91 0.32 3 5 0.27 0.64 Elementary 2003-04 1489 3.64 0.30 3 5 0.09 0.20 2004-05 1489 3.70 0.29 3 5 0.13 0.14 2005-06 1489 3.84 0.31 3 5 0.12 -0.07 2006-07 1489 3.84 0.29 3 5 -0.02 0.24 High 2003-04 348 3.83 0.26 3 5 0.74 3.23 ** 2004-05 348 3.86 0.28 3 5 0.57 2.48 2005-06 348 3.92 0.30 3 5 0.79 2.25 2006-07 348 3.96 0.29 3 5 0.75 1.70 Middle 2003-04 439 3.79 0.33 3 5 0.36 -0.02 2004-05 439 3.82 0.30 3 5 0.55 0.18 2005-06 439 3.98 0.27 3 5 0.59 0.42 2006-07 439 4.13 0.31 4 5 0.55 -0.12 Note. skewness > 1 ** kurtosis > 3 Descriptive statistics for the demographic variables obtained from the Florida School Indicators Report for 2003-04, 2004-05, and 2005-06 that were used in the analysis were computed and are listed in the Table C 13. The FSIR indicators for 2006-07 were not available. Many variables had skewness over 1.0 and kurtosis over 3.0. These variables included the counts of instructional staff, students, and LEP and gifted populations of students that were all positivel y skewed with high kurtosis. Two other categories of variables with high skew and kurtosis were the pr oportions of suspensions an d teachers teaching out of their subject area. Skew ranged from a low of -1.76 to high of 4.09. Kurtosis ranged from a low -1.0 of to a high of 23.11. To determine if this lack of normalit y impacted the analysis, exploratory factor analysis of the variables that were to be grouped into composites we re run with the data in both its original form and after it had been normalized through log transformation.

PAGE 353

Appendix C: Data Preparation Procedures (Continued) 339 Table C 13. Descriptive Statistics of Demographic Variables in the Florida School Indicators Reports Level and Variable N Mean Std Dev Min Max Skew Kurt 2003-04 All School Levels Total number of instructional staff 2327 58.74 30.04 0 213 1.87 4.54 ** Total number of students 2327 978.23 611.05 0 4655 2.15 6.06 ** Percent students with disabilities 2324 15.54 5.43 0.5 40.6 0.68 1.39 ** Percent students eligible for free or reduced price lunch program 2324 52.26 25.32 1 100 0.10 -0.94 Percent LEP students 2153 8.73 10.68 0 63.3 1.97 3.97 ** Percent gifted students 1835 4.97 5.75 0.1 52.3 2.91 12.30 ** Percent of students absent more than 21 days 2327 8.31 5.51 0 38.9 1.56 3.40 ** Total crime incidents 2324 0.07 0.11 0 1 3.11 15.17 ** Stability percent of students who remain for the year 2325 93.71 2.79 71.3 100 -0.97 2.70 ** Percent of students with inhouse suspensions 2327 7.20 10.94 0 63.6 1.60 1.64 ** Percent of students with outof-school suspensions 2327 7.19 8.05 0 75.9 1.81 4.69 ** Percent of teachers with an advanced degree 2327 33.47 11.17 0 78.2 0.34 0.26 Average number of years experience 2302 12.62 3.24 3.8 33.7 0.50 1.81 ** Percent of core academic classes taught by out-of-field teachers 2327 5.99 9.58 0 73.7 2.88 10.91 ** Elementary Total number of instructional staff 1527 46.34 13.40 0 123 0.31 1.17 ** Total number of students 1527 714.54 240.46 16 2328 0.62 1.82 ** Percent students with disabilities 1527 16.09 5.72 1.2 40.6 0.80 1.22 ** Percent students eligible for free or reduced price lunch program 1527 57.12 26.17 1 100 -0.14 -1.00 ** Percent LEP students 1399 10.79 12.21 0.1 63.3 1.56 1.97 ** Percent gifted students 1380 4.30 5.40 0.1 52.3 3.41 17.36 ** Percent of students absent more than 21 days 1527 6.29 3.15 0 24.5 0.82 1.31 ** Total crime incidents 1526 0.04 0.10 0 1 4.09 23.11 ** Stability percent of students who remain for the year 1527 93.96 2.82 71.3 100 -1.08 3.30 **

PAGE 354

Appendix C: Data Preparation Procedures (Continued) 340 Level and Variable N Mean Std Dev Min Max Skew Kurt Percent of students with inhouse suspensions 1527 1.77 3.55 0 36.55 4.03 22.25 ** Percent of students with outof-school suspensions 1527 3.26 3.71 0 36.1 2.67 11.46 ** Percent of teachers with an advanced degree 1527 32.35 11.46 0 70.8 0.51 0.25 Average number of years experience 1514 12.59 3.36 3.8 33.7 0.52 1.78 ** Percent of core academic classes taught by out-of-field teachers 1527 5.72 10.40 0 73.7 3.04 11.10 ** High Total number of instructional staff 352 104.24 43.87 0 213 0.03 -0.28 Total number of students 352 1909.10 905.16 0 4655 0.35 0.09 Percent students with disabilities 349 13.25 4.45 0.7 31.05 0.03 0.89 Percent students eligible for free or reduced price lunch program 349 35.51 17.14 1.8 93.3 0.45 0.17 Percent LEP students 329 4.50 4.76 0 23.1 1.47 1.75 ** Percent gifted students 22 7.04 9.92 0.4 39.8 2.54 6.35 ** Percent of students absent more than 21 days 352 13.81 7.71 0 35.6 0.44 -0.19 Total crime incidents 350 0.12 0.10 0 1 2.67 15.29 ** Stability percent of students who remain for the year 350 92.55 2.52 82.5 99.4 -0.48 1.26 ** Percent of students with inhouse suspensions 352 16.84 13.11 0 51.3 0.26 -0.88 Percent of students with outof-school suspensions 352 13.23 7.30 0 42 0.86 1.07 ** Percent of teachers with an advanced degree 352 38.74 9.98 0 78.2 -0.44 2.82 ** Average number of years experience 345 13.43 2.80 4.4 29.7 0.43 3.36 ** Percent of core academic classes taught by out-of-field teachers 352 5.77 7.72 0 54.1 2.63 10.28 ** Middle Total number of instructional staff 448 65.26 19.28 8 124 -0.01 0.21 Total number of students 448 1145.61 392.71 178 2662 0.39 0.63 Percent students with disabilities 448 15.47 4.57 0.5 27.9 -0.13 0.18 Percent students eligible for free or reduced price lunch program 448 48.78 21.32 3.7 100 0.11 -0.66 Percent LEP students 425 5.21 5.43 0.1 32.5 1.70 3.54 ** Percent gifted students 433 6.99 6.05 0.1 37.3 2.04 5.19 **

PAGE 355

Appendix C: Data Preparation Procedures (Continued) 341 Level and Variable N Mean Std Dev Min Max Skew Kurt Percent of students absent more than 21 days 448 10.89 5.75 0 38.9 0.90 2.04 ** Total crime incidents 448 0.09 0.11 0 0.8 2.22 8.56 ** Stability percent of students who remain for the year 448 93.78 2.67 79.5 98.9 -1.25 3.27 ** Percent of students with inhouse suspensions 448 18.15 12.27 0 63.6 0.12 -0.61 Percent of students with outof-school suspensions 448 15.84 9.56 0.1 75.9 1.14 3.48 ** Percent of teachers with an advanced degree 448 33.16 9.81 10.5 68 0.46 0.32 Average number of years experience 443 12.08 2.99 4.1 24.8 0.58 1.20 ** Percent of core academic classes taught by out-of-field teachers 448 7.07 7.71 0 45.3 1.45 2.38 ** 2004-05 All School Levels Total number of instructional staff 2341 61.65 31.10 0 228 1.85 4.66 ** Total number of students 2341 978.12 613.01 0 4723 2.14 5.98 ** Percent students with disabilities 2339 15.32 5.33 0.4 43.9 0.72 1.96 ** Percent students eligible for free or reduced price lunch program 2338 52.71 24.09 0.9 102.1 0.01 -0.90 Percent LEP students 2232 8.52 10.65 0 85.5 2.11 5.07 ** Percent gifted students 1811 4.96 5.85 0.1 54.9 2.96 12.64 ** Percent of students absent more than 21 days 2341 9.31 6.41 0 47.5 1.51 3.23 ** Total crime incidents 2333 0.05 0.09 0 0.8 2.98 14.17 ** Stability percent of students who remain for the year 2338 93.15 2.98 78.1 100 -0.76 0.92 Percent of students with inhouse suspensions 2341 6.89 10.46 0 51.5 1.66 1.88 ** Percent of students with outof-school suspensions 2341 6.86 7.80 0 57.9 1.80 3.96 ** Percent of teachers with an advanced degree 2341 33.10 10.77 0 100 0.40 0.80 Average number of years experience 2336 12.45 3.19 2.8 30 0.34 1.03 ** Percent of core academic classes taught by out-of-field teachers 2341 6.43 11.09 0 84.3 2.71 8.55 ** Elementary Total number of instructional staff 1540 49.11 14.64 0 126 0.37 1.02 ** Total number of students 1540 714.62 246.99 8 2260 0.62 1.43 **

PAGE 356

Appendix C: Data Preparation Procedures (Continued) 342 Level and Variable N Mean Std Dev Min Max Skew Kurt Percent students with disabilities 1540 15.83 5.60 1.6 43.9 0.90 1.93 ** Percent students eligible for free or reduced price lunch program 1540 56.61 24.83 0.9 100 -0.21 -0.93 Percent LEP students 1461 10.48 12.19 0 85.5 1.70 2.87 ** Percent gifted students 1360 4.25 5.48 0.1 54.9 3.50 18.21 ** Percent of students absent more than 21 days 1540 7.63 4.79 0.2 47.5 1.88 6.46 ** Total crime incidents 1536 0.03 0.09 0 0.8 4.03 22.62 ** Stability percent of students who remain for the year 1539 93.47 2.95 80.4 100 -0.85 0.96 Percent of students with inhouse suspensions 1540 1.73 3.28 0 37 3.81 21.34 ** Percent of students with outof-school suspensions 1540 3.15 3.54 0 27 2.13 5.57 ** Percent of teachers with an advanced degree 1540 32.00 10.87 0 71.5 0.49 0.36 Average number of years experience 1539 12.43 3.26 2.8 30 0.30 0.79 Percent of core academic classes taught by out-of-field teachers 1540 6.71 12.52 0 84.3 2.61 7.17 ** High Total number of instructional staff 353 108.41 45.35 0 228 0.07 -0.09 Total number of students 353 1931.27 896.26 0 4723 0.33 0.14 Percent students with disabilities 351 13.16 4.51 0.4 27.6 -0.19 0.41 Percent students eligible for free or reduced price lunch program 350 38.85 18.02 3.8 102.1 0.59 0.69 Percent LEP students 335 4.35 4.65 0 26 1.53 2.11 ** Percent gifted students 20 7.60 10.36 0.7 39.6 2.20 4.50 ** Percent of students absent more than 21 days 353 14.02 8.46 0 38.4 0.50 -0.16 Total crime incidents 350 0.11 0.09 0 0.8 2.64 13.08 ** Stability percent of students who remain for the year 351 91.64 2.91 78.1 98.8 -0.59 2.01 ** Percent of students with inhouse suspensions 353 15.73 12.50 0 47.1 0.40 -0.75 Percent of students with outof-school suspensions 353 12.50 7.49 0 39 0.85 0.85 Percent of teachers with an advanced degree 353 38.23 10.45 0 100 0.15 4.89 ** Average number of years experience 350 13.24 2.88 5.4 29.8 0.57 3.31 **

PAGE 357

Appendix C: Data Preparation Procedures (Continued) 343 Level and Variable N Mean Std Dev Min Max Skew Kurt Percent of core academic classes taught by out-of-field teachers 353 5.16 7.07 0 42.3 2.17 5.52 ** Middle Total number of instructional staff 448 67.89 19.94 9 129 0.11 0.28 Total number of students 448 1132.88 379.43 172 2558 0.39 0.48 Percent students with disabilities 448 15.26 4.48 0.5 28.7 -0.16 0.20 Percent students eligible for free or reduced price lunch program 448 50.15 21.15 3.3 100 0.04 -0.61 Percent LEP students 436 5.16 5.42 0.1 36.4 1.90 4.90 ** Percent gifted students 431 7.09 6.18 0.1 39.1 2.09 5.56 ** Percent of students absent more than 21 days 448 11.37 6.96 0 44.4 1.09 2.77 ** Total crime incidents 447 0.07 0.09 0 0.7 1.90 7.30 ** Stability percent of students who remain for the year 448 93.23 2.78 81.2 99.4 -0.78 0.80 Percent of students with inhouse suspensions 448 17.65 11.98 0 51.5 0.13 -0.68 Percent of students with outof-school suspensions 448 15.20 9.41 0 57.9 1.00 1.64 ** Percent of teachers with an advanced degree 448 32.82 9.47 10.7 60.7 0.32 -0.07 Average number of years experience 447 11.89 3.07 4.8 26.2 0.46 0.85 Percent of core academic classes taught by out-of-field teachers 448 6.49 7.93 0 40.3 1.60 2.25 ** 2005-06 All School Levels Total number of instructional staff 2341 62.65 31.26 0 250 1.89 5.09 ** Total number of students 2341 968.70 602.43 0 5060 2.17 6.19 ** Percent students with disabilities 2338 15.32 5.45 0.3 72.55 1.48 9.51 ** Percent students eligible for free or reduced price lunch program 2337 52.31 23.87 1.7 100 -0.04 -0.96 Percent LEP students 2248 8.79 10.71 0 78.9 2.05 4.69 ** Percent gifted students 2098 4.88 5.78 0 57.2 3.03 13.84 ** Percent of students absent more than 21 days 2341 9.38 6.59 0 57.6 1.59 3.79 ** Total crime incidents 2332 0.05 0.09 0 0.9 2.99 13.78 ** Stability percent of students who remain for the year 2339 92.99 3.22 63.1 99.5 -1.43 5.86 **

PAGE 358

Appendix C: Data Preparation Procedures (Continued) 344 Level and Variable N Mean Std Dev Min Max Skew Kurt Percent of students with inhouse suspensions 2341 6.66 10.45 0 56.5 1.76 2.32 ** Percent of students with outof-school suspensions 2341 6.44 7.77 0 64.9 2.03 5.51 ** Percent of teachers with an advanced degree 2341 31.72 11.93 0 100 -0.09 0.95 Average number of years experience 2338 12.64 3.21 4 31.3 0.49 1.54 ** Percent of core academic classes taught by out-of-field teachers 2341 6.95 11.52 0 80 2.51 7.20 ** Elementary Total number of instructional staff 1540 50.58 15.43 1 135 0.49 1.19 ** Total number of students 1540 716.14 243.58 11 2258 0.61 1.42 ** Percent students with disabilities 1540 16.05 5.77 1.9 72.55 1.77 10.39 ** Percent students eligible for free or reduced price lunch program 1540 56.13 24.69 1.7 100 -0.25 -0.97 Percent LEP students 1472 10.78 12.21 0.1 78.9 1.64 2.62 ** Percent gifted students 1374 4.16 5.55 0.1 57.2 3.69 20.49 ** Percent of students absent more than 21 days 1540 7.55 5.12 0.1 57.6 2.33 10.04 ** Total crime incidents 1535 0.04 0.09 0 0.9 3.87 20.31 ** Stability percent of students who remain for the year 1540 93.25 3.26 63.1 99.5 -1.76 8.18 ** Percent of students with inhouse suspensions 1540 1.75 3.24 0 26.4 3.51 15.83 ** Percent of students with outof-school suspensions 1540 3.13 3.81 0 29.6 2.66 10.12 ** Percent of teachers with an advanced degree 1540 30.62 11.89 0 71.2 -0.01 0.56 Average number of years experience 1540 12.63 3.29 4 31 0.51 1.29 ** Percent of core academic classes taught by out-of-field teachers 1540 7.01 12.76 0 80 2.44 6.12 ** High Total number of instructional staff 353 109.52 45.80 0 250 0.15 0.13 Total number of students 353 1930.55 877.98 0 5060 0.28 0.27 Percent students with disabilities 350 13.12 4.52 0.3 30.85 -0.08 0.66 Percent students eligible for free or reduced price lunch program 349 38.67 17.16 2.5 93.7 0.17 -0.34 Percent LEP students 337 4.41 4.65 0 28.3 1.61 2.78 ** Percent gifted students 291 4.74 5.03 0 37.4 2.67 11.00 **

PAGE 359

Appendix C: Data Preparation Procedures (Continued) 345 Level and Variable N Mean Std Dev Min Max Skew Kurt Percent of students absent more than 21 days 353 15.15 8.64 0 48.7 0.59 0.64 Total crime incidents 349 0.10 0.08 0 0.7 2.37 11.49 ** Stability percent of students who remain for the year 351 91.81 2.96 77 99 -0.57 1.68 ** Percent of students with inhouse suspensions 353 14.85 12.90 0 51.8 0.46 -0.80 Percent of students with outof-school suspensions 353 11.43 8.06 0 43.7 0.97 1.37 ** Percent of teachers with an advanced degree 353 36.64 12.20 0 100 -0.40 3.50 ** Average number of years experience 350 13.29 2.80 5.3 31.3 0.64 4.86 ** Percent of core academic classes taught by out-of-field teachers 353 6.68 8.34 0 48.1 2.12 5.78 ** Middle Total number of instructional staff 448 67.20 19.68 9 130 0.17 0.31 Total number of students 448 1078.98 354.95 154 2300 0.38 0.47 Percent students with disabilities 448 14.53 4.30 0.8 28.7 -0.08 0.25 Percent students eligible for free or reduced price lunch program 448 49.79 21.16 3.9 100 0.02 -0.79 Percent LEP students 439 5.45 5.69 0.1 37.4 2.05 5.78 ** Percent gifted students 433 7.27 6.34 0.1 39.5 2.10 5.78 ** Percent of students absent more than 21 days 448 11.11 6.11 0 35.7 0.67 0.62 Total crime incidents 448 0.07 0.09 0 0.6 1.77 4.75 ** Stability percent of students who remain for the year 448 93.04 3.08 78.9 99 -1.09 2.14 ** Percent of students with inhouse suspensions 448 17.08 12.64 0 56.5 0.24 -0.65 Percent of students with outof-school suspensions 448 13.90 10.07 0 64.9 1.12 2.31 ** Percent of teachers with an advanced degree 448 31.63 10.81 0 62 -0.25 0.70 Average number of years experience 448 12.18 3.16 4.6 25.3 0.46 0.97 Percent of core academic classes taught by out-of-field teachers 448 6.96 8.88 0 72.1 2.26 8.51 ** Note. skewness > 1 ** kurtosis > 3

PAGE 360

Appendix C: Data Preparation Procedures (Continued) 346 Descriptive statistics were calculated for the tech nology integration variables for all four years (2003-04, 2004-05, 2005-06, and 2006-07) and are listed in Table C 14. Skewness for the variables for all four years ranged between -9.86 and 47.61; and Kurtosis ranged between -1.65 and 2287.37. All of the variables with the highest absolute amounts of skewness and kurtosis were counts of computers. Table C 14. Descriptive Statistics of the Technology Integratio n Variables from the Florida Innovates (STAR) Survey Label N Mean Std Dev Min Max Skew Kurt 2003-04 All School Levels Modern multi-media computers in Media center (desktops) 2327 13.32 13.13 0 120 2.34 8.67 Modern multi-media computers in Classrooms (desktops) 2327 106.39 104.12 0 1735 3.52 32.13 Modern multi-media computers in Computer labs primarily serving general education (desktops) 2327 42.61 47.08 0 525 3.51 20.43 Modern multi-media computers in Mobile computer labs (laptops) 2327 8.24 22.21 0 352 5.75 54.82 Older computer or not multi-media in Media center (desktops) 2327 3.26 7.27 0 92 4.72 32.66 Older computer or not multi-media in Classrooms (desktops) 2327 46.29 63.80 0 600 2.83 12.84 Older computer or not multi-media in Computer labs primarily serving general education (desktops) 2327 8.56 22.24 0 271 4.79 34.95 Older computer or not multi-media in Mobile computer labs (laptops) 2327 1.03 7.81 0 176 12.36 196.89 Percent student computers with Concept mapping software 2327 2.29 1.36 1 5 0.92 -0.42 Percent student computers with Graphics software 2327 4.39 1.12 1 5 -1.71 1.61 Percent student computers with Multimedia authoring software 2327 2.56 1.40 1 5 0.65 -0.90 Percent student computers with Presentation software 2327 4.16 1.19 1 5 -1.16 0.03 Percent student computers with Spreadsheet software 2327 4.44 1.02 1 5 -1.81 2.26 Percent student computers with Video editing software 2327 2.06 1.03 1 5 1.40 1.81 Percent student computers with Web authoring software 2327 2.02 1.10 1 5 1.51 1.82 Percent student computers with Basic word processing software 2327 4.78 0.76 1 5 -3.81 13.90

PAGE 361

Appendix C: Data Preparation Procedures (Continued) 347 Label N Mean Std Dev Min Max Skew Kurt Percent student computers with Robust word processing software 2327 4.60 0.92 1 5 -2.45 5.19 Percent student computers with FCAT Explorer software 2327 4.08 1.28 1 5 -1.18 0.08 Percent student computers with Other test prep tools software 2327 3.20 1.52 1 5 -0.11 -1.50 Percent student computers with Integrated Learning Systems software 2327 3.19 1.54 1 5 -0.11 -1.51 Percent student computers with Content-specific skills practice/tutorials software 2327 3.76 1.31 1 5 -0.57 -1.07 Percent student computers with Content-specific simulation software 2327 2.87 1.45 1 5 0.26 -1.33 Percent student computers with Other content-specific resources software 2327 3.13 1.46 1 5 0.02 -1.44 Percent student computers with General Reference tools software 2327 3.93 1.34 1 5 -0.87 -0.67 Administrative tasks 2327 4.31 1.04 1 5 -1.33 0.54 Delivery of lessons 2327 2.94 0.98 1 5 0.51 -0.65 Desktop video production 2327 1.84 0.68 1 5 1.15 3.60 Email to other school or district staff 2327 4.61 0.88 1 5 -2.31 4.40 Email to students or parents 2327 3.07 1.22 1 5 0.19 -1.11 Presentations 2327 2.82 0.98 1 5 0.70 -0.36 Research 2327 3.86 1.00 1 5 -0.48 -0.73 Analysis of student assessment information 2327 3.85 1.18 1 5 -0.61 -0.86 Video conferencing 2327 1.23 0.51 1 5 2.92 12.63 Webpage publishing 2327 2.00 0.78 1 5 1.56 4.15 Degree students use Drill and practice software 2327 2.05 0.96 1 5 0.57 -0.42 Degree students use Integrated Learning Systems 2327 2.03 1.17 1 5 1.07 0.30 Degree students use Multimedia 2327 3.00 0.95 1 5 -0.14 -0.34 Degree students use Simulation software 2327 3.42 0.95 1 5 -0.24 -0.20 Degree students use Tool-based software 2327 2.13 1.02 1 5 0.55 -0.49 % of technology $ devoted to professional development 2327 13.54 14.91 0 100 2.19 7.80 Level of school-based technical support 2177 2.21 1.07 0 4 0.40 -1.07 Level of school-based instructional technology specialist support 1871 2.10 1.14 1 4 0.49 -1.26 Level of dependability of the Internet connection 2324 4.17 0.71 1 5 -0.51 0.10

PAGE 362

Appendix C: Data Preparation Procedures (Continued) 348 Label N Mean Std Dev Min Max Skew Kurt Degree of delays when using the Internet 2324 3.85 0.63 1 5 -2.00 6.18 Time at your school for a technical issue to be resolved 2327 3.70 1.14 1 5 -0.24 -1.36 Elementary Modern multi-media computers in Media center (desktops) 1531 8.67 7.02 0 60 1.79 5.32 Modern multi-media computers in Classrooms (desktops) 1531 96.51 69.36 0 420 1.04 1.50 Modern multi-media computers in Computer labs primarily serving general education (desktops) 1531 28.11 21.70 0 215 1.10 4.38 Modern multi-media computers in Mobile computer labs (laptops) 1531 5.22 13.85 0 150 4.14 24.83 Older computer or not multi-media in Media center (desktops) 1531 2.40 4.39 0 56 3.62 24.14 Older computer or not multi-media in Classrooms (desktops) 1531 45.74 51.94 0 340 1.84 4.74 Older computer or not multi-media in Computer labs primarily serving general education (desktops) 1531 5.82 14.17 0 120 3.02 10.70 Older computer or not multi-media in Mobile computer labs (laptops) 1531 0.81 6.60 0 176 16.34 367.22 Percent student computers with Concept mapping software 1531 2.34 1.41 1 5 0.81 -0.70 Percent student computers with Graphics software 1531 4.36 1.16 1 5 -1.69 1.56 Percent student computers with Multimedia authoring software 1531 2.58 1.43 1 5 0.59 -1.03 Percent student computers with Presentation software 1531 3.93 1.29 1 5 -0.80 -0.75 Percent student computers with Spreadsheet software 1531 4.29 1.13 1 5 -1.46 0.96 Percent student computers with Video editing software 1531 2.02 1.08 1 5 1.34 1.39 Percent student computers with Web authoring software 1531 1.88 1.07 1 5 1.73 2.64 Percent student computers with Basic word processing software 1531 4.74 0.83 1 5 -3.44 11.08 Percent student computers with Robust word processing software 1531 4.48 1.03 1 5 -2.01 3.02 Percent student computers with FCAT Explorer software 1531 4.04 1.31 1 5 -1.18 0.08 Percent student computers with Other test prep tools software 1531 3.22 1.59 1 5 -0.19 -1.55

PAGE 363

Appendix C: Data Preparation Procedures (Continued) 349 Label N Mean Std Dev Min Max Skew Kurt Percent student computers with Integrated Learning Systems software 1531 3.35 1.60 1 5 -0.34 -1.48 Percent student computers with Content-specific skills practice/tutorials software 1531 4.03 1.25 1 5 -1.05 -0.15 Percent student computers with Content-specific simulation software 1531 2.93 1.54 1 5 0.13 -1.50 Percent student computers with Other content-specific resources software 1531 3.28 1.51 1 5 -0.20 -1.45 Percent student computers with General Reference tools software 1531 3.89 1.35 1 5 -0.85 -0.67 Administrative tasks 1531 4.11 1.13 1 5 -0.97 -0.35 Delivery of lessons 1531 2.85 0.99 1 5 0.63 -0.55 Desktop video production 1531 1.75 0.70 1 5 1.23 3.64 Email to other school or district staff 1531 4.55 0.93 1 5 -2.08 3.27 Email to students or parents 1531 2.90 1.19 1 5 0.35 -0.94 Presentations 1531 2.67 0.95 1 5 0.91 0.13 Research 1531 3.76 1.02 1 5 -0.39 -0.81 Analysis of student assessment information 1531 3.90 1.16 1 5 -0.68 -0.73 Video conferencing 1531 1.19 0.50 1 5 3.49 16.93 Webpage publishing 1531 1.92 0.75 1 5 1.47 4.24 Degree students use Drill and practice software 1531 1.89 0.92 1 5 0.80 0.04 Degree students use Integrated Learning Systems 1531 1.95 1.22 1 5 1.24 0.53 Degree students use Multimedia 1531 3.15 0.93 1 5 -0.22 -0.21 Degree students use Simulation software 1531 3.40 0.98 1 5 -0.18 -0.35 Degree students use Tool-based software 1531 2.33 1.04 1 5 0.34 -0.61 % of technology $ devoted to professional development 1531 13.13 15.19 0 100 2.26 7.94 Level of school-based technical support 1430 2.14 1.03 1 4 0.44 -0.97 Level of school-based instructional technology specialist support 1218 2.04 1.13 1 4 0.53 -1.21 Level of dependability of the Internet connection 1529 4.15 0.70 1 5 -0.46 0.04 Degree of delays when using the Internet 1529 3.85 0.66 1 5 -1.86 5.30 Time at your school for a technical issue to be resolved 1531 3.64 1.15 1 5 -0.16 -1.41

PAGE 364

Appendix C: Data Preparation Procedures (Continued) 350 Label N Mean Std Dev Min Max Skew Kurt High Modern multi-media computers in Media center (desktops) 350 28.44 19.05 0 120 1.22 2.94 Modern multi-media computers in Classrooms (desktops) 350 147.48 185.97 0 1735 2.95 16.24 Modern multi-media computers in Computer labs primarily serving general education (desktops) 350 87.06 83.63 0 525 2.03 5.42 Modern multi-media computers in Mobile computer labs (laptops) 350 17.17 40.43 0 352 4.38 25.24 Older computer or not multi-media in Media center (desktops) 350 5.22 11.72 0 78 3.25 12.12 Older computer or not multi-media in Classrooms (desktops) 350 48.39 91.42 0 600 3.19 12.16 Older computer or not multi-media in Computer labs primarily serving general education (desktops) 350 14.89 35.52 0 271 3.97 19.50 Older computer or not multi-media in Mobile computer labs (laptops) 350 1.13 9.19 0 120 10.66 123.71 Percent student computers with Concept mapping software 350 2.09 1.19 1 5 1.32 0.94 Percent student computers with Graphics software 350 4.43 1.06 1 5 -1.66 1.28 Percent student computers with Multimedia authoring software 350 2.53 1.29 1 5 0.81 -0.48 Percent student computers with Presentation software 350 4.67 0.72 2 5 -2.29 4.44 Percent student computers with Spreadsheet software 350 4.75 0.64 1 5 -2.84 8.23 Percent student computers with Video editing software 350 2.18 0.77 1 5 2.18 5.86 Percent student computers with Web authoring software 350 2.56 1.09 1 5 1.24 0.47 Percent student computers with Basic word processing software 350 4.95 0.36 2 5 -7.39 55.13 Percent student computers with Robust word processing software 350 4.86 0.49 2 5 -4.31 20.07 Percent student computers with FCAT Explorer software 350 4.16 1.18 1 5 -1.08 -0.27 Percent student computers with Other test prep tools software 350 3.16 1.30 1 5 0.28 -1.40 Percent student computers with Integrated Learning Systems software 350 2.65 1.20 1 5 0.74 -0.42 Percent student computers with Content-specific skills practice/tutorials software 350 3.07 1.15 1 5 0.51 -1.08

PAGE 365

Appendix C: Data Preparation Procedures (Continued) 351 Label N Mean Std Dev Min Max Skew Kurt Percent student computers with Content-specific simulation software 350 2.68 1.14 1 5 0.81 -0.32 Percent student computers with Other content-specific resources software 350 2.71 1.22 1 5 0.70 -0.66 Percent student computers with General Reference tools software 350 3.99 1.30 1 5 -0.89 -0.70 Administrative tasks 350 4.68 0.75 2 5 -2.43 4.96 Delivery of lessons 350 3.15 0.87 1 5 0.22 -0.64 Desktop video production 350 2.08 0.58 1 5 1.63 6.06 Email to other school or district staff 350 4.71 0.75 1 5 -2.80 7.38 Email to students or parents 350 3.51 1.17 1 5 -0.20 -1.18 Presentations 350 3.19 0.96 1 5 0.33 -0.78 Research 350 4.10 0.92 2 5 -0.76 -0.31 Analysis of student assessment information 350 3.68 1.23 1 5 -0.39 -1.19 Video conferencing 350 1.37 0.54 1 5 1.52 4.67 Webpage publishing 350 2.21 0.69 1 5 2.14 6.06 Degree students use Drill and practice software 350 2.37 0.97 1 5 0.09 -0.90 Degree students use Integrated Learning Systems 350 2.31 1.03 1 5 0.62 0.03 Degree students use Multimedia 350 2.47 0.84 1 5 0.00 -0.43 Degree students use Simulation software 350 3.57 0.85 1 5 -0.53 0.62 Degree students use Tool-based software 350 1.44 0.69 1 4 1.52 1.78 % of technology $ devoted to professional development 350 14.70 14.65 0 100 2.42 10.26 Level of school-based technical support 338 2.36 1.16 1 4 0.30 -1.37 Level of school-based instructional technology specialist support 289 2.16 1.19 1 4 0.45 -1.36 Level of dependability of the Internet connection 350 4.25 0.68 2 5 -0.57 0.14 Degree of delays when using the Internet 350 3.89 0.52 1 5 -2.56 10.86 Time at your school for a technical issue to be resolved 350 3.81 1.11 2 5 -0.35 -1.26 Middle Modern multi-media computers in Media center (desktops) 446 17.40 13.28 0 92 1.65 4.78 Modern multi-media computers in Classrooms (desktops) 446 108.07 106.46 0 564 1.51 2.68

PAGE 366

Appendix C: Data Preparation Procedures (Continued) 352 Label N Mean Std Dev Min Max Skew Kurt Modern multi-media computers in Computer labs primarily serving general education (desktops) 446 57.52 44.50 0 350 1.56 5.34 Modern multi-media computers in Mobile computer labs (laptops) 446 11.61 23.04 0 128 2.57 7.30 Older computer or not multi-media in Media center (desktops) 446 4.69 9.75 0 92 3.77 20.97 Older computer or not multi-media in Classrooms (desktops) 446 46.50 73.77 0 531 2.57 8.69 Older computer or not multi-media in Computer labs primarily serving general education (desktops) 446 12.98 28.79 0 260 3.48 17.42 Older computer or not multi-media in Mobile computer labs (laptops) 446 1.72 10.10 0 120 7.81 71.03 Percent student computers with Concept mapping software 446 2.29 1.32 1 5 1.04 -0.08 Percent student computers with Graphics software 446 4.45 1.05 1 5 -1.78 1.80 Percent student computers with Multimedia authoring software 446 2.52 1.35 1 5 0.76 -0.65 Percent student computers with Presentation software 446 4.54 0.84 1 5 -1.95 3.26 Percent student computers with Spreadsheet software 446 4.70 0.69 1 5 -2.60 6.62 Percent student computers with Video editing software 446 2.13 1.02 1 5 1.50 2.12 Percent student computers with Web authoring software 446 2.08 1.07 1 5 1.46 1.87 Percent student computers with Basic word processing software 446 4.80 0.73 1 5 -3.98 15.14 Percent student computers with Robust word processing software 446 4.79 0.68 1 5 -3.83 15.20 Percent student computers with FCAT Explorer software 446 4.16 1.23 1 5 -1.18 0.05 Percent student computers with Other test prep tools software 446 3.15 1.47 1 5 -0.01 -1.44 Percent student computers with Integrated Learning Systems software 446 3.07 1.45 1 5 0.07 -1.41 Percent student computers with Content-specific skills practice/tutorials software 446 3.38 1.32 1 5 -0.07 -1.38 Percent student computers with Content-specific simulation software 446 2.79 1.35 1 5 0.41 -1.07 Percent student computers with Other content-specific resources software 446 2.96 1.39 1 5 0.27 -1.28

PAGE 367

Appendix C: Data Preparation Procedures (Continued) 353 Label N Mean Std Dev Min Max Skew Kurt Percent student computers with General Reference tools software 446 4.00 1.31 1 5 -0.90 -0.67 Administrative tasks 446 4.72 0.69 2 5 -2.63 6.34 Delivery of lessons 446 3.05 0.96 1 5 0.42 -0.68 Desktop video production 446 1.95 0.65 1 5 1.21 4.14 Email to other school or district staff 446 4.73 0.78 1 5 -3.03 8.56 Email to students or parents 446 3.34 1.21 1 5 -0.02 -1.21 Presentations 446 3.02 0.96 1 5 0.50 -0.78 Research 446 3.98 0.94 2 5 -0.54 -0.68 Analysis of student assessment information 446 3.84 1.18 1 5 -0.56 -0.95 Video conferencing 446 1.24 0.51 1 5 2.61 10.19 Webpage publishing 446 2.13 0.90 1 5 1.58 3.05 Degree students use Drill and practice software 446 2.33 0.95 1 5 0.31 -0.46 Degree students use Integrated Learning Systems 446 2.07 1.05 1 5 0.81 0.04 Degree students use Multimedia 446 2.88 0.93 1 5 -0.18 -0.44 Degree students use Simulation software 446 3.36 0.87 1 5 -0.27 -0.10 Degree students use Tool-based software 446 1.95 0.89 1 5 0.62 -0.34 % of technology $ devoted to professional development 446 14.06 14.09 0 100 1.77 5.27 Level of school-based technical support 409 2.33 1.10 0 4 0.30 -1.19 Level of school-based instructional technology specialist support 364 2.21 1.16 1 4 0.37 -1.34 Level of dependability of the Internet connection 445 4.18 0.73 1 5 -0.60 0.28 Degree of delays when using the Internet 445 3.84 0.61 1 5 -2.24 7.26 Time at your school for a technical issue to be resolved 446 3.80 1.13 1 5 -0.41 -1.19 2004-05 All School Levels Modern multi-media computers in Media center (desktops) 2327 13.76 14.19 0 130 2.55 10.38 Modern multi-media computers in Classrooms (desktops) 2327 113.67 108.88 0 2106 4.45 55.52 Modern multi-media computers in Computer labs primarily serving general education (desktops) 2327 42.08 48.68 0 525 3.63 20.95 Modern multi-media computers in Mobile computer labs (desktops) 2327 1.62 8.61 0 106 7.06 58.58

PAGE 368

Appendix C: Data Preparation Procedures (Continued) 354 Label N Mean Std Dev Min Max Skew Kurt Older computer or not multi-media in Media center (desktops) 2327 3.64 7.29 0 67 3.66 17.71 Older computer or not multi-media in Classrooms (desktops) 2327 50.46 79.26 0 2106 9.07 198.66 Older computer or not multi-media in Computer labs primarily serving general education (desktops) 2327 8.22 20.46 0 220 4.30 26.44 Older computer or not multi-media in Mobile computer labs (desktops) 2327 0.94 6.61 0 120 10.59 141.36 Modern multi-media computers in Media center (laptop) 2326 2.72 60.18 0 2890 47.56 2282.47 Modern multi-media computers in Classrooms (laptops) 2327 11.64 92.87 0 2890 22.69 602.43 Modern multi-media computers in Computer labs primarily serving general education (laptops) 2326 4.21 61.62 0 2890 44.34 2072.27 Modern multi-media computers in Mobile computer labs (laptops) 2327 10.88 26.54 0 399 5.35 49.00 Older computer or not multi-media in Media center (laptops) 2327 0.38 3.03 0 70 13.90 240.43 Older computer or not multi-media in Classrooms (laptops) 2327 1.85 13.32 0 351 16.31 347.76 Older computer or not multi-media in Computer labs primarily serving general education (laptops) 2327 0.71 8.23 0 220 18.64 405.80 Older computer or not multi-media in Mobile computer labs (laptops) 2327 0.18 2.56 0 75 20.58 499.50 Percent student computers with Concept mapping software 2327 2.27 1.38 1 5 0.93 -0.44 Percent student computers with Graphics software 2327 4.37 1.17 1 5 -1.67 1.32 Percent student computers with Multimedia authoring software 2327 2.38 1.37 1 5 0.82 -0.60 Percent student computers with Presentation software 2327 4.33 1.10 1 5 -1.47 0.89 Percent student computers with Spreadsheet software 2327 4.57 0.92 1 5 -2.25 4.28 Percent student computers with Video editing software 2327 2.14 1.14 1 5 1.20 0.75 Percent student computers with Web authoring software 2327 2.03 1.14 1 5 1.42 1.40 Percent student computers with Basic word processing software 2327 4.84 0.67 1 5 -4.65 21.29 Percent student computers with Robust word processing software 2327 4.65 0.89 1 5 -2.73 6.71 Percent student computers with FCAT Explorer software 2327 4.32 1.15 1 5 -1.64 1.57

PAGE 369

Appendix C: Data Preparation Procedures (Continued) 355 Label N Mean Std Dev Min Max Skew Kurt Percent student computers with Other test prep tools software 2327 3.45 1.52 1 5 -0.39 -1.37 Percent student computers with Integrated Learning Systems software 2327 3.32 1.54 1 5 -0.24 -1.47 Percent student computers with Content-specific skills practice/tutorials software 2327 3.87 1.27 1 5 -0.71 -0.85 Percent student computers with Content-specific simulation software 2327 2.94 1.49 1 5 0.16 -1.42 Percent student computers with Other content-specific resources software 2327 3.24 1.47 1 5 -0.12 -1.43 Percent student computers with General Reference tools software 2327 4.07 1.31 1 5 -1.10 -0.22 Administrative tasks 2327 4.48 0.92 1 5 -1.73 1.96 Delivery of lessons 2327 3.09 0.99 1 5 0.30 -0.77 Desktop video production 2327 1.87 0.73 1 5 1.26 3.47 Email to other school or district staff 2327 4.74 0.69 1 5 -2.87 8.01 Email to students or parents 2327 3.27 1.25 1 5 0.03 -1.24 Presentations 2327 2.99 0.98 1 5 0.44 -0.66 Research 2327 3.95 0.98 1 5 -0.61 -0.56 Analysis of student assessment information 2327 4.11 1.07 1 5 -0.98 -0.13 Video conferencing 2327 1.27 0.58 1 5 2.97 12.14 Webpage publishing 2327 2.09 0.86 1 5 1.68 3.77 Degree students use Drill and practice software 2327 3.63 1.01 1 5 -0.63 -0.12 Degree students use Integrated Learning Systems 2327 3.49 1.20 1 5 -0.61 -0.51 Degree students use Multimedia 2327 2.36 1.02 1 5 0.73 0.01 Degree students use Simulation software 2327 2.02 0.97 1 5 0.83 0.09 Degree students use Tool-based software 2327 3.13 1.08 1 5 0.03 -0.89 % of technology $ devoted to professional development 2327 11.89 13.10 0 100 1.95 6.86 Level of school-based technical support 2202 2.35 1.09 1 4 0.03 -1.34 Level of school-based instructional technology specialist support 1922 2.06 1.11 1 4 0.49 -1.23 Level of dependability of the Internet connection 2327 4.41 0.74 1 5 -1.12 1.01 Degree of delays when using the Internet 2327 3.85 0.66 1 5 -2.05 6.05 Time at your school for a technical issue to be resolved 2327 3.64 1.18 1 5 -0.17 -1.46

PAGE 370

Appendix C: Data Preparation Procedures (Continued) 356 Label N Mean Std Dev Min Max Skew Kurt Elementary Modern multi-media computers in Media center (desktops) 1531 8.74 7.06 0 46 1.55 3.41 Modern multi-media computers in Classrooms (desktops) 1531 102.05 72.59 0 476 1.06 1.65 Modern multi-media computers in Computer labs primarily serving general education (desktops) 1531 27.98 22.03 0 143 0.84 1.46 Modern multi-media computers in Mobile computer labs (desktops) 1531 0.93 5.32 0 75 7.14 60.07 Older computer or not multi-media in Media center (desktops) 1531 2.78 4.62 0 39 2.56 8.96 Older computer or not multi-media in Classrooms (desktops) 1531 49.07 54.02 0 350 1.76 4.31 Older computer or not multi-media in Computer labs primarily serving general education (desktops) 1531 5.70 13.47 0 143 3.12 13.69 Older computer or not multi-media in Mobile computer labs (desktops) 1531 0.82 5.38 0 100 9.77 129.64 Modern multi-media computers in Media center (laptop) 1531 0.95 3.87 0 42 5.77 39.36 Modern multi-media computers in Classrooms (laptops) 1531 7.13 28.54 0 340 6.49 50.10 Modern multi-media computers in Computer labs primarily serving general education (laptops) 1531 2.09 9.13 0 143 6.56 61.01 Modern multi-media computers in Mobile computer labs (laptops) 1531 7.04 17.10 0 184 4.02 23.59 Older computer or not multi-media in Media center (laptops) 1531 0.25 1.82 0 40 13.23 226.59 Older computer or not multi-media in Classrooms (laptops) 1531 1.72 10.34 0 199 10.99 154.68 Older computer or not multi-media in Computer labs primarily serving general education (laptops) 1531 0.43 4.99 0 143 19.52 481.34 Older computer or not multi-media in Mobile computer labs (laptops) 1531 0.11 1.44 0 29 16.50 290.39 Percent student computers with Concept mapping software 1531 2.33 1.43 1 5 0.83 -0.71 Percent student computers with Graphics software 1531 4.35 1.20 1 5 -1.64 1.20 Percent student computers with Multimedia authoring software 1531 2.37 1.42 1 5 0.80 -0.70 Percent student computers with Presentation software 1531 4.14 1.20 1 5 -1.10 -0.17 Percent student computers with Spreadsheet software 1531 4.45 1.02 1 5 -1.88 2.50

PAGE 371

Appendix C: Data Preparation Procedures (Continued) 357 Label N Mean Std Dev Min Max Skew Kurt Percent student computers with Video editing software 1531 2.07 1.19 1 5 1.18 0.51 Percent student computers with Web authoring software 1531 1.85 1.09 1 5 1.71 2.47 Percent student computers with Basic word processing software 1531 4.81 0.73 1 5 -4.18 16.90 Percent student computers with Robust word processing software 1531 4.56 0.97 1 5 -2.36 4.65 Percent student computers with FCAT Explorer software 1531 4.31 1.16 1 5 -1.69 1.74 Percent student computers with Other test prep tools software 1531 3.47 1.60 1 5 -0.47 -1.40 Percent student computers with Integrated Learning Systems software 1531 3.47 1.60 1 5 -0.45 -1.42 Percent student computers with Content-specific skills practice/tutorials software 1531 4.14 1.19 1 5 -1.20 0.25 Percent student computers with Content-specific simulation software 1531 2.98 1.58 1 5 0.07 -1.56 Percent student computers with Other content-specific resources software 1531 3.39 1.52 1 5 -0.34 -1.39 Percent student computers with General Reference tools software 1531 4.03 1.33 1 5 -1.07 -0.27 Administrative tasks 1531 4.33 1.02 1 5 -1.36 0.66 Delivery of lessons 1531 2.99 1.00 1 5 0.41 -0.75 Desktop video production 1531 1.76 0.72 1 5 1.29 3.54 Email to other school or district staff 1531 4.71 0.72 1 5 -2.72 7.07 Email to students or parents 1531 3.07 1.22 1 5 0.23 -1.11 Presentations 1531 2.83 0.96 1 5 0.64 -0.36 Research 1531 3.86 1.00 1 5 -0.49 -0.71 Analysis of student assessment information 1531 4.12 1.08 1 5 -1.01 -0.06 Video conferencing 1531 1.23 0.58 1 5 3.54 16.24 Webpage publishing 1531 2.00 0.83 1 5 1.70 4.26 Degree students use Drill and practice software 1531 3.85 0.89 1 5 -0.83 0.86 Degree students use Integrated Learning Systems 1531 3.59 1.22 1 5 -0.79 -0.26 Degree students use Multimedia 1531 2.14 0.88 1 5 0.74 0.42 Degree students use Simulation software 1531 2.00 0.97 1 5 0.77 -0.16 Degree students use Tool-based software 1531 2.85 1.02 1 5 0.22 -0.70 % of technology $ devoted to professional development 1531 11.46 13.18 0 100 1.96 6.67

PAGE 372

Appendix C: Data Preparation Procedures (Continued) 358 Label N Mean Std Dev Min Max Skew Kurt Level of school-based technical support 1439 2.24 1.06 1 4 0.16 -1.30 Level of school-based instructional technology specialist support 1237 1.97 1.07 1 4 0.61 -1.04 Level of dependability of the Internet connection 1531 4.38 0.74 1 5 -1.05 0.85 Degree of delays when using the Internet 1531 3.84 0.65 1 5 -2.00 5.70 Time at your school for a technical issue to be resolved 1531 3.54 1.19 1 5 -0.04 -1.50 High Modern multi-media computers in Media center (desktops) 350 30.75 21.65 0 130 1.35 3.11 Modern multi-media computers in Classrooms (desktops) 350 164.69 197.92 0 2106 3.84 28.10 Modern multi-media computers in Computer labs primarily serving general education (desktops) 350 83.89 86.23 0 525 2.15 5.66 Modern multi-media computers in Mobile computer labs (desktops) 350 3.33 13.79 0 102 4.89 25.26 Older computer or not multi-media in Media center (desktops) 350 5.09 11.04 0 63 2.85 8.35 Older computer or not multi-media in Classrooms (desktops) 350 56.37 146.33 0 2106 8.71 111.36 Older computer or not multi-media in Computer labs primarily serving general education (desktops) 350 14.54 33.06 0 220 3.46 14.03 Older computer or not multi-media in Mobile computer labs (desktops) 350 1.24 8.91 0 115 9.05 94.36 Modern multi-media computers in Media center (laptop) 349 11.50 154.83 0 2890 18.57 346.26 Modern multi-media computers in Classrooms (laptops) 350 31.42 223.23 0 2890 10.44 115.21 Modern multi-media computers in Computer labs primarily serving general education (laptops) 349 13.72 156.67 0 2890 17.91 329.13 Modern multi-media computers in Mobile computer labs (laptops) 350 20.18 44.65 0 399 4.65 29.81 Older computer or not multi-media in Media center (laptops) 350 1.03 6.02 0 70 8.49 81.68 Older computer or not multi-media in Classrooms (laptops) 350 1.59 16.51 0 300 17.13 308.62 Older computer or not multi-media in Computer labs primarily serving general education (laptops) 350 1.70 15.06 0 220 11.91 154.48 Older computer or not multi-media in Mobile computer labs (laptops) 350 0.55 5.57 0 75 11.37 135.98

PAGE 373

Appendix C: Data Preparation Procedures (Continued) 359 Label N Mean Std Dev Min Max Skew Kurt Percent student computers with Concept mapping software 350 2.05 1.18 1 5 1.33 1.02 Percent student computers with Graphics software 350 4.45 1.05 1 5 -1.69 1.39 Percent student computers with Multimedia authoring software 350 2.39 1.19 1 5 0.95 0.06 Percent student computers with Presentation software 350 4.77 0.65 1 5 -3.24 10.76 Percent student computers with Spreadsheet software 350 4.83 0.56 1 5 -4.01 17.18 Percent student computers with Video editing software 350 2.30 0.92 1 5 1.68 2.70 Percent student computers with Web authoring software 350 2.59 1.10 1 5 1.19 0.38 Percent student computers with Basic word processing software 350 4.91 0.50 1 5 -6.28 40.76 Percent student computers with Robust word processing software 350 4.82 0.67 1 5 -4.21 17.69 Percent student computers with FCAT Explorer software 350 4.31 1.08 1 5 -1.34 0.42 Percent student computers with Other test prep tools software 350 3.32 1.30 1 5 0.06 -1.41 Percent student computers with Integrated Learning Systems software 350 2.82 1.26 1 5 0.45 -0.92 Percent student computers with Content-specific skills practice/tutorials software 350 3.15 1.21 1 5 0.32 -1.21 Percent student computers with Content-specific simulation software 350 2.78 1.19 1 5 0.55 -0.75 Percent student computers with Other content-specific resources software 350 2.88 1.25 1 5 0.49 -0.96 Percent student computers with General Reference tools software 350 4.13 1.25 1 5 -1.11 -0.22 Administrative tasks 350 4.79 0.58 2 5 -3.13 9.98 Delivery of lessons 350 3.26 0.90 1 5 0.13 -0.50 Desktop video production 350 2.13 0.65 1 5 1.85 5.93 Email to other school or district staff 350 4.75 0.67 2 5 -2.88 7.83 Email to students or parents 350 3.69 1.15 1 5 -0.34 -1.20 Presentations 350 3.41 0.93 2 5 0.10 -0.86 Research 350 4.19 0.88 2 5 -0.87 -0.04 Analysis of student assessment information 350 3.99 1.05 1 5 -0.73 -0.53 Video conferencing 350 1.45 0.61 1 4 1.33 2.20 Webpage publishing 350 2.28 0.80 1 5 2.03 4.58

PAGE 374

Appendix C: Data Preparation Procedures (Continued) 360 Label N Mean Std Dev Min Max Skew Kurt Degree students use Drill and practice software 350 3.29 1.12 1 5 -0.14 -0.97 Degree students use Integrated Learning Systems 350 3.32 1.17 1 5 -0.28 -0.87 Degree students use Multimedia 350 3.08 1.13 1 5 0.18 -1.05 Degree students use Simulation software 350 2.00 1.03 1 5 1.07 0.74 Degree students use Tool-based software 350 3.97 0.93 1 5 -0.67 -0.25 % of technology $ devoted to professional development 350 12.70 13.72 0 100 2.48 10.74 Level of school-based technical support 340 2.70 1.11 1 4 -0.40 -1.18 Level of school-based instructional technology specialist support 302 2.33 1.23 1 4 0.15 -1.59 Level of dependability of the Internet connection 350 4.44 0.74 2 5 -1.18 0.74 Degree of delays when using the Internet 350 3.87 0.64 1 5 -1.94 6.39 Time at your school for a technical issue to be resolved 350 3.83 1.11 2 5 -0.37 -1.25 Middle Modern multi-media computers in Media center (desktops) 446 17.63 13.61 0 95 1.51 4.23 Modern multi-media computers in Classrooms (desktops) 446 113.51 103.00 0 548 1.14 1.01 Modern multi-media computers in Computer labs primarily serving general education (desktops) 446 57.67 50.84 0 364 1.94 6.81 Modern multi-media computers in Mobile computer labs (desktops) 446 2.62 11.65 0 106 5.60 35.66 Older computer or not multi-media in Media center (desktops) 446 5.44 10.08 0 67 2.71 8.61 Older computer or not multi-media in Classrooms (desktops) 446 50.61 77.29 0 478 2.38 6.77 Older computer or not multi-media in Computer labs primarily serving general education (desktops) 446 11.89 25.32 0 180 3.09 12.06 Older computer or not multi-media in Mobile computer labs (desktops) 446 1.12 8.17 0 120 10.55 128.55 Modern multi-media computers in Media center (laptop) 446 1.90 7.05 0 60 4.66 24.09 Modern multi-media computers in Classrooms (laptops) 446 11.61 53.14 0 841 10.19 139.50 Modern multi-media computers in Computer labs primarily serving general education (laptops) 446 4.07 16.35 0 161 5.08 30.80

PAGE 375

Appendix C: Data Preparation Procedures (Continued) 361 Label N Mean Std Dev Min Max Skew Kurt Modern multi-media computers in Mobile computer labs (laptops) 446 16.77 30.98 0 253 2.73 10.43 Older computer or not multi-media in Media center (laptops) 446 0.32 2.79 0 40 11.67 148.49 Older computer or not multi-media in Classrooms (laptops) 446 2.51 18.62 0 351 15.48 279.50 Older computer or not multi-media in Computer labs primarily serving general education (laptops) 446 0.89 9.43 0 180 16.30 297.91 Older computer or not multi-media in Mobile computer labs (laptops) 446 0.14 1.60 0 30 16.34 289.28 Percent student computers with Concept mapping software 446 2.23 1.31 1 5 0.98 -0.20 Percent student computers with Graphics software 446 4.40 1.15 1 5 -1.74 1.54 Percent student computers with Multimedia authoring software 446 2.43 1.37 1 5 0.81 -0.60 Percent student computers with Presentation software 446 4.62 0.80 1 5 -2.31 4.86 Percent student computers with Spreadsheet software 446 4.76 0.65 1 5 -3.06 9.93 Percent student computers with Video editing software 446 2.26 1.11 1 5 1.29 1.04 Percent student computers with Web authoring software 446 2.22 1.19 1 5 1.22 0.65 Percent student computers with Basic word processing software 446 4.90 0.56 1 5 -6.04 36.67 Percent student computers with Robust word processing software 446 4.80 0.66 1 5 -3.71 14.04 Percent student computers with FCAT Explorer software 446 4.35 1.14 1 5 -1.68 1.65 Percent student computers with Other test prep tools software 446 3.49 1.39 1 5 -0.33 -1.26 Percent student computers with Integrated Learning Systems software 446 3.21 1.43 1 5 -0.10 -1.34 Percent student computers with Content-specific skills practice/tutorials software 446 3.53 1.29 1 5 -0.26 -1.28 Percent student computers with Content-specific simulation software 446 2.92 1.36 1 5 0.24 -1.21 Percent student computers with Other content-specific resources software 446 3.03 1.38 1 5 0.16 -1.29 Percent student computers with General Reference tools software 446 4.15 1.25 1 5 -1.15 -0.15 Administrative tasks 446 4.76 0.60 2 5 -2.74 7.16

PAGE 376

Appendix C: Data Preparation Procedures (Continued) 362 Label N Mean Std Dev Min Max Skew Kurt Delivery of lessons 446 3.27 0.97 1 5 0.11 -0.73 Desktop video production 446 2.00 0.76 1 5 1.28 3.25 Email to other school or district staff 446 4.81 0.58 1 5 -3.56 13.48 Email to students or parents 446 3.63 1.25 1 5 -0.36 -1.14 Presentations 446 3.23 0.94 1 5 0.24 -0.72 Research 446 4.08 0.97 1 5 -0.83 -0.14 Analysis of student assessment information 446 4.20 1.06 1 5 -1.11 0.09 Video conferencing 446 1.25 0.52 1 5 2.69 10.37 Webpage publishing 446 2.24 0.95 1 5 1.56 2.44 Degree students use Drill and practice software 446 3.16 1.08 1 5 -0.17 -0.82 Degree students use Integrated Learning Systems 446 3.30 1.10 1 5 -0.30 -0.66 Degree students use Multimedia 446 2.55 1.07 1 5 0.60 -0.39 Degree students use Simulation software 446 2.07 0.96 1 5 0.83 0.32 Degree students use Tool-based software 446 3.43 0.98 1 5 -0.19 -0.79 % of technology $ devoted to professional development 446 12.71 12.25 0 80 1.39 3.18 Level of school-based technical support 423 2.45 1.08 1 4 -0.11 -1.30 Level of school-based instructional technology specialist support 383 2.15 1.12 1 4 0.34 -1.33 Level of dependability of the Internet connection 446 4.46 0.73 1 5 -1.36 2.01 Degree of delays when using the Internet 446 3.86 0.68 1 5 -2.28 7.01 Time at your school for a technical issue to be resolved 446 3.84 1.16 1 5 -0.46 -1.26 2005-06 All School Levels Modern multi-media computers in Media center (desktops) 2327 10.65 13.69 0 120 2.58 9.44 Modern multi-media computers in Classrooms (desktops) 2327 79.67 106.77 0 2000 5.30 62.15 Modern multi-media computers in Computer labs primarily serving general education (desktops) 2327 31.31 41.31 0 538 3.87 25.36 Modern multi-media computers in Mobile computer labs (laptops) 2327 14.13 47.79 0 859 6.69 67.33 Older computer or not multi-media in Media center (desktops) 2327 6.48 9.90 0 89 2.96 12.55 Older computer or not multi-media in Classrooms (desktops) 2327 73.50 74.03 0 721 1.72 5.57

PAGE 377

Appendix C: Data Preparation Procedures (Continued) 363 Label N Mean Std Dev Min Max Skew Kurt Older computer or not multi-media in Computer labs primarily serving general education (desktops) 2327 15.82 28.65 0 370 4.09 30.71 Modern multi-media computers in Media center (laptops) 2327 1.94 10.13 0 214 10.77 162.35 Modern multi-media computers in Classrooms (laptops) 2327 12.29 83.00 0 2520 20.80 560.10 Modern multi-media computers in Computer labs primarily serving general education (laptops) 2327 3.47 23.93 0 496 13.28 220.38 Older computer or not multi-media in Media center (laptops) 2327 0.65 4.00 0 84 10.41 148.28 Older computer or not multi-media in Classrooms (laptops) 2326 3.06 18.09 0 314 10.63 134.49 Older computer or not multi-media in Computer labs primarily serving general education (laptops) 2327 0.84 8.56 0 238 17.95 400.49 Older computer or not multi-media in Mobile computer labs (laptops) 2327 3.43 16.35 0 361 10.41 156.58 Percent student computers with Concept mapping software 2327 2.54 1.52 1 5 0.60 -1.15 Percent student computers with Graphics software 2327 4.45 1.12 1 5 -1.94 2.40 Percent student computers with Multimedia authoring software 2327 2.71 1.45 1 5 0.46 -1.20 Percent student computers with Presentation software 2327 4.47 0.96 1 5 -1.86 2.58 Percent student computers with Spreadsheet software 2327 4.61 0.87 1 5 -2.48 5.61 Percent student computers with Video editing software 2327 2.29 1.26 1 5 0.99 -0.07 Percent student computers with Web authoring software 2327 1.99 1.15 1 5 1.45 1.43 Percent student computers with Basic word processing software 2327 4.87 0.60 1 5 -5.08 26.12 Percent student computers with Robust word processing software 2327 4.63 0.95 1 5 -2.72 6.43 Percent student computers with FCAT Explorer software 2327 4.75 0.71 1 5 -3.38 11.72 Percent student computers with Other test prep tools software 2327 3.70 1.51 1 5 -0.70 -1.07 Percent student computers with Integrated Learning Systems software 2327 3.20 1.54 1 5 -0.07 -1.55 Percent student computers with Content-specific skills practice/tutorials software 2327 3.60 1.38 1 5 -0.38 -1.32

PAGE 378

Appendix C: Data Preparation Procedures (Continued) 364 Label N Mean Std Dev Min Max Skew Kurt Percent student computers with Content-specific simulation software 2327 2.06 1.33 1 5 1.19 0.15 Percent student computers with Other content-specific resources software 2327 3.06 1.56 1 5 0.07 -1.55 Percent student computers with General Reference tools software 2327 4.11 1.33 1 5 -1.16 -0.18 Administrative tasks 2326 4.67 0.74 1 5 -2.54 6.11 Delivery of lessons 2326 3.25 1.03 1 5 0.17 -0.98 Desktop video production 2326 1.91 0.79 1 5 1.10 2.01 Email to other school or district staff 2326 4.79 0.63 1 5 -3.43 12.04 Email to students or parents 2326 3.35 1.22 1 5 -0.08 -1.16 Presentations 2326 3.05 1.01 1 5 0.37 -0.83 Research 2326 4.03 0.96 1 5 -0.69 -0.47 Analysis of student assessment information 2326 4.28 0.98 1 5 -1.23 0.55 Video conferencing 2326 1.30 0.63 1 5 2.85 10.65 Webpage publishing 2326 2.09 0.91 1 5 1.41 2.43 Degree students use Drill and practice software 2326 3.72 1.09 1 5 -0.76 -0.04 Degree students use Integrated Learning Systems 2326 3.71 1.26 1 5 -0.82 -0.35 Degree students use Multimedia 2326 2.70 1.25 1 5 0.37 -0.93 Degree students use Presentation 2326 2.62 1.24 1 5 0.49 -0.85 Degree students use Simulation software 2326 1.98 1.11 1 5 0.98 0.04 Degree students use Research software 2326 3.70 1.09 1 5 -0.41 -0.84 Degree students use Tool-based software 2326 3.43 1.21 1 5 -0.25 -1.04 % of technology $ devoted to professional development 2326 11.74 13.45 0 100 2.10 7.51 Level of school-based technical support 2326 2.36 1.15 1 6 0.26 -0.72 Level of school-based instructional technology specialist support 2326 2.29 1.36 1 5 0.58 -0.97 Level of dependability of the Internet connection 2327 4.50 0.74 1 5 -1.51 2.08 Degree of delays when using the Internet 2327 3.64 1.01 1 5 -1.37 1.21 Time at your school for a technical issue to be resolved 2327 3.47 1.14 1 5 0.04 -1.35 Elementary Modern multi-media computers in Media center (desktops) 1531 6.71 7.56 0 90 2.40 12.36 Modern multi-media computers in Classrooms (desktops) 1531 70.62 68.73 0 436 1.46 2.67

PAGE 379

Appendix C: Data Preparation Procedures (Continued) 365 Label N Mean Std Dev Min Max Skew Kurt Modern multi-media computers in Computer labs primarily serving general education (desktops) 1531 21.40 20.93 0 199 1.51 6.60 Modern multi-media computers in Mobile computer labs (laptops) 1531 9.01 30.73 0 240 4.92 26.08 Older computer or not multi-media in Media center (desktops) 1531 4.76 6.44 0 89 3.70 30.35 Older computer or not multi-media in Classrooms (desktops) 1531 75.80 63.58 0 393 0.91 0.76 Older computer or not multi-media in Computer labs primarily serving general education (desktops) 1531 10.77 17.52 0 143 1.95 5.33 Modern multi-media computers in Media center (laptops) 1531 1.27 6.79 0 168 12.72 257.24 Modern multi-media computers in Classrooms (laptops) 1531 8.33 32.90 0 360 5.39 33.84 Modern multi-media computers in Computer labs primarily serving general education (laptops) 1531 1.86 10.27 0 192 10.27 141.04 Older computer or not multi-media in Media center (laptops) 1531 0.45 2.66 0 30 7.86 69.21 Older computer or not multi-media in Classrooms (laptops) 1531 3.20 17.59 0 237 9.58 105.15 Older computer or not multi-media in Computer labs primarily serving general education (laptops) 1531 0.63 6.42 0 190 20.81 546.89 Older computer or not multi-media in Mobile computer labs (laptops) 1531 2.07 9.55 0 128 8.04 82.27 Percent student computers with Concept mapping software 1531 2.61 1.57 1 5 0.50 -1.32 Percent student computers with Graphics software 1531 4.42 1.17 1 5 -1.89 2.17 Percent student computers with Multimedia authoring software 1531 2.72 1.50 1 5 0.39 -1.32 Percent student computers with Presentation software 1531 4.31 1.07 1 5 -1.46 1.07 Percent student computers with Spreadsheet software 1531 4.49 0.99 1 5 -2.03 3.22 Percent student computers with Video editing software 1531 2.21 1.31 1 5 0.99 -0.18 Percent student computers with Web authoring software 1531 1.81 1.08 1 5 1.76 2.67 Percent student computers with Basic word processing software 1531 4.82 0.71 1 5 -4.27 18.01 Percent student computers with Robust word processing software 1531 4.52 1.06 1 5 -2.27 3.99

PAGE 380

Appendix C: Data Preparation Procedures (Continued) 366 Label N Mean Std Dev Min Max Skew Kurt Percent student computers with FCAT Explorer software 1531 4.73 0.73 1 5 -3.19 10.65 Percent student computers with Other test prep tools software 1531 3.67 1.58 1 5 -0.71 -1.14 Percent student computers with Integrated Learning Systems software 1531 3.36 1.65 1 5 -0.33 -1.57 Percent student computers with Content-specific skills practice/tutorials software 1531 3.94 1.27 1 5 -0.86 -0.55 Percent student computers with Content-specific simulation software 1531 1.97 1.38 1 5 1.26 0.18 Percent student computers with Other content-specific resources software 1531 3.19 1.62 1 5 -0.13 -1.61 Percent student computers with General Reference tools software 1531 4.10 1.34 1 5 -1.16 -0.17 Administrative tasks 1530 4.56 0.84 1 5 -2.06 3.57 Delivery of lessons 1530 3.13 1.05 1 5 0.31 -0.94 Desktop video production 1530 1.79 0.76 1 5 1.13 2.08 Email to other school or district staff 1530 4.77 0.67 1 5 -3.27 10.78 Email to students or parents 1530 3.16 1.21 1 5 0.09 -1.09 Presentations 1530 2.86 0.99 1 5 0.59 -0.56 Research 1530 3.97 0.99 1 5 -0.64 -0.57 Analysis of student assessment information 1530 4.28 0.99 1 5 -1.25 0.57 Video conferencing 1530 1.27 0.63 1 5 3.17 12.67 Webpage publishing 1530 1.99 0.90 1 5 1.49 2.84 Degree students use Drill and practice software 1530 3.84 1.01 1 5 -0.90 0.58 Degree students use Integrated Learning Systems 1530 3.73 1.27 1 5 -0.92 -0.17 Degree students use Multimedia 1530 2.36 1.08 1 5 0.58 -0.40 Degree students use Presentation 1530 2.16 1.02 1 5 0.86 0.20 Degree students use Simulation software 1530 1.90 1.06 1 5 1.03 0.16 Degree students use Research software 1530 3.42 1.06 1 5 -0.17 -0.88 Degree students use Tool-based software 1530 3.07 1.13 1 5 0.01 -0.90 % of technology $ devoted to professional development 1530 11.16 13.13 0 100 2.03 6.95 Level of school-based technical support 1530 2.24 1.15 1 6 0.47 -0.44 Level of school-based instructional technology specialist support 1530 2.23 1.38 1 5 0.72 -0.80

PAGE 381

Appendix C: Data Preparation Procedures (Continued) 367 Label N Mean Std Dev Min Max Skew Kurt Level of dependability of the Internet connection 1531 4.46 0.77 1 5 -1.41 1.76 Degree of delays when using the Internet 1531 3.59 1.04 1 5 -1.31 0.92 Time at your school for a technical issue to be resolved 1531 3.37 1.15 1 5 0.15 -1.35 High Modern multi-media computers in Media center (desktops) 350 24.38 21.85 0 120 1.08 1.25 Modern multi-media computers in Classrooms (desktops) 350 116.77 201.32 0 2000 4.27 27.68 Modern multi-media computers in Computer labs primarily serving general education (desktops) 350 63.89 74.74 0 538 2.32 7.40 Modern multi-media computers in Mobile computer labs (laptops) 350 26.53 80.79 0 859 5.70 41.56 Older computer or not multi-media in Media center (desktops) 350 11.35 16.40 0 85 1.77 3.11 Older computer or not multi-media in Classrooms (desktops) 350 69.66 101.19 0 721 2.51 8.35 Older computer or not multi-media in Computer labs primarily serving general education (desktops) 350 30.62 49.91 0 370 3.25 14.82 Modern multi-media computers in Media center (laptops) 350 3.09 13.00 0 158 7.04 65.82 Modern multi-media computers in Classrooms (laptops) 350 29.21 189.35 0 2520 10.92 131.65 Modern multi-media computers in Computer labs primarily serving general education (laptops) 350 6.26 40.45 0 496 10.44 120.54 Older computer or not multi-media in Media center (laptops) 350 1.05 4.55 0 32 4.98 25.22 Older computer or not multi-media in Classrooms (laptops) 349 3.10 20.57 0 314 11.47 157.24 Older computer or not multi-media in Computer labs primarily serving general education (laptops) 350 0.98 8.75 0 128 12.08 158.22 Older computer or not multi-media in Mobile computer labs (laptops) 350 5.70 20.68 0 210 5.62 39.02 Percent student computers with Concept mapping software 350 2.31 1.37 1 5 1.00 -0.23 Percent student computers with Graphics software 350 4.47 1.05 1 5 -1.81 1.82 Percent student computers with Multimedia authoring software 350 2.70 1.23 1 5 0.77 -0.62 Percent student computers with Presentation software 350 4.82 0.57 2 5 -3.53 12.44

PAGE 382

Appendix C: Data Preparation Procedures (Continued) 368 Label N Mean Std Dev Min Max Skew Kurt Percent student computers with Spreadsheet software 350 4.88 0.47 2 5 -4.32 19.45 Percent student computers with Video editing software 350 2.53 1.07 1 5 1.22 0.53 Percent student computers with Web authoring software 350 2.61 1.15 1 5 1.02 0.00 Percent student computers with Basic word processing software 350 4.96 0.31 2 5 -8.55 75.59 Percent student computers with Robust word processing software 350 4.87 0.57 1 5 -5.16 27.49 Percent student computers with FCAT Explorer software 350 4.77 0.72 1 5 -3.40 10.92 Percent student computers with Other test prep tools software 350 3.69 1.31 1 5 -0.46 -1.20 Percent student computers with Integrated Learning Systems software 350 2.70 1.16 1 5 0.80 -0.41 Percent student computers with Content-specific skills practice/tutorials software 350 2.64 1.26 1 5 0.75 -0.58 Percent student computers with Content-specific simulation software 350 2.21 1.13 1 5 1.26 0.97 Percent student computers with Other content-specific resources software 350 2.60 1.28 1 5 0.78 -0.55 Percent student computers with General Reference tools software 350 4.18 1.27 1 5 -1.20 -0.11 Administrative tasks 350 4.89 0.40 2 5 -4.46 22.50 Delivery of lessons 350 3.45 0.93 2 5 0.11 -0.85 Desktop video production 350 2.16 0.67 1 5 1.46 3.99 Email to other school or district staff 350 4.83 0.54 2 5 -3.76 14.70 Email to students or parents 350 3.78 1.09 1 5 -0.41 -0.98 Presentations 350 3.44 0.94 2 5 0.08 -0.88 Research 350 4.13 0.89 2 5 -0.73 -0.35 Analysis of student assessment information 350 4.15 1.00 1 5 -0.91 -0.20 Video conferencing 350 1.38 0.59 1 5 1.70 4.46 Webpage publishing 350 2.31 0.81 1 5 1.63 2.96 Degree students use Drill and practice software 350 3.55 1.24 1 5 -0.57 -0.69 Degree students use Integrated Learning Systems 350 3.72 1.28 1 5 -0.67 -0.74 Degree students use Multimedia 350 3.81 1.20 1 5 -0.77 -0.45 Degree students use Presentation 350 3.77 1.11 1 5 -0.49 -0.99 Degree students use Simulation software 350 2.09 1.19 1 5 0.86 -0.37

PAGE 383

Appendix C: Data Preparation Procedures (Continued) 369 Label N Mean Std Dev Min Max Skew Kurt Degree students use Research software 350 4.49 0.77 1 5 -1.51 1.91 Degree students use Tool-based software 350 4.43 0.87 1 5 -1.56 1.74 % of technology $ devoted to professional development 350 13.08 14.09 0 100 2.26 9.02 Level of school-based technical support 350 2.69 1.15 1 6 -0.25 -0.89 Level of school-based instructional technology specialist support 350 2.45 1.33 1 5 0.26 -1.30 Level of dependability of the Internet connection 350 4.61 0.66 2 5 -1.77 2.84 Degree of delays when using the Internet 350 3.68 0.97 1 5 -1.46 1.68 Time at your school for a technical issue to be resolved 350 3.64 1.10 2 5 -0.15 -1.30 Middle Modern multi-media computers in Media center (desktops) 446 13.37 14.11 0 99 2.02 6.81 Modern multi-media computers in Classrooms (desktops) 446 81.60 100.97 0 548 1.64 2.54 Modern multi-media computers in Computer labs primarily serving general education (desktops) 446 39.78 42.29 0 364 2.24 9.96 Modern multi-media computers in Mobile computer labs (laptops) 446 21.98 57.44 0 400 4.02 17.77 Older computer or not multi-media in Media center (desktops) 446 8.53 11.19 0 65 1.63 2.70 Older computer or not multi-media in Classrooms (desktops) 446 68.61 81.63 0 421 1.67 2.73 Older computer or not multi-media in Computer labs primarily serving general education (desktops) 446 21.54 31.44 0 200 2.01 5.42 Modern multi-media computers in Media center (laptops) 446 3.30 15.50 0 214 8.50 92.70 Modern multi-media computers in Classrooms (laptops) 446 12.62 62.36 0 936 9.24 115.05 Modern multi-media computers in Computer labs primarily serving general education (laptops) 446 6.80 36.34 0 420 7.35 61.30 Older computer or not multi-media in Media center (laptops) 446 0.99 6.54 0 84 9.20 94.93 Older computer or not multi-media in Classrooms (laptops) 446 2.56 17.76 0 300 12.88 193.72 Older computer or not multi-media in Computer labs primarily serving general education (laptops) 446 1.46 13.44 0 238 13.78 223.63

PAGE 384

Appendix C: Data Preparation Procedures (Continued) 370 Label N Mean Std Dev Min Max Skew Kurt Older computer or not multi-media in Mobile computer labs (laptops) 446 6.29 27.02 0 361 8.54 90.07 Percent student computers with Concept mapping software 446 2.49 1.45 1 5 0.67 -0.94 Percent student computers with Graphics software 446 4.54 0.99 1 5 -2.15 3.49 Percent student computers with Multimedia authoring software 446 2.68 1.42 1 5 0.55 -1.07 Percent student computers with Presentation software 446 4.75 0.63 1 5 -3.06 10.51 Percent student computers with Spreadsheet software 446 4.83 0.49 1 5 -3.72 16.98 Percent student computers with Video editing software 446 2.38 1.20 1 5 1.10 0.24 Percent student computers with Web authoring software 446 2.14 1.20 1 5 1.29 0.81 Percent student computers with Basic word processing software 446 4.96 0.31 1 5 -9.70 102.44 Percent student computers with Robust word processing software 446 4.80 0.71 1 5 -4.10 17.00 Percent student computers with FCAT Explorer software 446 4.83 0.65 1 5 -4.26 18.63 Percent student computers with Other test prep tools software 446 3.81 1.42 1 5 -0.76 -0.90 Percent student computers with Integrated Learning Systems software 446 3.02 1.31 1 5 0.32 -1.19 Percent student computers with Content-specific skills practice/tutorials software 446 3.15 1.37 1 5 0.23 -1.40 Percent student computers with Content-specific simulation software 446 2.25 1.29 1 5 1.04 -0.03 Percent student computers with Other content-specific resources software 446 2.95 1.45 1 5 0.28 -1.38 Percent student computers with General Reference tools software 446 4.13 1.33 1 5 -1.15 -0.26 Administrative tasks 446 4.89 0.44 2 5 -4.61 23.47 Delivery of lessons 446 3.50 0.94 2 5 -0.12 -0.88 Desktop video production 446 2.14 0.87 1 5 1.09 1.43 Email to other school or district staff 446 4.84 0.52 2 5 -3.66 13.84 Email to students or parents 446 3.67 1.21 1 5 -0.41 -1.07 Presentations 446 3.40 0.95 1 5 0.06 -0.86 Research 446 4.15 0.89 2 5 -0.78 -0.29 Analysis of student assessment information 446 4.39 0.91 1 5 -1.46 1.43

PAGE 385

Appendix C: Data Preparation Procedures (Continued) 371 Label N Mean Std Dev Min Max Skew Kurt Video conferencing 446 1.34 0.66 1 5 2.64 8.91 Webpage publishing 446 2.28 0.95 1 5 1.32 1.66 Degree students use Drill and practice software 446 3.44 1.20 1 5 -0.38 -0.84 Degree students use Integrated Learning Systems 446 3.64 1.23 1 5 -0.63 -0.63 Degree students use Multimedia 446 3.00 1.27 1 5 0.11 -1.10 Degree students use Presentation 446 3.27 1.17 1 5 0.03 -1.23 Degree students use Simulation software 446 2.17 1.19 1 5 0.88 -0.16 Degree students use Research software 446 4.03 0.98 1 5 -0.74 -0.37 Degree students use Tool-based software 446 3.87 1.09 1 5 -0.66 -0.59 % of technology $ devoted to professional development 446 12.69 13.90 0 97 2.17 7.78 Level of school-based technical support 446 2.51 1.11 1 6 -0.04 -0.78 Level of school-based instructional technology specialist support 446 2.40 1.32 1 5 0.38 -1.10 Level of dependability of the Internet connection 446 4.56 0.70 1 5 -1.68 2.86 Degree of delays when using the Internet 446 3.77 0.94 1 5 -1.50 2.09 Time at your school for a technical issue to be resolved 446 3.66 1.09 1 5 -0.17 -1.22 2006-07 All School Levels Modern multi-media computers in Media center (desktops) 2327 11.16 15.25 0 186 3.34 20.68 Modern multi-media computers in Media center (laptops) 2327 1.73 12.49 0 446 24.94 797.45 Modern multi-media computers in Classrooms (desktops) 2327 81.74 93.73 0 909 2.22 8.80 Modern multi-media computers in Classrooms (laptops) 2327 12.31 83.86 0 2580 20.96 563.86 Modern multi-media computers in Computer labs primarily serving general education (desktops) 2327 28.69 42.81 0 685 5.40 53.08 Modern multi-media computers in Computer labs primarily serving general education (laptops) 2327 2.54 20.57 0 600 19.34 471.73 Modern multi-media computers in Mobile computer labs (desktops) 2327 23.07 64.28 0 1204 5.97 63.90 Modern multi-media computers in Mobile computer labs (laptops) 2327 5.80 41.91 0 925 11.30 166.64

PAGE 386

Appendix C: Data Preparation Procedures (Continued) 372 Label N Mean Std Dev Min Max Skew Kurt Older computer or not multi-media in Media center (desktops) 2327 5.23 9.60 0 99 3.45 16.92 Older computer or not multi-media in Media center (laptops) 2327 0.63 4.41 0 128 14.66 331.90 Older computer or not multi-media in Classrooms (desktops) 2327 64.48 79.13 0 722 2.14 8.34 Older computer or not multi-media in Classrooms (laptops) 2327 3.07 15.81 0 260 9.04 101.42 Older computer or not multi-media in Computer labs primarily serving general education (desktops) 2327 12.96 26.62 0 411 4.05 31.34 Older computer or not multi-media in Computer labs primarily serving general education (laptops) 2327 0.48 3.94 0 97 13.06 228.33 Older computer or not multi-media in Mobile computer labs (desktops) 2327 6.02 23.39 0 474 8.11 102.24 Older computer or not multi-media in Mobile computer labs (laptops) 2327 1.55 41.77 0 1975 45.52 2145.63 Percent student computers with Concept mapping software 2327 1.58 1.55 0 4 0.56 -1.25 Percent student computers with Graphics software 2327 3.50 1.10 0 4 -2.11 3.08 Percent student computers with Multimedia authoring software 2327 1.80 1.54 0 4 0.32 -1.44 Percent student computers with Presentation software 2327 3.64 0.82 0 4 -2.52 5.93 Percent student computers with Spreadsheet software 2327 3.71 0.77 0 4 -3.10 9.66 Percent student computers with Video editing software 2327 1.54 1.46 0 4 0.65 -1.02 Percent student computers with Web authoring software 2327 0.91 1.16 0 4 1.53 1.61 Percent student computers with Basic word processing software 2327 3.90 0.53 0 4 -5.92 36.12 Percent student computers with Robust word processing software 2327 3.67 0.92 0 4 -2.99 8.05 Percent student computers with FCAT Explorer software 2327 3.85 0.56 0 4 -4.58 22.83 Percent student computers with Other test prep tools software 2327 2.83 1.48 0 4 -0.85 -0.85 Percent student computers with Integrated Learning Systems software 2327 2.23 1.56 0 4 -0.10 -1.57 Percent student computers with Content-specific skills practice/tutorials software 2327 2.53 1.44 0 4 -0.34 -1.40

PAGE 387

Appendix C: Data Preparation Procedures (Continued) 373 Label N Mean Std Dev Min Max Skew Kurt Percent student computers with Content-specific simulation software 2327 0.96 1.29 0 4 1.36 0.67 Percent student computers with Other content-specific resources software 2327 2.02 1.58 0 4 0.11 -1.57 Percent student computers with General Reference tools software 2327 3.18 1.33 0 4 -1.32 0.21 Administrative tasks 2327 3.80 0.58 0 4 -3.29 11.35 Delivery of lessons 2327 2.50 1.04 0 4 -0.16 -0.88 Desktop video production 2327 1.04 0.86 0 4 1.19 2.02 Email to other school or district staff 2327 3.87 0.49 0 4 -4.48 22.32 Email to students or parents 2327 2.61 1.23 0 4 -0.34 -1.18 Presentations 2327 2.34 1.05 0 4 0.07 -1.01 Research 2327 3.18 0.91 0 4 -0.90 -0.01 Analysis of student assessment information 2327 3.39 0.89 0 4 -1.42 1.26 Video conferencing 2327 0.36 0.66 0 4 2.32 6.73 Webpage publishing 2327 1.20 0.99 0 4 1.31 1.64 Degree students use Drill and practice software 2327 2.76 1.15 0 4 -0.86 -0.01 Degree students use Integrated Learning Systems 2327 2.77 1.27 0 4 -0.89 -0.26 Degree students use Multimedia 2327 1.77 1.28 0 4 0.31 -1.04 Degree students use presentation 2327 1.88 1.25 0 4 0.29 -1.07 Degree students use Simulation software 2327 0.91 1.11 0 4 1.06 0.10 Degree students use Tool-based software 2327 2.52 1.22 0 4 -0.34 -1.03 Degree students use research software 2327 2.86 1.07 0 4 -0.60 -0.68 % of technology $ devoted to professional development 2327 12.77 16.09 0 100 2.47 8.28 Level of school-based technical support 2327 3.36 1.06 0 5 -0.21 -0.92 Level of school-based instructional technology specialist support 2327 1.85 1.23 0 4 0.31 -1.03 Level of dependability of the Internet connection 2327 3.63 0.65 0 4 -1.92 3.94 Degree of delays when using the Internet 2327 2.73 0.97 0 4 -1.45 1.77 Time at your school for a technical issue to be resolved 2327 2.67 1.13 1 4 -0.18 -1.38 Elementary Modern multi-media computers in Media center (desktops) 1531 6.98 8.16 0 111 2.84 20.15

PAGE 388

Appendix C: Data Preparation Procedures (Continued) 374 Label N Mean Std Dev Min Max Skew Kurt Modern multi-media computers in Media center (laptops) 1531 1.03 4.33 0 40 5.41 31.46 Modern multi-media computers in Classrooms (desktops) 1531 80.14 77.11 0 561 1.30 2.06 Modern multi-media computers in Classrooms (laptops) 1531 8.51 32.84 0 434 6.05 47.88 Modern multi-media computers in Computer labs primarily serving general education (desktops) 1531 19.48 19.54 0 160 0.95 1.69 Modern multi-media computers in Computer labs primarily serving general education (laptops) 1531 1.73 9.51 0 242 13.69 290.79 Modern multi-media computers in Mobile computer labs (desktops) 1531 14.44 41.36 0 395 4.10 20.05 Modern multi-media computers in Mobile computer labs (laptops) 1531 3.08 20.44 0 262 7.96 70.04 Older computer or not multi-media in Media center (desktops) 1531 3.61 5.50 0 47 2.40 8.35 Older computer or not multi-media in Media center (laptops) 1531 0.40 2.55 0 34 8.78 86.42 Older computer or not multi-media in Classrooms (desktops) 1531 64.85 67.22 0 469 1.04 0.93 Older computer or not multi-media in Classrooms (laptops) 1531 3.24 15.25 0 190 7.75 71.50 Older computer or not multi-media in Computer labs primarily serving general education (desktops) 1531 9.32 18.64 0 206 3.28 18.47 Older computer or not multi-media in Computer labs primarily serving general education (laptops) 1531 0.51 4.09 0 97 13.36 243.51 Older computer or not multi-media in Mobile computer labs (desktops) 1531 4.60 18.29 0 184 6.13 44.00 Older computer or not multi-media in Mobile computer labs (laptops) 1531 0.66 7.56 0 122 13.61 194.40 Percent student computers with Concept mapping software 1531 1.64 1.60 0 4 0.46 -1.41 Percent student computers with Graphics software 1531 3.46 1.15 0 4 -2.03 2.75 Percent student computers with Multimedia authoring software 1531 1.76 1.60 0 4 0.32 -1.51 Percent student computers with Presentation software 1531 3.51 0.94 0 4 -1.99 3.22 Percent student computers with Spreadsheet software 1531 3.61 0.89 0 4 -2.55 6.08 Percent student computers with Video editing software 1531 1.41 1.50 0 4 0.72 -0.98

PAGE 389

Appendix C: Data Preparation Procedures (Continued) 375 Label N Mean Std Dev Min Max Skew Kurt Percent student computers with Web authoring software 1531 0.71 1.09 0 4 1.91 3.04 Percent student computers with Basic word processing software 1531 3.86 0.62 0 4 -5.14 26.69 Percent student computers with Robust word processing software 1531 3.58 1.02 0 4 -2.55 5.45 Percent student computers with FCAT Explorer software 1531 3.85 0.55 0 4 -4.79 25.66 Percent student computers with Other test prep tools software 1531 2.85 1.53 0 4 -0.93 -0.77 Percent student computers with Integrated Learning Systems software 1531 2.43 1.66 0 4 -0.40 -1.55 Percent student computers with Content-specific skills practice/tutorials software 1531 2.89 1.35 0 4 -0.84 -0.70 Percent student computers with Content-specific simulation software 1531 0.85 1.32 0 4 1.50 0.86 Percent student computers with Other content-specific resources software 1531 2.15 1.65 0 4 -0.09 -1.65 Percent student computers with General Reference tools software 1531 3.18 1.34 0 4 -1.35 0.32 Administrative tasks 1531 3.73 0.66 0 4 -2.79 7.75 Delivery of lessons 1531 2.39 1.08 0 4 -0.04 -1.00 Desktop video production 1531 0.93 0.86 0 4 1.26 2.24 Email to other school or district staff 1531 3.86 0.50 0 4 -4.29 20.41 Email to students or parents 1531 2.42 1.24 0 4 -0.12 -1.27 Presentations 1531 2.16 1.05 0 4 0.25 -0.95 Research 1531 3.13 0.95 0 4 -0.83 -0.20 Analysis of student assessment information 1531 3.41 0.89 0 4 -1.50 1.53 Video conferencing 1531 0.33 0.68 0 4 2.59 7.79 Webpage publishing 1531 1.15 1.01 0 4 1.35 1.72 Degree students use Drill and practice software 1531 2.85 1.07 0 4 -0.99 0.54 Degree students use Integrated Learning Systems 1531 2.79 1.28 0 4 -0.96 -0.13 Degree students use Multimedia 1531 1.40 1.11 0 4 0.57 -0.48 Degree students use presentation 1531 1.44 1.09 0 4 0.69 -0.28 Degree students use Simulation software 1531 0.80 1.04 0 4 1.18 0.44 Degree students use Tool-based software 1531 2.16 1.17 0 4 -0.04 -1.01 Degree students use research software 1531 2.58 1.06 0 4 -0.32 -0.84

PAGE 390

Appendix C: Data Preparation Procedures (Continued) 376 Label N Mean Std Dev Min Max Skew Kurt % of technology $ devoted to professional development 1531 13.11 17.05 0 100 2.43 7.61 Level of school-based technical support 1531 3.23 1.06 0 5 -0.04 -0.95 Level of school-based instructional technology specialist support 1531 1.70 1.19 0 4 0.46 -0.79 Level of dependability of the Internet connection 1531 3.62 0.67 0 4 -1.94 4.13 Degree of delays when using the Internet 1531 2.71 0.99 0 4 -1.42 1.56 Time at your school for a technical issue to be resolved 1531 2.56 1.15 1 4 -0.03 -1.43 High Modern multi-media computers in Media center (desktops) 350 24.52 24.38 0 180 1.60 5.10 Modern multi-media computers in Media center (laptops) 350 4.47 25.97 0 446 14.46 241.21 Modern multi-media computers in Classrooms (desktops) 350 98.98 145.00 0 909 2.37 6.82 Modern multi-media computers in Classrooms (laptops) 350 29.54 192.34 0 2580 10.82 129.48 Modern multi-media computers in Computer labs primarily serving general education (desktops) 350 54.67 77.96 0 550 2.92 11.55 Modern multi-media computers in Computer labs primarily serving general education (laptops) 350 6.83 45.66 0 600 10.38 119.56 Modern multi-media computers in Mobile computer labs (desktops) 350 47.09 110.95 0 1204 4.93 37.05 Modern multi-media computers in Mobile computer labs (laptops) 350 12.53 80.63 0 925 7.87 68.44 Older computer or not multi-media in Media center (desktops) 350 10.47 17.34 0 99 2.04 4.36 Older computer or not multi-media in Media center (laptops) 350 1.04 5.34 0 56 6.80 52.28 Older computer or not multi-media in Classrooms (desktops) 350 70.60 115.66 0 722 2.70 8.78 Older computer or not multi-media in Classrooms (laptops) 350 3.01 16.45 0 260 11.83 173.79 Older computer or not multi-media in Computer labs primarily serving general education (desktops) 350 23.33 44.93 0 411 3.42 18.30 Older computer or not multi-media in Computer labs primarily serving general education (laptops) 350 0.27 2.99 0 45 12.70 171.89 Older computer or not multi-media in Mobile computer labs (desktops) 350 9.02 34.15 0 474 8.86 104.98

PAGE 391

Appendix C: Data Preparation Procedures (Continued) 377 Label N Mean Std Dev Min Max Skew Kurt Older computer or not multi-media in Mobile computer labs (laptops) 350 5.97 105.57 0 1975 18.70 349.70 Percent student computers with Concept mapping software 350 1.39 1.37 0 4 0.93 -0.43 Percent student computers with Graphics software 350 3.54 1.02 0 4 -1.99 2.35 Percent student computers with Multimedia authoring software 350 1.93 1.35 0 4 0.46 -1.22 Percent student computers with Presentation software 350 3.91 0.40 1 4 -5.62 34.55 Percent student computers with Spreadsheet software 350 3.94 0.30 1 4 -6.01 41.58 Percent student computers with Video editing software 350 1.81 1.28 0 4 0.76 -0.90 Percent student computers with Web authoring software 350 1.60 1.10 0 4 1.13 0.18 Percent student computers with Basic word processing software 350 3.97 0.29 1 4 -9.83 98.48 Percent student computers with Robust word processing software 350 3.86 0.61 0 4 -5.03 26.10 Percent student computers with FCAT Explorer software 350 3.82 0.62 0 4 -3.73 13.78 Percent student computers with Other test prep tools software 350 2.67 1.37 0 4 -0.44 -1.29 Percent student computers with Integrated Learning Systems software 350 1.63 1.15 0 4 0.91 -0.13 Percent student computers with Content-specific skills practice/tutorials software 350 1.60 1.24 0 4 0.83 -0.41 Percent student computers with Content-specific simulation software 350 1.14 1.03 0 4 1.37 1.77 Percent student computers with Other content-specific resources software 350 1.58 1.26 0 4 0.86 -0.40 Percent student computers with General Reference tools software 350 3.15 1.33 0 4 -1.18 -0.24 Administrative tasks 350 3.93 0.34 1 4 -6.60 50.04 Delivery of lessons 350 2.69 0.91 0 4 -0.32 -0.45 Desktop video production 350 1.28 0.75 0 4 1.67 3.37 Email to other school or district staff 350 3.86 0.54 0 4 -4.78 25.32 Email to students or parents 350 3.04 1.03 0 4 -0.77 -0.46 Presentations 350 2.72 0.93 0 4 -0.21 -0.74 Research 350 3.30 0.86 0 4 -1.12 0.66 Analysis of student assessment information 350 3.30 0.94 0 4 -1.23 0.66

PAGE 392

Appendix C: Data Preparation Procedures (Continued) 378 Label N Mean Std Dev Min Max Skew Kurt Video conferencing 350 0.44 0.62 0 4 1.53 3.71 Webpage publishing 350 1.37 0.88 0 4 1.58 2.38 Degree students use Drill and practice software 350 2.61 1.30 0 4 -0.64 -0.76 Degree students use Integrated Learning Systems 350 2.77 1.27 0 4 -0.81 -0.47 Degree students use Multimedia 350 2.92 1.19 0 4 -0.85 -0.41 Degree students use presentation 350 3.00 1.04 0 4 -0.79 -0.48 Degree students use Simulation software 350 1.13 1.24 0 4 0.83 -0.47 Degree students use Tool-based software 350 3.51 0.82 0 4 -1.95 3.69 Degree students use research software 350 3.61 0.70 0 4 -2.02 4.34 % of technology $ devoted to professional development 350 11.04 10.96 0 50 1.34 2.07 Level of school-based technical support 350 3.66 1.01 1 5 -0.53 -0.76 Level of school-based instructional technology specialist support 350 2.15 1.24 0 4 0.01 -1.26 Level of dependability of the Internet connection 350 3.66 0.60 1 4 -1.72 2.58 Degree of delays when using the Internet 350 2.76 0.93 0 4 -1.39 1.91 Time at your school for a technical issue to be resolved 350 2.90 1.10 1 4 -0.51 -1.10 Middle Modern multi-media computers in Media center (desktops) 446 15.04 17.47 0 186 3.30 22.96 Modern multi-media computers in Media center (laptops) 446 2.00 14.63 0 280 16.13 297.28 Modern multi-media computers in Classrooms (desktops) 446 73.68 93.17 0 538 1.66 2.56 Modern multi-media computers in Classrooms (laptops) 446 11.84 61.25 0 967 10.25 139.26 Modern multi-media computers in Computer labs primarily serving general education (desktops) 446 39.89 50.50 0 685 5.26 59.39 Modern multi-media computers in Computer labs primarily serving general education (laptops) 446 1.96 15.74 0 255 12.33 174.96 Modern multi-media computers in Mobile computer labs (desktops) 446 33.83 72.36 0 560 3.47 14.56 Modern multi-media computers in Mobile computer labs (laptops) 446 9.85 50.69 0 460 5.94 36.68 Older computer or not multi-media in Media center (desktops) 446 6.67 10.43 0 74 2.20 6.28

PAGE 393

Appendix C: Data Preparation Procedures (Continued) 379 Label N Mean Std Dev Min Max Skew Kurt Older computer or not multi-media in Media center (laptops) 446 1.10 7.50 0 128 12.28 189.48 Older computer or not multi-media in Classrooms (desktops) 446 58.39 81.43 0 542 2.05 5.18 Older computer or not multi-media in Classrooms (laptops) 446 2.52 17.13 0 240 10.19 117.70 Older computer or not multi-media in Computer labs primarily serving general education (desktops) 446 17.33 27.85 0 134 1.70 2.30 Older computer or not multi-media in Computer labs primarily serving general education (laptops) 446 0.51 4.03 0 64 11.34 153.58 Older computer or not multi-media in Mobile computer labs (desktops) 446 8.52 27.83 0 290 6.00 45.98 Older computer or not multi-media in Mobile computer labs (laptops) 446 1.16 12.76 0 230 15.09 250.73 Percent student computers with Concept mapping software 446 1.52 1.51 0 4 0.65 -1.07 Percent student computers with Graphics software 446 3.62 0.98 0 4 -2.49 4.89 Percent student computers with Multimedia authoring software 446 1.84 1.48 0 4 0.34 -1.35 Percent student computers with Presentation software 446 3.87 0.47 0 4 -4.90 28.76 Percent student computers with Spreadsheet software 446 3.88 0.46 0 4 -4.73 25.76 Percent student computers with Video editing software 446 1.76 1.40 0 4 0.57 -1.13 Percent student computers with Web authoring software 446 1.05 1.18 0 4 1.39 1.17 Percent student computers with Basic word processing software 446 3.95 0.33 0 4 -8.34 78.17 Percent student computers with Robust word processing software 446 3.81 0.70 0 4 -4.27 18.35 Percent student computers with FCAT Explorer software 446 3.86 0.56 0 4 -4.75 23.80 Percent student computers with Other test prep tools software 446 2.89 1.40 0 4 -0.85 -0.80 Percent student computers with Integrated Learning Systems software 446 2.03 1.32 0 4 0.33 -1.22 Percent student computers with Content-specific skills practice/tutorials software 446 2.06 1.41 0 4 0.26 -1.40 Percent student computers with Content-specific simulation software 446 1.21 1.32 0 4 1.13 0.10

PAGE 394

Appendix C: Data Preparation Procedures (Continued) 380 Label N Mean Std Dev Min Max Skew Kurt Percent student computers with Other content-specific resources software 446 1.93 1.49 0 4 0.28 -1.41 Percent student computers with General Reference tools software 446 3.20 1.29 0 4 -1.31 0.17 Administrative tasks 446 3.89 0.37 1 4 -4.29 21.95 Delivery of lessons 446 2.72 0.91 0 4 -0.30 -0.61 Desktop video production 446 1.22 0.86 0 4 1.14 1.49 Email to other school or district staff 446 3.90 0.42 1 4 -4.77 24.65 Email to students or parents 446 2.96 1.14 0 4 -0.76 -0.65 Presentations 446 2.65 0.96 1 4 -0.17 -0.92 Research 446 3.28 0.83 1 4 -0.94 0.15 Analysis of student assessment information 446 3.41 0.85 1 4 -1.32 0.82 Video conferencing 446 0.40 0.63 0 4 1.92 5.23 Webpage publishing 446 1.26 0.98 0 4 1.19 1.25 Degree students use Drill and practice software 446 2.55 1.24 0 4 -0.57 -0.69 Degree students use Integrated Learning Systems 446 2.72 1.23 0 4 -0.72 -0.51 Degree students use Multimedia 446 2.13 1.24 0 4 -0.04 -1.14 Degree students use presentation 446 2.49 1.09 0 4 -0.16 -1.14 Degree students use Simulation software 446 1.11 1.19 0 4 0.81 -0.43 Degree students use Tool-based software 446 2.96 1.06 0 4 -0.78 -0.31 Degree students use research software 446 3.24 0.92 1 4 -1.04 0.10 % of technology $ devoted to professional development 446 12.94 16.05 0 100 2.51 8.61 Level of school-based technical support 446 3.55 1.05 0 5 -0.57 -0.38 Level of school-based instructional technology specialist support 446 2.11 1.26 0 4 0.05 -1.25 Level of dependability of the Internet connection 446 3.66 0.63 1 4 -1.92 3.50 Degree of delays when using the Internet 446 2.78 0.93 0 4 -1.57 2.51 Time at your school for a technical issue to be resolved 446 2.89 1.06 1 4 -0.41 -1.13 Note. skewness > 1 ** kurtosis > 3 In 2003, eighteen schools had missing variables for number of students. Three of these schools also had missing variables for number of students for 2004-05 and 200-06. These schools were inspected

PAGE 395

Appendix C: Data Preparation Procedures (Continued) 381 and found to be PreK schools, so they were removed from the data set. Two schools reported zero students. These schools were inspect ed and found to be 9th grade centers that were associat ed with high schools. It is assumed that these schools have their students reported in the high school count, so they were removed from the analysis. The other 15 schools were matched with schools in 2004-05 and 2005-06 to see if the number of students had been provided. All of these schools had the number of students provided in both years, so the missing number of studen ts in 2003-04 was replaced with the number of st udents in 2004-05. To obtain counts of computers, all of the variab les for modern computer types were summed and then all of the old computer types were summed. The results were added together to get a total computer count. If the total number of computers was not equal to zero, the number of students in the school was divided by this total computer count. In 2003, eighteen schools had the variable for number of students missing. Seven schools reported they had no computers. These schools generated missing data. The data were sorted by students per computer, and then the dataset was visually inspected. There appeared to be many entries that had the exact same number for modern computers and non-modern computers. An additional variable was made to determine which schools had the exact same number entered for each computer type for both modern and non-modern. There were 98 schools that had entered information into the STAR survey this way. In 2004-05, three schools had no information for the number of students, one school had no computers, and 50 schools had the same data for modern and non modern computers. In 2005-06, four schools had no computers, three had no information about the number of students, and 70 had the same number of computers for both modern and non modern computer types. In 2006-07, three schools reported that they had no computers, all schools supplied the number of students enrolled, and only 7 schools entered the same information for modern and non modern computers. It was decided to delete the schools that appeared to have the same information entered twice; 2125 schools remained in the dataset, and the descriptive statistics were run again. Further inspection of the computer counts for just the modern computers found that 25 schools had no modern computers, 27 additional schools reported the same number of computers in media centers and in the general education labs, another 10 schools reported the same number of computers in media centers and mobile labs, 13 additional schools reported the exact same number for media centers and classrooms,

PAGE 396

Appendix C: Data Preparation Procedures (Continued) 382 and 5 schools reported the exact same number for classrooms and mobile labs. This means that 55 schools that had modern computers reported the same number of computers in different categories. Although this information may be accurate, it may also indicate that the person completing the survey misunderstood the purpose of the item by focusing on the use of the computer as opposed to th e location or category of computer. Some schools may have used the computer s in their media centers as labs for classroom teachers, and others may have viewed this same scen ario as the media center being used as a classroom. Other schools may have housed their mobile computer labs in the media center or in the classroom and included these computers in the counts for the storage location as well as the mobile lab count. In addition, 3 schools indicated that they had more modern student computers than students, and 39 schools indicated that they had less than 2 students per modern computer. Additional years were investigated to determine if the counts of computers exhibited similar patterns. In 2004-05, 21 schools reported they had no modern computers or non-modern computers, 61 schools had the exact same number of modern and non modern computers. When each indicator of modern and non-modern computers were compared separately after schools that reported having no computers were removed, 50 schools reported the exact same number of modern and non-modern desktop computers in each of the following location: media centers, classr ooms, general education labs, and mobile labs; and 1194 schools reported the exact same number of lapt ops in each of the followin g locations: media centers, classrooms, general education labs, and mobile labs. When the separate categories of modern computers was examined after schools with no computers were removed, 18 schools reported the same number of desktop computers in media centers and general education computer labs, 6 schools reported the same number of desktop computers in the media centers and classrooms, 12 schools reported the same number of laptop computers in the media center and mobile com puter labs 51 schools reported the exact same number of laptop computers in classrooms and mobile computer labs, 11 schools reported the same number of laptop and desktop computers in media centers, general education labs, and mobile computer labs, and 5 schools reported the exact same numbers of computer s in media centers, classrooms, and mobile computer labs. These categories were mutually exclusive, so in all, there were 103 schools that appeared to report

PAGE 397

Appendix C: Data Preparation Procedures (Continued) 383 computers in multiple categories. In addition, 12 schools reported having more modern computers than students; while 87 more schools had less than 2 students per modern computer. In 2005-06, four schools had no computers and 69 had no modern computers, and 72 schools reported having the exact same number of modern and non-modern computers. Sixty-eight schools reported the exact same number of modern and non-modern student desktop computers in media centers, classrooms, and general education labs. Eighteen schools reported having the same number of modern and non-modern laptop computers in media centers, classrooms, and mobile computer carts. When examining only numbers of modern student computers, 27 schools reported the exact same number of computers in media centers and general education labs; 17 schools reported the exact same number of computers in media centers and mobile computer labs; 12 schools reported the exact same number of computers in media centers and classrooms; 37 schools reported the same number of computers in media centers, general education labs, and mobile computer labs; and two reported the same number of computers in media centers, classrooms, general education labs, and mob ile computer labs. These categories were mutually exclusive, so in all, 95 schools appear to have re ported having the same computers in multiple places. In addition after excluding 69 schools that reported ha ving no modern computers, three schools reported having more modern student computers than students, and 68 more schools reported having less than 2 students per modern computer. In 2006-07, six schools reported on the STAR survey that they had no students, and another school reported that it had one student enrolled. Three schools reported that they had no modern or nonmodern computers, and only 7 schools entered the same information for modern and non modern computers. When this was broken down by specific categories of modern and non-modern computers only two had the exact same counts for modern and non-modern student desktop computers in media centers, classrooms, general education labs and mobile computer labs, and no laptops had the exact same number for category. One hundred twenty-six schools reported having no modern computers. After these schools were excluded, 3 schools reported the exact same number of modern student computers in media centers and general education labs; 4 reported the same numbers of modern computers in the media center and mobile computer lab; 8 schools reported the exact same number of computers in media centers and

PAGE 398

Appendix C: Data Preparation Procedures (Continued) 384 classrooms; 20 schools reported the exact same num ber of computers in the classrooms and mobile computer labs; 39 schools reported the exact same number of computers in the media centers, general education labs, and mobile computer labs, while only one school reported the same counts between media centers, classrooms and mobile computer labs, and only one school reported the same counts of modern student computers in media centers, classrooms, general education labs, and mobile computer labs. These categories were mutually exclusive, so that in all, 76 schools seemed to report computers in more than one category. In addition after excluding the schools that reported having no students and schools reporting no modern student computers, 21 schools reported having more modern student computers than students, and an additional 91 schools reported having le ss than two students per modern computer. Each year seemed to follow the same pattern with some schools reporting computers in more than one category, and some schools reporting having more modern computers than students. Also, each year the number of schools with less than 2 students per computer has increased. Removing all schools with questionable entries, no modern computers, or with no students would decrease the sample for the study from 2327 schools to 1841 or a reduction of almost 21%. Skewness of the technology indicator variables ranged between -6.14 and 42.53; and kurtosis ranged between -3.34 and 1819.36. Although was an improvement, the data were still not normal. Since it cannot be verified that the counts of the computers by the schools identified have been duplicated, nor can it be verified that the schools not identified entered accurate information, choosing to remove 21% of the schools does not seem to be a viab le alternative. Therefore, for th is study, the category of access to computer hardware was removed from the analysis. It was assumed that access to the software was an adequate proxy for access to the hardware. However w ith the removal of the variables of computers in classrooms and computers in labs, the relationship between the location of the computer and school level achievement cannot be determined in this study. Because many of the variables in this study were not normally dist ributed, the data were transformed using the natural log. To determine if this transformation was necessary the exploratory factor analysis was conducted with both the original data an d the transformed data. The results were the same, so the original data were used in the rest of the analysis.

PAGE 399

Appendix C: Data Preparation Procedures (Continued) 385 Exploratory Factor Analysis The results of the exploratory factor analysis ar e delineated for each year for each of the composite variables used in the analysis. Student learning environment. First, exploratory factor analysis was conducted with the variables predicted to measure the aspects of the student lear ning environment for each year separately. These variables had correlated communality estimates that ranged from .21 to .57 in 2003-04, .20 to .55 in 200405, and .19 to .56 in 2005-06. Data for 2006-07 was not available. The standardized regression coefficients of the rotated factor patterns and the correlations of the factor structure obtain ed from the common factor analysis with oblique rotation are depicted in Table C 15 For all three years, both the original data and the transformed data loaded on only one factor. One variable was dropped because factor loadings were below .3. That was the ratio of students per instructional staff. A composite score was created to measure the positive student learning environment by summing the percentage of students who did not serve out-ofschool suspensions, the percentage of students who did not serve in-school suspensions, the percentage of students who were not absent more than 21 days, and the percentage stability rate. The total crime incidents per student (times 100) was subtracted from the result. Table C 15. Common Factor Analysis with Oblique Rotation: Student Learning Environments Factor Pattern Item 2003 2004 2005 Percent of students with out-of-school suspensions 82 82 75 Percent of students with in-house suspensions 65 63 63 Percent of Students with Over 21 Days Absences 69 58 58 Total Crime Incidents/student 58 57 49 Percent Instability 47 48 46 Note: Printed values are multiplied by 100 and ro unded to the nearest integer. Values greater than 0.3 are flagged by an '*'.

PAGE 400

Appendix C: Data Preparation Procedures (Continued) 386 Teacher qualifications. Next, exploratory factor analysis was conducted with the variables predicted to measure the teacher qualifications for each year separately. These variables had correlated communality estimates that ranged from .03 to .06 in 2003-04, .06 to .10 in 2004-05, and .04 to .08 in 2005-06. Data for 2006-07 were not available. The standa rdized regression coefficien ts of the rotated factor patterns and the correlations of the f actor structure obtained from the co mmon factor analysis with oblique rotation are depicted in Table C 16. For all three years, only one factor was obtained with both the original data and the transformed data. Table C 16. Common Factor Analysis with Oblique Rotation: Teacher Qualifications Factor Pattern Item 2003 2004 2005 Percent of Teachers with an advanced degree 39 45 43 Average number of years experience 38 45 33 Percent of classes taught by teach ers with certification 31 38 34 Note: Printed values are multiplied by 100 and rounde d to the nearest integer. Values greater than 0.3 are flagged by an '*'. Student access to software. Exploratory factor analysis was condu cted with variables predicted to measure the types of software available on student computers for each year separately. The variables had correlated communality estimates that ranged from .13 to .57 in 2003-04, .09 to .56 in 2004-05, .13 to .60 in 2005-06, and .11 to .54 in 2006-07. The standardized regression coefficients of the rotated factor patterns and the correlations of the factor structure obtained from the common factor analysis with oblique rotation are depicted in Table C 17. For consistency of interpre tability over all four years, the number of factors was specified to be three. Both the original data and the transformed data loaded the same three factors on the same variables: Content Software, Office/ Productio n Software, and Advanced Production Software. These three factors were used in the multi-level modeling analysis as separate composite variables. After designating all missing items as zero, the composite was made from the mean of all of the included variables.

PAGE 401

Appendix C: Data Preparation Procedures (Continued) 387 Table C 17. Common Factor Analysis with Oblique Rotation: Student Access to Software Rotated Factor Pattern (Standardized Regression Coefficients) Factor St ructure (Correlations) Item Content Software Office/ Production Software Advanced Production Software Content Software Office/ Production Software Advanced Production Software 2003 Percent student computers with Other test prep tools software 66* 1 -11 62* 7 18 Percent student computers with Contentspecific skills practice/tutorials software 66* -5 4 67* 6 31* Percent student computers with Contentspecific simulation software 63* -8 11 67* 6 36* Percent student computers with Other content-specific resources software 63* -5 14 68* 9 39* Percent student computers with FCAT Explorer software 57* 18 -17 52* 20 15 Percent student computers with Integrated Learning Systems software 54* -5 0 54* 3 22 Percent student computers with General Reference tools software 50* 16 6 55* 25 33* Percent student computers with Spreadsheet software -1 80* 2 11 81* 30* Percent student computers with Presentation software 2 72* -4 11 71* 23 Percent student computers with Robust word processing software -2 63* 6 10 65* 28 Percent student computers with Graphics software 6 35* 20 20 43* 35*

PAGE 402

Appendix C: Data Preparation Procedures (Continued) 388 Rotated Factor Pattern (Standardized Regression Coefficients) Factor Structure (Correlations) Item Content Software Office/ Production Software Advanced Production Software Content Software Office/ Production Software Advanced Production Software Percent student computers with Basic word processing software 3 31* 5 10 33* 18 Percent student computers with Multimedia authoring software 3 4 56* 28 25 59* Percent student computers with Video editing software -5 9 48* 17 26 49* Percent student computers with Concept mapping software 9 -2 46* 29 16 49* Percent student computers with Web authoring software -7 17 39* 13 30* 42* 2004 Percent student computers with Contentspecific skills practice/tutorials software 69* -7 2 69* 4 28 Percent student computers with Other content-specific resources software 68* -3 7 71* 10 34* Percent student computers with Other test prep tools software 66* 2 -9 63* 8 19 Percent student computers with Contentspecific simulation software 64* -3 8 66* 9 33* Percent student computers with Integrated Learning Systems software 58* -4 -4 56* 3 19 Percent student computers with FCAT Explorer software 51* 15 -10 49* 19 17 Percent student computers with General Reference tools software 48* 12 10 54* 23 34*

PAGE 403

Appendix C: Data Preparation Procedures (Continued) 389 Rotated Factor Pattern (Standardized Regression Coefficients) Factor Structure (Correlations) Item Content Software Office/ Production Software Advanced Production Software Content Software Office/ Production Software Advanced Production Software Percent student computers with Spreadsheet software -1 81* 1 11 81* 30* Percent student computers with Presentation software 4 73* -6 13 72* 23 Percent student computers with Robust word processing software 0 57* 9 12 60* 30* Percent student computers with Graphics software 3 28 27 18 38* 39* Percent student computers with Basic word processing software 2 25 9 9 29 19 Percent student computers with Multimedia authoring software 4 0 55* 26 21 57* Percent student computers with Video editing software -6 7 52* 16 25 52* Percent student computers with Concept mapping software 12 -2 46* 31* 16 50* Percent student computers with Web authoring software -8 15 38* 10 27 40* 2005 Percent student computers with Spreadsheet software 83* -4 3 83* 13 30* Percent student computers with Presentation software 78* 1 -6 76* 15 22 Percent student computers with Robust word processing software 59* -5 10 62* 11 28 Percent student computers with Basic word processing software 35* 8 -1 37* 15 15

PAGE 404

Appendix C: Data Preparation Procedures (Continued) 390 Rotated Factor Pattern (Standardized Regression Coefficients) Factor Structure (Correlations) Item Content Software Office/ Production Software Advanced Production Software Content Software Office/ Production Software Advanced Production Software Percent student computers with Graphics software 29 16 16 38* 28 33* Percent student computers with Other content-specific resources software 3 65* -3 15 65* 25 Percent student computers with Contentspecific skills practice/tutorials software -7 62* -1 5 60* 22 Percent student computers with Contentspecific simulation software -1 51* 6 12 54* 27 Percent student computers with Other test prep tools software 8 43* -4 15 43* 16 Percent student computers with General Reference tools software 14 42* 10 26 49* 32* Percent student computers with Integrated Learning Systems software -6 36* 8 3 38* 20 Percent student computers with FCAT Explorer software 24 31* -6 28 33* 15 Percent student computers with Multimedia authoring software -4 1 70* 22 30 70* Percent student computers with Video editing software 4 -5 67* 27 24 66* Percent student computers with Concept mapping software -2 21 41* 17 38* 49* Percent student computers with Web authoring software 15 -1 36* 28 17 41* 2006 Percent student computers with Spreadsheet software 79* -5 3 79* 10 26

PAGE 405

Appendix C: Data Preparation Procedures (Continued) 391 Rotated Factor Pattern (Standardized Regression Coefficients) Factor Structure (Correlations) Item Content Software Office/ Production Software Advanced Production Software Content Software Office/ Production Software Advanced Production Software Percent student computers with Presentation software 76* -1 -5 74* 10 20 Percent student computers with Robust word processing software 45* 1 13 50* 14 28 Percent student computers with Basic word processing software 34* 6 0 35* 12 14 Percent student computers with Graphics software 25 13 20 33* 26 33* Percent student computers with Contentspecific skills practice/tutorials software -6 60* -1 4 59* 23 Percent student computers with Other content-specific resources software -1 59* 4 10 61* 29 Percent student computers with Other test prep tools software 8 50* -12 12 46* 12 Percent student computers with Contentspecific simulation software 3 43* 11 13 48* 30 Percent student computers with General Reference tools software 13 40* 12 23 47* 33* Percent student computers with Integrated Learning Systems software -4 38* 1 3 38* 16 Percent student computers with FCAT Explorer software 24 27 -5 27 29 14 Percent student computers with Multimedia authoring software -4 1 74* 21 32* 73*

PAGE 406

Appendix C: Data Preparation Procedures (Continued) 392 Rotated Factor Pattern (Standardized Regression Coefficients) Factor Structure (Correlations) Item Content Software Office/ Production Software Advanced Production Software Content Software Office/ Production Software Advanced Production Software Percent student computers with Video editing software 4 -6 66* 25 23 65* Percent student computers with Concept mapping software -3 15 46* 14 34* 52* Percent student computers with Web authoring software 10 -6 41* 22 13 41* Note: Printed values are multiplied by 100 and rounded to the nearest integer. Values greater than 0.3 are flagged by an '*'. Teachers regularly use technology. Exploratory factor analysis was conducted with variables predicted to measure the percent of teachers who regularly use technology for different tasks. The exploratory factor analysis was conducted for eac h year separately. The variables had correlated communality estimates that ranged from .14 to .47 in 2003-04, .14 to .47 in 2004-05, .15 to .49 in 2005-06, and .13 to .49 in 2006-07. The standardized regression coefficients of the rotated factor patterns and the correlations of the factor structur e obtained from the co mmon factor analysis with oblique rotation are depicted in Table C 18. For consistency of interpreta bility over all four years, the number of factors was specified to be two. Both the original data and the transformed data loaded the same two factors with the same variables: delivery of instruction and administrative purposes. These two factors were used in the multi-level modeling analysis as separate composite variables. After designating all missing items as zero, the composite was made from the mean of all of the included variables. Table C 18. Common Factor Analysis with Oblique Rotation: Teachers Regularly use Technology Rotated Factor Pattern (Standardized Regression Coefficients) Fact or Structure (Correlations) Items Delivery of Instruction Administrative Purposes Delivery of Instruction Administrative Purposes 2003 Presentations 63 15 73 56

PAGE 407

Appendix C: Data Preparation Procedures (Continued) 393 Rotated Factor Pattern (Standardized Regression Coefficients) Fact or Structure (Correlations) Items Delivery of Instruction Administrative Purposes Delivery of Instruction Administrative Purposes Delivery of lessons 58 12 66 50 Desktop video production 53 -2 52 33 Video conferencing 48 -11 41 21 Webpage publishing 37 13 45 37 Email to other school or district staff -13 67 31 59 Administrative tasks 1 53 36 54 Email to students or parents 14 48 45 57 Research 29 33 50 52 Analysis of student assessment information 20 32 40 44 2004 Desktop video production 57 -8 52 29 Presentations 56 20 69 56 Delivery of lessons 50 25 66 57 Video conferencing 50 -15 41 17 Webpage publishing 38 8 43 32 Email to other school or district staff -13 62 27 54 Administrative tasks -8 61 31 56 Analysis of student assessment information 6 44 34 47 Research 26 38 50 55 Email to students or parents 25 33 46 49 2005 Presentations 63 15 73 56 Desktop video production 63 -7 59 33 Delivery of lessons 58 17 69 55 Video 50 -14 41 18

PAGE 408

Appendix C: Data Preparation Procedures (Continued) 394 Rotated Factor Pattern (Standardized Regression Coefficients) Fact or Structure (Correlations) Items Delivery of Instruction Administrative Purposes Delivery of Instruction Administrative Purposes conferencing Webpage publishing 34 15 44 37 Email to other school or district staff -13 61 26 53 Administrative tasks -5 54 29 51 Research 20 47 50 60 Analysis of student assessment information 5 45 33 48 Email to students or parents 21 37 45 51 2006 Desktop video production 64 -13 55 29 Presentations 55 26 72 62 Video conferencing 51 -17 40 16 Delivery of lessons 46 29 65 60 Webpage publishing 40 1 41 28 Administrative tasks -15 52 19 42 Email to other school or district staff -14 51 20 42 Research 18 48 50 60 Analysis of student assessment information 4 43 33 46 Email to students or parents 24 29 42 44 Note: Printed values are multiplied by 100 and rounde d to the nearest integer. Values greater than 0.3 are flagged by an '*'. Frequency students use software. Exploratory factor analysis was conducted with variables predicted to measure the frequency that students use di fferent types of software for each year separately. The variables had correlated communality estimates that ranged from .03 to .21 in 2003-04, .13 to .29 in 2004-05, .15 to .38 in 2005-06, and .17 to .38 in 2006-07. The standardized regression coefficients of the

PAGE 409

Appendix C: Data Preparation Procedures (Continued) 395 rotated factor patterns and the correlations of the fact or structure obtained from the common factor analysis with oblique rotation are depicted in Table C 19. For co nsistency of interpretability over all four years, the number of factors was specified to be two. Both the original data and the transformed data loaded the same two factors with the same variables: production tool and content delivery. These two factors were used in the multi-level modeling analysis as separate composite variables. The composite was made from the sum of all of the included variables. Table C 19. Common Factor Analysis with Oblique Rotation: Frequency Students Use Software Rotated Factor Pattern (Standardized Regression Coefficients) Factor Structure (Correlations) Item Production Tool Content Delivery Production Tool Content Delivery 2003 Frequency students use Multimedia 58 1 58 4 Frequency students use Tool-based software 49 -14 49 -11 Frequency students use Simulation software 39 20 40 22 Frequency students use Drill and practice software -4 36 -2 36 Frequency students use Integrated Learning Systems 3 30 5 30 2004 Frequency students use Multimedia 65 0 65 23 Frequency students use Tool-based software 63 -4 61 18 Frequency students use Simulation software 38 13 43 27 Frequency students use Drill and practice software -1 49 16 49 Frequency students use Integrated Learning Systems 5 47 21 48 2005 Frequency students use Multimedia 70 -2 69 30 Frequency students use Tool-based software 65 1 66 31 Frequency students use Simulation software 41 8 45 27

PAGE 410

Appendix C: Data Preparation Procedures (Continued) 396 Rotated Factor Pattern (Standardized Regression Coefficients) Factor St ructure (Correlations) Item Production Tool Content Delivery Production Tool Content Delivery Frequency students use Drill and practice software 2 50 24 50 Frequency students use Integrated Learning Systems 4 48 26 50 2006 Frequency students use Multimedia 71 0 71 35 Frequency students use Tool-based software 67 -2 66 31 Frequency students use Simulation software 37 19 46 37 Frequency students use Drill and practice software 2 52 28 53 Frequency students use Integrated Learning Systems 3 51 28 52 Note: Printed values are multiplied by 100 and roun ded to the nearest integer. Values greater than 0.3 are flagged by an '*'. Support for Technology. Exploratory factor analysis was condu cted with variables predicted to measure the level of school support for technology for each year separately. The variables had correlated communality estimates that ranged from .03 to .21 in 2003-04, .13 to .29 in 2004-05, .15 to .38 in 2005-06, and .17 to .38 in 2006-07. The standardized regression coefficients of the rotated factor patterns and the correlations of the factor structur e obtained from the co mmon factor analysis with oblique rotation are depicted in Table C 20. For consistency of interpreta bility over all four years, the number of factors was specified to be two. One item was removed from the anal ysis because its factor loadings were less than .3. Both the original data and the transformed data load ed the same two factors with the same variables: human/ time and hardware/ Internet. These two factors were used in th e multi-level modeling analysis as separate composite variables. The composite was made from the sum of all of the included variables.

PAGE 411

Appendix C: Data Preparation Procedures (Continued) 397 Table C 20. Common Factor Analysis with Oblique Rotation: Support for Technology Rotated Factor Pattern (Standardized Regression Coefficients) Factor St ructure (Correlations) Human/ Time Hardware/ Internet Human/ Time Hardware/ Internet 2003 Level of school-based technical support 65 -5 64 -1 Level of school-based instructional technology specialist support 63 -1 63 3 Time at your school for a technical issue to be resolved 42 10 42 12 Degree of delays when using the Internet -2 58 1 57 Level of dependability of the Internet connection 5 57 9 58 2004 Level of school-based technical support 71 -3 71 2 Level of school-based instructional technology specialist support 64 -3 64 1 Time at your school for a technical issue to be resolved 48 9 49 12 Level of dependability of the Internet connection 6 53 10 54 Degree of delays when using the Internet -4 53 0 53 2005 Level of school-based technical support 68 -1 68 9 Level of school-based instructional technology specialist support 65 -5 64 4 Time at your school for a technical issue to be resolved 36 13 38 18 Level of dependability of the Internet connection 4 53 11 54 Degree of delays when using the Internet 0 51 7 51 2006 Level of school-based technical support 67 3 66 -1 Level of school-based instructional technology specialist support 60 5 60 1

PAGE 412

Appendix C: Data Preparation Procedures (Continued) 398 Rotated Factor Pattern (Standardized Regression Coefficients) Factor Structure (Correlations) Human/ Time Hardware/ Internet Human/ Time Hardware/ Internet Time at your school for a technical issue to be resolved -43 13 -44 16 Level of dependability of the Internet connection -6 52 -10 53 Degree of delays when using the Internet -4 -52 0 -52 Note: Printed values are multiplied by 100 and roun ded to the nearest integer. Values greater than 0.3 are flagged by an '*'. Descriptive Statistics of Predictor Variables Table C 21. Descriptive Statistics of Predictor Variables for FCAT Reading Outcome Variable School Level N School Year Mean STD Min Max Skew Kurt Percent Students on Free or Reduced Lunch Program All Schools 22852003-04 52.0625.251.0100.0 0.10 -0.94 2004-05 52.5324.080.9100.0 0.02 -0.90 2005-06 52.1623.891.7100.0 -0.03 -0.97 2006-07 52.1623.891.7100.0 -0.03 -0.97 Elementary 14962003-04 56.8726.131.0100.0 -0.14 -1.01 2004-05 56.4124.880.9100.0 -0.20 -0.94 2005-06 55.9924.771.7100.0 -0.24 -0.99 2006-07 55.9924.771.7100.0 -0.24 -0.99 High 3452003-04 35.4817.181.893.3 0.46 0.18 2004-05 38.8318.053.8100.0 0.58 0.63 2005-06 38.6517.212.593.7 0.17 -0.34 2006-07 38.6517.212.593.7 0.17 -0.34 Middle 4442003-04 48.7421.243.7100.0 0.12 -0.65 2004-05 50.1021.093.3100.0 0.05 -0.60 2005-06 49.7621.113.9100.0 0.02 -0.79 2006-07 49.7621.113.9100.0 0.02 -0.79 Percent Minority Students All Schools 22862003-04 50.1828.200.0100.0 0.31 -1.11 2004-05 51.3528.370.0100.0 0.28 -1.14 2005-06 52.3828.380.0100.0 0.23 -1.16 2006-07 52.3828.380.0100.0 0.23 -1.16 Elementary 14962003-04 51.9228.970.0100.0 0.21 -1.22 2004-05 53.0629.100.0100.0 0.19 -1.23 2005-06 54.2928.950.0100.0 0.14 -1.24 2006-07 54.2928.950.0100.0 0.14 -1.24 High 3462003-04 44.6626.012.6100.0 0.60 -0.63

PAGE 413

Appendix C: Data Preparation Procedures (Continued) 399 Variable School Level N School Year Mean STD Min Max Skew Kurt 2004-05 45.5926.470.799.9 0.55 -0.75 2005-06 45.9426.991.099.9 0.50 -0.82 2006-07 45.9426.991.099.9 0.50 -0.82 Middle 4442003-04 48.5926.584.8100.0 0.41 -0.89 2004-05 50.0826.624.699.8 0.36 -0.96 2005-06 50.9726.685.699.9 0.32 -1.00 2006-07 50.9726.685.699.9 0.32 -1.00 Percent LEP students All Schools 21202003-04 8.7510.690.063.3 1.97 3.96 2004-05 8.5310.560.061.6 2.01 4.18 2005-06 8.8110.630.065.7 1.98 4.12 2006-07 8.8110.630.065.7 1.98 4.12 Elementary 13742003-04 10.8312.220.163.3 1.56 1.96 2004-05 10.5212.070.061.6 1.60 2.14 2005-06 10.8312.110.165.7 1.57 2.15 2006-07 10.8312.110.165.7 1.57 2.15 High 3252003-04 4.504.780.023.1 1.48 1.75 2004-05 4.354.660.026.0 1.54 2.11 2005-06 4.434.670.028.3 1.61 2.75 2006-07 4.434.670.028.3 1.61 2.75 Middle 4212003-04 5.235.450.132.5 1.69 3.50 2004-05 5.185.440.136.4 1.89 4.87 2005-06 5.475.710.137.4 2.04 5.74 2006-07 5.475.710.137.4 2.04 5.74 Percent Students with Disabilities All Schools 22852003-04 15.515.410.540.6 0.69 1.44 2004-05 15.305.320.443.9 0.74 1.99 2005-06 15.315.440.372.6 1.52 9.77 2006-07 15.315.440.372.6 1.52 9.77 Elementary 14962003-04 16.045.711.240.6 0.81 1.28 2004-05 15.815.611.643.9 0.92 1.95 2005-06 16.045.771.972.6 1.81 10.63 2006-07 16.045.771.972.6 1.81 10.63 High 3452003-04 13.264.430.731.1 0.03 0.96 2004-05 13.204.470.427.6 -0.16 0.42 2005-06 13.164.480.330.9 -0.04 0.66 2006-07 13.164.480.330.9 -0.04 0.66 Middle 4442003-04 15.444.570.527.9 -0.13 0.17 2004-05 15.234.480.528.7 -0.16 0.20 2005-06 14.514.290.828.7 -0.07 0.26 2006-07 14.514.290.828.7 -0.07 0.26 Percent Gifted students All Schools 18072003-04 4.995.770.152.3 2.91 12.23 2004-05 5.005.870.154.9 2.96 12.66 2005-06 4.915.800.057.2 3.03 13.89 2006-07 4.915.800.057.2 3.03 13.89 Elementary 13562003-04 4.325.430.152.3 3.41 17.27

PAGE 414

Appendix C: Data Preparation Procedures (Continued) 400 Variable School Level N School Year Mean STD Min Max Skew Kurt 2004-05 4.285.500.154.9 3.51 18.29 2005-06 4.205.570.157.2 3.69 20.55 2006-07 4.205.570.157.2 3.69 20.55 High 212003-04 7.0410.170.439.8 2.49 5.99 2004-05 7.6610.640.739.6 2.14 4.13 2005-06 4.705.060.037.4 2.69 11.04 2006-07 4.705.060.037.4 2.69 11.04 Middle 4302003-04 7.006.060.137.3 2.04 5.18 2004-05 7.096.180.139.1 2.10 5.59 2005-06 7.276.340.139.5 2.11 5.84 2006-07 7.276.340.139.5 2.11 5.84 Positive Learning Environment All Schools 22862003-04 363.8028.47131.4398.7 -1.56 3.58 2004-05 364.1626.65220.3398.1 -1.38 1.92 2005-06 364.7726.02221.3397.8 -1.38 1.96 2006-07 364.7726.02221.3397.8 -1.38 1.96 Elementary 14962003-04 377.4716.69241.5398.7 -2.95 13.80 2004-05 376.9514.83272.9397.9 -2.23 8.48 2005-06 376.6215.67252.8397.8 -2.34 9.06 2006-07 376.6215.67252.8397.8 -2.34 9.06 High 3462003-04 336.0526.61232.4396.0 -0.53 0.89 2004-05 338.3226.55259.0397.2 -0.49 0.26 2005-06 340.2624.88261.7396.6 -0.23 0.12 2006-07 340.2624.88261.7396.6 -0.23 0.12 Middle 4442003-04 339.3829.33131.4394.9 -1.52 6.16 2004-05 341.2227.76220.3398.1 -0.83 1.21 2005-06 343.9028.34221.3397.8 -0.94 1.19 2006-07 343.9028.34221.3397.8 -0.94 1.19 Positive Teacher Qualifications All Schools 22862003-04 140.0316.8261.3200.7 -0.66 1.69 2004-05 139.1418.0659.9194.4 -0.84 1.69 2005-06 137.4218.9640.4191.6 -0.96 1.72 2006-07 137.4218.9640.4191.6 -0.96 1.72 Elementary 14962003-04 139.1617.5761.3187.9 -0.69 1.71 2004-05 137.7719.1459.9192.4 -0.87 1.64 2005-06 136.2820.0140.4191.6 -0.99 1.60 2006-07 136.2820.0140.4191.6 -0.99 1.60 High 3462003-04 146.3014.6179.9200.7 -0.88 2.85 2004-05 146.1714.5090.2194.4 -0.62 1.44 2005-06 143.0416.2880.5191.6 -0.80 1.18 2006-07 143.0416.2880.5191.6 -0.80 1.18 Middle 4442003-04 138.0714.6481.8177.4 -0.26 0.71 2004-05 138.2615.3989.5173.7 -0.45 0.26 2005-06 136.8716.3761.6174.3 -0.68 1.45 2006-07 136.8716.3761.6174.3 -0.68 1.45 Percent of Student Computers with Content Software All Schools 22862003-04 50.5922.630.087.5 -0.03 -0.90

PAGE 415

Appendix C: Data Preparation Procedures (Continued) 401 Variable School Level N School Year Mean STD Min Max Skew Kurt 2004-05 54.2322.470.087.5 -0.21 -0.88 2005-06 52.1418.860.087.5 -0.05 -0.57 2006-07 52.6318.340.087.5 -0.07 -0.58 Elementary 14962003-04 53.0622.320.087.5 -0.18 -0.81 2004-05 56.6422.270.087.5 -0.36 -0.75 2005-06 54.3118.640.087.5 -0.22 -0.41 2006-07 55.1517.580.087.5 -0.25 -0.38 High 3462003-04 43.4621.591.887.5 0.42 -0.61 2004-05 46.8621.495.487.5 0.23 -0.96 2005-06 45.2018.095.487.5 0.44 -0.11 2006-07 44.4318.177.187.5 0.43 -0.31 Middle 4442003-04 47.8222.990.087.5 0.15 -0.91 2004-05 51.8722.460.087.5 -0.09 -0.80 2005-06 50.2218.697.187.5 0.18 -0.68 2006-07 50.5218.843.687.5 0.25 -0.54 Percent of Student Computers with Office/ Production Software All Schools 22862003-04 74.9016.800.087.5 -1.66 2.68 2004-05 76.6815.660.087.5 -1.78 3.15 2005-06 78.1414.960.087.5 -2.17 5.07 2006-07 80.0412.930.087.5 -2.36 6.95 Elementary 14962003-04 72.2318.160.087.5 -1.40 1.66 2004-05 74.5416.900.087.5 -1.48 1.79 2005-06 75.9216.680.087.5 -1.84 3.28 2006-07 78.1914.340.087.5 -2.05 5.10 High 3462003-04 80.9211.5510.087.5 -2.51 8.05 2004-05 81.5111.540.087.5 -3.04 12.83 2005-06 82.5910.3610.087.5 -3.50 16.37 2006-07 83.678.4312.587.5 -3.42 17.94 Middle 4442003-04 79.2012.9810.087.5 -2.09 5.25 2004-05 80.1412.350.087.5 -2.42 7.84 2005-06 82.159.1537.387.5 -2.00 4.09 2006-07 83.468.9127.287.5 -2.76 8.71 Percent of Student Computers with Advanced Production Software All Schools 22862003-04 22.1318.290.087.5 1.15 1.09 2004-05 21.8518.630.087.5 1.15 1.10 2005-06 25.9321.190.087.5 0.87 0.11 2006-07 28.0622.920.087.5 0.68 -0.50 Elementary 14962003-04 21.9818.480.087.5 1.07 0.87 2004-05 21.2518.840.087.5 1.08 0.82 2005-06 25.4621.210.087.5 0.79 -0.08 2006-07 26.9823.120.087.5 0.67 -0.58 High 3462003-04 22.7716.920.087.5 1.43 1.94 2004-05 22.8816.800.087.5 1.43 2.22 2005-06 27.4320.480.087.5 1.15 0.74 2006-07 31.0521.670.087.5 0.87 -0.03 Middle 4442003-04 22.1218.680.087.5 1.29 1.34 2004-05 23.0819.240.087.5 1.27 1.35

PAGE 416

Appendix C: Data Preparation Procedures (Continued) 402 Variable School Level N School Year Mean STD Min Max Skew Kurt 2005-06 26.3221.650.087.5 0.99 0.27 2006-07 29.3822.970.087.5 0.67 -0.54 Percent of Teachers Who Regularly Use Technology to Delivery Instruction All Schools 22862003-04 20.0012.140.087.5 1.10 1.71 2004-05 22.2912.650.087.5 0.97 1.26 2005-06 23.8413.740.087.5 0.82 0.70 2006-07 27.7514.160.087.5 0.58 0.28 Elementary 14962003-04 18.2311.660.087.5 1.26 2.50 2004-05 20.2812.290.087.5 1.17 2.16 2005-06 21.4413.260.087.5 1.00 1.23 2006-07 25.7314.280.087.5 0.76 0.59 High 3462003-04 24.5811.455.068.0 0.76 0.54 2004-05 26.9312.325.068.0 0.86 0.70 2005-06 28.2812.725.067.8 0.52 -0.23 2006-07 31.9212.910.077.4 0.49 0.46 Middle 4442003-04 22.4112.892.578.1 1.08 1.36 2004-05 25.4312.510.070.0 0.73 0.26 2005-06 28.5014.052.587.5 0.71 0.77 2006-07 31.3013.265.078.1 0.35 0.02 Percent of Teachers Who Regularly Use Technology for Administrative Purposes All Schools 22862003-04 61.6317.040.087.5 -0.74 0.28 2004-05 65.7515.665.087.5 -0.86 0.55 2005-06 68.5814.730.087.5 -1.07 1.31 2006-07 72.1812.690.087.5 -1.04 1.58 Elementary 14962003-04 59.2617.400.087.5 -0.67 0.11 2004-05 63.4715.965.087.5 -0.79 0.38 2005-06 66.6715.260.087.5 -1.06 1.22 2006-07 70.7013.060.087.5 -0.98 1.54 High 3462003-04 66.3115.755.087.5 -0.91 0.78 2004-05 69.7813.9712.587.5 -1.10 1.28 2005-06 71.7712.9727.787.5 -0.97 0.72 2006-07 74.9011.5617.587.5 -1.26 2.27 Middle 4442003-04 65.9815.1214.987.5 -0.85 0.63 2004-05 70.3114.1812.587.5 -0.93 0.80 2005-06 72.5412.9017.487.5 -1.01 1.04 2006-07 75.0511.3922.887.5 -1.04 1.03 Frequency Students Use Content Delivery Software All Schools 22862003-04 5.931.620.08.0 -0.65 0.02 2004-05 5.121.810.08.0 -0.32 -0.47 2005-06 5.431.950.08.0 -0.54 -0.37 2006-07 5.532.000.08.0 -0.66 -0.23 Elementary 14962003-04 6.161.590.08.0 -0.84 0.37 2004-05 5.431.670.08.0 -0.41 -0.23 2005-06 5.571.870.08.0 -0.62 -0.14 2006-07 5.651.910.08.0 -0.70 -0.08 High 3462003-04 5.331.631.08.0 -0.33 -0.19 2004-05 4.612.030.08.0 -0.08 -0.81

PAGE 417

Appendix C: Data Preparation Procedures (Continued) 403 Variable School Level N School Year Mean STD Min Max Skew Kurt 2005-06 5.262.140.08.0 -0.47 -0.65 2006-07 5.372.230.08.0 -0.61 -0.53 Middle 4442003-04 5.611.560.08.0 -0.43 -0.07 2004-05 4.461.830.08.0 0.00 -0.63 2005-06 5.071.990.08.0 -0.29 -0.73 2006-07 5.282.080.08.0 -0.51 -0.51 Frequency Students Use Production Tool Software All Schools 22862003-04 6.482.090.012.0 -0.17 -0.06 2004-05 4.522.320.012.0 0.46 -0.02 2005-06 5.142.770.012.0 0.33 -0.56 2006-07 5.232.810.012.0 0.31 -0.60 Elementary 14962003-04 6.142.140.012.0 -0.07 -0.09 2004-05 4.012.120.011.0 0.47 0.05 2005-06 4.382.480.012.0 0.53 -0.09 2006-07 4.402.520.012.0 0.53 -0.13 High 3462003-04 7.521.633.012.0 -0.04 0.22 2004-05 6.032.340.012.0 0.17 -0.21 2005-06 7.322.391.012.0 -0.39 -0.24 2006-07 7.552.420.012.0 -0.31 -0.18 Middle 4442003-04 6.811.931.012.0 -0.09 -0.26 2004-05 5.052.300.012.0 0.46 -0.03 2005-06 6.042.770.012.0 0.17 -0.54 2006-07 6.212.700.012.0 0.09 -0.67 Level of Human Tech Support All Schools 22862003-04 6.472.770.012.0 0.06 -0.71 2004-05 6.592.850.012.0 0.04 -0.91 2005-06 7.132.772.014.0 0.03 -1.10 2006-07 7.912.591.013.0 0.00 -0.97 Elementary 14962003-04 6.292.710.012.0 0.09 -0.66 2004-05 6.262.800.012.0 0.15 -0.86 2005-06 6.862.782.014.0 0.17 -1.09 2006-07 7.522.541.013.0 0.15 -0.91 High 3462003-04 6.862.801.012.0 0.11 -0.79 2004-05 7.452.831.012.0 -0.19 -0.84 2005-06 7.782.673.013.0 -0.22 -0.88 2006-07 8.702.434.013.0 -0.29 -0.74 Middle 4442003-04 6.752.921.012.0 -0.10 -0.78 2004-05 7.022.860.012.0 -0.14 -0.88 2005-06 7.552.712.014.0 -0.19 -1.00 2006-07 8.572.621.013.0 -0.28 -0.86 Level of Hardware/ Internet Dependability All Schools 22862003-04 6.011.150.08.0 -1.32 3.19 2004-05 6.251.170.08.0 -1.43 2.70 2005-06 6.131.470.08.0 -1.24 1.10 2006-07 6.361.360.08.0 -1.43 1.94 Elementary 14962003-04 5.981.180.08.0 -1.29 2.99

PAGE 418

Appendix C: Data Preparation Procedures (Continued) 404 Variable School Level N School Year Mean STD Min Max Skew Kurt 2004-05 6.211.160.08.0 -1.31 2.14 2005-06 6.041.510.08.0 -1.16 0.81 2006-07 6.321.400.08.0 -1.44 1.91 High 3462003-04 6.150.971.08.0 -1.16 3.22 2004-05 6.321.171.08.0 -1.45 2.52 2005-06 6.271.351.08.0 -1.45 2.08 2006-07 6.411.291.08.0 -1.28 1.46 Middle 4442003-04 6.001.170.08.0 -1.41 3.36 2004-05 6.321.190.08.0 -1.84 4.89 2005-06 6.331.380.08.0 -1.36 1.62 2006-07 6.431.281.08.0 -1.44 2.15 Table C 22. Descriptive Statistics of Predictor Variables for FCAT Math Outcome Variable School Level N School Year Mean STD Min Max Skew Kurt Percent Students on Free or Reduced Lunch Program All Years 2300 2003-04 52.21 25.34 1.0 100.0 0.10 -0.94 2004-05 52.62 24.09 0.9 100.0 0.02 -0.90 2005-06 52.25 23.90 1.7 100.0 -0.04 -0.97 2006-07 52.25 23.90 1.7 100.0 -0.04 -0.97 Elementary 1511 2003-04 57.05 26.21 1.0 100.0 -0.14 -1.01 2004-05 56.52 24.86 0.9 100.0 -0.21 -0.93 2005-06 56.09 24.75 1.7 100.0 -0.25 -0.99 2006-07 56.09 24.75 1.7 100.0 -0.25 -0.99 High 345 2003-04 35.48 17.18 1.8 93.3 0.46 0.18 2004-05 38.83 18.05 3.8 100.0 0.58 0.63 2005-06 38.65 17.21 2.5 93.7 0.17 -0.34 2006-07 38.65 17.21 2.5 93.7 0.17 -0.34 Middle 444 2003-04 48.74 21.24 3.7 100.0 0.12 -0.65 2004-05 50.10 21.09 3.3 100.0 0.05 -0.60 2005-06 49.76 21.11 3.9 100.0 0.02 -0.79 2006-07 49.76 21.11 3.9 100.0 0.02 -0.79 Percent Minority Students All Years 2301 2003-04 50.26 28.22 0.0 100.0 0.30 -1.11 2004-05 51.45 28.39 0.0 100.0 0.27 -1.14 2005-06 52.48 28.40 0.0 100.0 0.23 -1.17 2006-07 52.48 28.40 0.0 100.0 0.23 -1.17 Elementary 1511 2003-04 52.03 28.98 0.0 100.0 0.20 -1.22 2004-05 53.19 29.13 0.0 100.0 0.18 -1.24 2005-06 54.42 28.97 0.0 100.0 0.14 -1.25 2006-07 54.42 28.97 0.0 100.0 0.14 -1.25 High 346 2003-04 44.66 26.01 2.6 100.0 0.60 -0.63 2004-05 45.59 26.47 0.7 99.9 0.55 -0.75

PAGE 419

Appendix C: Data Preparation Procedures (Continued) 405 Variable School Level N School Year Mean STD Min Max Skew Kurt 2005-06 45.94 26.99 1.0 99.9 0.50 -0.82 2006-07 45.94 26.99 1.0 99.9 0.50 -0.82 Middle 444 2003-04 48.59 26.58 4.8 100.0 0.41 -0.89 2004-05 50.08 26.62 4.6 99.8 0.36 -0.96 2005-06 50.97 26.68 5.6 99.9 0.32 -1.00 2006-07 50.97 26.68 5.6 99.9 0.32 -1.00 Percent LEP students All Years 2135 2003-04 8.77 10.71 0.0 63.3 1.96 3.92 2004-05 8.55 10.57 0.0 61.6 2.01 4.14 2005-06 8.83 10.65 0.0 65.7 1.97 4.09 2006-07 8.83 10.65 0.0 65.7 1.97 4.09 Elementary 1389 2003-04 10.84 12.24 0.1 63.3 1.55 1.94 2004-05 10.53 12.08 0.0 61.6 1.60 2.12 2005-06 10.84 12.13 0.1 65.7 1.57 2.14 2006-07 10.84 12.13 0.1 65.7 1.57 2.14 High 325 2003-04 4.50 4.78 0.0 23.1 1.48 1.75 2004-05 4.35 4.66 0.0 26.0 1.54 2.11 2005-06 4.43 4.67 0.0 28.3 1.61 2.75 2006-07 4.43 4.67 0.0 28.3 1.61 2.75 Middle 421 2003-04 5.23 5.45 0.1 32.5 1.69 3.50 2004-05 5.18 5.44 0.1 36.4 1.89 4.87 2005-06 5.47 5.71 0.1 37.4 2.04 5.74 2006-07 5.47 5.71 0.1 37.4 2.04 5.74 Percent Students with Disabilities All Years 2300 2003-04 15.51 5.41 0.5 40.6 0.68 1.42 2004-05 15.30 5.32 0.4 43.9 0.73 1.99 2005-06 15.30 5.44 0.3 72.6 1.51 9.72 2006-07 15.30 5.44 0.3 72.6 1.51 9.72 Elementary 1511 2003-04 16.05 5.71 1.2 40.6 0.81 1.27 2004-05 15.80 5.60 1.6 43.9 0.91 1.95 2005-06 16.03 5.77 1.9 72.6 1.79 10.58 2006-07 16.03 5.77 1.9 72.6 1.79 10.58 High 345 2003-04 13.26 4.43 0.7 31.1 0.03 0.96 2004-05 13.20 4.47 0.4 27.6 -0.16 0.42 2005-06 13.16 4.48 0.3 30.9 -0.04 0.66 2006-07 13.16 4.48 0.3 30.9 -0.04 0.66 Middle 444 2003-04 15.44 4.57 0.5 27.9 -0.13 0.17 2004-05 15.23 4.48 0.5 28.7 -0.16 0.20 2005-06 14.51 4.29 0.8 28.7 -0.07 0.26 2006-07 14.51 4.29 0.8 28.7 -0.07 0.26 Percent Gifted students All Years 1817 2003-04 4.98 5.77 0.1 52.3 2.91 12.23 2004-05 4.99 5.87 0.1 54.9 2.96 12.64 2005-06 4.90 5.79 0.0 57.2 3.03 13.87 2006-07 4.90 5.79 0.0 57.2 3.03 13.87 Elementary 1366 2003-04 4.31 5.42 0.1 52.3 3.40 17.24

PAGE 420

Appendix C: Data Preparation Procedures (Continued) 406 Variable School Level N School Year Mean STD Min Max Skew Kurt 2004-05 4.28 5.49 0.1 54.9 3.50 18.23 2005-06 4.19 5.56 0.1 57.2 3.69 20.47 2006-07 4.19 5.56 0.1 57.2 3.69 20.47 High 21 2003-04 7.04 10.17 0.4 39.8 2.49 5.99 2004-05 7.66 10.64 0.7 39.6 2.14 4.13 2005-06 4.70 5.06 0.0 37.4 2.69 11.04 2006-07 4.70 5.06 0.0 37.4 2.69 11.04 Middle 430 2003-04 7.00 6.06 0.1 37.3 2.04 5.18 2004-05 7.09 6.18 0.1 39.1 2.10 5.59 2005-06 7.27 6.34 0.1 39.5 2.11 5.84 2006-07 7.27 6.34 0.1 39.5 2.11 5.84 Positive Learning Environment All Years 2301 2003-04 363.90 28.42 131.4 398.7 -1.56 3.61 2004-05 364.23 26.60 220.3 398.1 -1.39 1.94 2005-06 364.82 25.98 221.3 397.8 -1.39 1.98 2006-07 364.82 25.98 221.3 397.8 -1.39 1.98 Elementary 1511 2003-04 377.49 16.64 241.5 398.7 -2.95 13.86 2004-05 376.93 14.83 272.9 397.9 -2.22 8.41 2005-06 376.59 15.67 252.8 397.8 -2.33 8.97 2006-07 376.59 15.67 252.8 397.8 -2.33 8.97 High 346 2003-04 336.05 26.61 232.4 396.0 -0.53 0.89 2004-05 338.32 26.55 259.0 397.2 -0.49 0.26 2005-06 340.26 24.88 261.7 396.6 -0.23 0.12 2006-07 340.26 24.88 261.7 396.6 -0.23 0.12 Middle 444 2003-04 339.38 29.33 131.4 394.9 -1.52 6.16 2004-05 341.22 27.76 220.3 398.1 -0.83 1.21 2005-06 343.90 28.34 221.3 397.8 -0.94 1.19 2006-07 343.90 28.34 221.3 397.8 -0.94 1.19 Positive Teacher Qualifications All Years 2301 2003-04 139.97 16.83 61.3 200.7 -0.66 1.67 2004-05 139.06 18.12 59.9 194.4 -0.85 1.68 2005-06 137.38 19.00 40.4 191.6 -0.96 1.71 2006-07 137.38 19.00 40.4 191.6 -0.96 1.71 Elementary 1511 2003-04 139.08 17.57 61.3 187.9 -0.68 1.68 2004-05 137.67 19.21 59.9 192.4 -0.87 1.62 2005-06 136.23 20.05 40.4 191.6 -0.99 1.58 2006-07 136.23 20.05 40.4 191.6 -0.99 1.58 High 346 2003-04 146.30 14.61 79.9 200.7 -0.88 2.85 2004-05 146.17 14.50 90.2 194.4 -0.62 1.44 2005-06 143.04 16.28 80.5 191.6 -0.80 1.18 2006-07 143.04 16.28 80.5 191.6 -0.80 1.18 Middle 444 2003-04 138.07 14.64 81.8 177.4 -0.26 0.71 2004-05 138.26 15.39 89.5 173.7 -0.45 0.26 2005-06 136.87 16.37 61.6 174.3 -0.68 1.45 2006-07 136.87 16.37 61.6 174.3 -0.68 1.45

PAGE 421

Appendix C: Data Preparation Procedures (Continued) 407 Variable School Level N School Year Mean STD Min Max Skew Kurt Percent of Student Computers with Content Software All Years 2301 2003-04 50.58 22.63 0.0 87.5 -0.03 -0.90 2004-05 54.25 22.48 0.0 87.5 -0.21 -0.89 2005-06 52.17 18.86 0.0 87.5 -0.05 -0.57 2006-07 52.71 18.36 0.0 87.5 -0.07 -0.59 Elementary 1511 2003-04 53.03 22.32 0.0 87.5 -0.18 -0.80 2004-05 56.65 22.27 0.0 87.5 -0.36 -0.75 2005-06 54.35 18.65 0.0 87.5 -0.22 -0.41 2006-07 55.26 17.61 0.0 87.5 -0.26 -0.39 High 346 2003-04 43.46 21.59 1.8 87.5 0.42 -0.61 2004-05 46.86 21.49 5.4 87.5 0.23 -0.96 2005-06 45.20 18.09 5.4 87.5 0.44 -0.11 2006-07 44.43 18.17 7.1 87.5 0.43 -0.31 Middle 444 2003-04 47.82 22.99 0.0 87.5 0.15 -0.91 2004-05 51.87 22.46 0.0 87.5 -0.09 -0.80 2005-06 50.22 18.69 7.1 87.5 0.18 -0.68 2006-07 50.52 18.84 3.6 87.5 0.25 -0.54 Percent of Student Computers with Office/ Production Software All Years 2301 2003-04 74.86 16.83 0.0 87.5 -1.65 2.65 2004-05 76.68 15.66 0.0 87.5 -1.78 3.16 2005-06 78.11 14.98 0.0 87.5 -2.16 5.00 2006-07 80.07 12.90 0.0 87.5 -2.36 6.99 Elementary 1511 2003-04 72.20 18.18 0.0 87.5 -1.40 1.64 2004-05 74.56 16.88 0.0 87.5 -1.49 1.81 2005-06 75.90 16.69 0.0 87.5 -1.83 3.24 2006-07 78.25 14.29 0.0 87.5 -2.06 5.15 High 346 2003-04 80.92 11.55 10.0 87.5 -2.51 8.05 2004-05 81.51 11.54 0.0 87.5 -3.04 12.83 2005-06 82.59 10.36 10.0 87.5 -3.50 16.37 2006-07 83.67 8.43 12.5 87.5 -3.42 17.94 Middle 444 2003-04 79.20 12.98 10.0 87.5 -2.09 5.25 2004-05 80.14 12.35 0.0 87.5 -2.42 7.84 2005-06 82.15 9.15 37.3 87.5 -2.00 4.09 2006-07 83.46 8.91 27.2 87.5 -2.76 8.71 Percent of Student Computers with Advanced Production Software All Years 2301 2003-04 22.15 18.30 0.0 87.5 1.15 1.06 2004-05 21.87 18.68 0.0 87.5 1.14 1.06 2005-06 25.96 21.26 0.0 87.5 0.87 0.10 2006-07 28.12 22.95 0.0 87.5 0.67 -0.51 Elementary 1511 2003-04 22.01 18.50 0.0 87.5 1.06 0.83 2004-05 21.29 18.90 0.0 87.5 1.07 0.77 2005-06 25.52 21.31 0.0 87.5 0.78 -0.09 2006-07 27.07 23.17 0.0 87.5 0.66 -0.60 High 346 2003-04 22.77 16.92 0.0 87.5 1.43 1.94 2004-05 22.88 16.80 0.0 87.5 1.43 2.22 2005-06 27.43 20.48 0.0 87.5 1.15 0.74

PAGE 422

Appendix C: Data Preparation Procedures (Continued) 408 Variable School Level N School Year Mean STD Min Max Skew Kurt 2006-07 31.05 21.67 0.0 87.5 0.87 -0.03 Middle 444 2003-04 22.12 18.68 0.0 87.5 1.29 1.34 2004-05 23.08 19.24 0.0 87.5 1.27 1.35 2005-06 26.32 21.65 0.0 87.5 0.99 0.27 2006-07 29.38 22.97 0.0 87.5 0.67 -0.54 Percent of Teachers Who Regularly Use Technology to Delivery Instruction All Years 2301 2003-04 19.99 12.15 0.0 87.5 1.10 1.70 2004-05 22.27 12.65 0.0 87.5 0.97 1.25 2005-06 23.83 13.74 0.0 87.5 0.81 0.69 2006-07 27.72 14.18 0.0 87.5 0.58 0.27 Elementary 1511 2003-04 18.23 11.68 0.0 87.5 1.25 2.47 2004-05 20.27 12.30 0.0 87.5 1.17 2.13 2005-06 21.44 13.26 0.0 87.5 0.99 1.20 2006-07 25.71 14.31 0.0 87.5 0.75 0.56 High 346 2003-04 24.58 11.45 5.0 68.0 0.76 0.54 2004-05 26.93 12.32 5.0 68.0 0.86 0.70 2005-06 28.28 12.72 5.0 67.8 0.52 -0.23 2006-07 31.92 12.91 0.0 77.4 0.49 0.46 Middle 444 2003-04 22.41 12.89 2.5 78.1 1.08 1.36 2004-05 25.43 12.51 0.0 70.0 0.73 0.26 2005-06 28.50 14.05 2.5 87.5 0.71 0.77 2006-07 31.30 13.26 5.0 78.1 0.35 0.02 Percent of Teachers Who Regularly Use Technology for Administrative Purposes All Years 2301 2003-04 61.60 17.07 0.0 87.5 -0.75 0.28 2004-05 65.75 15.66 5.0 87.5 -0.86 0.54 2005-06 68.55 14.73 0.0 87.5 -1.07 1.30 2006-07 72.17 12.70 0.0 87.5 -1.04 1.59 Elementary 1511 2003-04 59.24 17.43 0.0 87.5 -0.67 0.12 2004-05 63.48 15.96 5.0 87.5 -0.79 0.38 2005-06 66.64 15.25 0.0 87.5 -1.05 1.21 2006-07 70.69 13.06 0.0 87.5 -0.98 1.54 High 346 2003-04 66.31 15.75 5.0 87.5 -0.91 0.78 2004-05 69.78 13.97 12.5 87.5 -1.10 1.28 2005-06 71.77 12.97 27.7 87.5 -0.97 0.72 2006-07 74.90 11.56 17.5 87.5 -1.26 2.27 Middle 444 2003-04 65.98 15.12 14.9 87.5 -0.85 0.63 2004-05 70.31 14.18 12.5 87.5 -0.93 0.80 2005-06 72.54 12.90 17.4 87.5 -1.01 1.04 2006-07 75.05 11.39 22.8 87.5 -1.04 1.03 Frequency Students Use Content Delivery Software All Years 2301 2003-04 5.93 1.62 0.0 8.0 -0.65 0.01 2004-05 5.12 1.81 0.0 8.0 -0.33 -0.47 2005-06 5.43 1.95 0.0 8.0 -0.54 -0.37 2006-07 5.53 2.00 0.0 8.0 -0.66 -0.23 Elementary 1511 2003-04 6.15 1.59 0.0 8.0 -0.83 0.36

PAGE 423

Appendix C: Data Preparation Procedures (Continued) 409 Variable School Level N School Year Mean STD Min Max Skew Kurt 2004-05 5.44 1.67 0.0 8.0 -0.41 -0.22 2005-06 5.57 1.87 0.0 8.0 -0.62 -0.14 2006-07 5.65 1.92 0.0 8.0 -0.70 -0.07 High 346 2003-04 5.33 1.63 1.0 8.0 -0.33 -0.19 2004-05 4.61 2.03 0.0 8.0 -0.08 -0.81 2005-06 5.26 2.14 0.0 8.0 -0.47 -0.65 2006-07 5.37 2.23 0.0 8.0 -0.61 -0.53 Middle 444 2003-04 5.61 1.56 0.0 8.0 -0.43 -0.07 2004-05 4.46 1.83 0.0 8.0 0.00 -0.63 2005-06 5.07 1.99 0.0 8.0 -0.29 -0.73 2006-07 5.28 2.08 0.0 8.0 -0.51 -0.51 Frequency Students Use Production Tool Software All Years 2301 2003-04 6.47 2.10 0.0 12.0 -0.18 -0.05 2004-05 4.52 2.32 0.0 12.0 0.46 -0.03 2005-06 5.13 2.76 0.0 12.0 0.33 -0.56 2006-07 5.21 2.81 0.0 12.0 0.31 -0.60 Elementary 1511 2003-04 6.13 2.14 0.0 12.0 -0.09 -0.08 2004-05 4.01 2.13 0.0 11.0 0.46 0.02 2005-06 4.37 2.48 0.0 12.0 0.54 -0.09 2006-07 4.39 2.52 0.0 12.0 0.53 -0.12 High 346 2003-04 7.52 1.63 3.0 12.0 -0.04 0.22 2004-05 6.03 2.34 0.0 12.0 0.17 -0.21 2005-06 7.32 2.39 1.0 12.0 -0.39 -0.24 2006-07 7.55 2.42 0.0 12.0 -0.31 -0.18 Middle 444 2003-04 6.81 1.93 1.0 12.0 -0.09 -0.26 2004-05 5.05 2.30 0.0 12.0 0.46 -0.03 2005-06 6.04 2.77 0.0 12.0 0.17 -0.54 2006-07 6.21 2.70 0.0 12.0 0.09 -0.67 Level of Human Tech Support All Years 2301 2003-04 6.45 2.77 0.0 12.0 0.07 -0.71 2004-05 6.58 2.85 0.0 12.0 0.05 -0.91 2005-06 7.12 2.77 2.0 14.0 0.04 -1.10 2006-07 7.90 2.59 1.0 13.0 0.01 -0.97 Elementary 1511 2003-04 6.27 2.70 0.0 12.0 0.10 -0.66 2004-05 6.25 2.80 0.0 12.0 0.15 -0.85 2005-06 6.85 2.77 2.0 14.0 0.17 -1.09 2006-07 7.51 2.54 1.0 13.0 0.15 -0.91 High 346 2003-04 6.86 2.80 1.0 12.0 0.11 -0.79 2004-05 7.45 2.83 1.0 12.0 -0.19 -0.84 2005-06 7.78 2.67 3.0 13.0 -0.22 -0.88 2006-07 8.70 2.43 4.0 13.0 -0.29 -0.74 Middle 444 2003-04 6.75 2.92 1.0 12.0 -0.10 -0.78 2004-05 7.02 2.86 0.0 12.0 -0.14 -0.88 2005-06 7.55 2.71 2.0 14.0 -0.19 -1.00 2006-07 8.57 2.62 1.0 13.0 -0.28 -0.86

PAGE 424

Appendix C: Data Preparation Procedures (Continued) 410 Variable School Level N School Year Mean STD Min Max Skew Kurt Level of Hardware/ Internet Dependability All Years 2301 2003-04 6.01 1.15 0.0 8.0 -1.32 3.17 2004-05 6.25 1.17 0.0 8.0 -1.43 2.72 2005-06 6.13 1.47 0.0 8.0 -1.23 1.08 2006-07 6.36 1.36 0.0 8.0 -1.43 1.92 Elementary 1511 2003-04 5.99 1.18 0.0 8.0 -1.28 2.96 2004-05 6.22 1.16 0.0 8.0 -1.31 2.16 2005-06 6.04 1.51 0.0 8.0 -1.15 0.79 2006-07 6.32 1.40 0.0 8.0 -1.44 1.88 High 346 2003-04 6.15 0.97 1.0 8.0 -1.16 3.22 2004-05 6.32 1.17 1.0 8.0 -1.45 2.52 2005-06 6.27 1.35 1.0 8.0 -1.45 2.08 2006-07 6.41 1.29 1.0 8.0 -1.28 1.46 Middle 444 2003-04 6.00 1.17 0.0 8.0 -1.41 3.36 2004-05 6.32 1.19 0.0 8.0 -1.84 4.89 2005-06 6.33 1.38 0.0 8.0 -1.36 1.62 2006-07 6.43 1.28 1.0 8.0 -1.44 2.15 Table C 23. Descriptive Statistics of Predictor Variables for FCAT Writing Outcome Variable N School Level School Year Mean STD Min Max Skew Kurt Percent Students on Free or Reduced Lunch Program 2263 All Years 2003-04 52.08 25.34 1.0 100.0 0.11 -0.95 2004-05 52.54 24.13 0.9 100.0 0.02 -0.91 2005-06 52.17 23.95 1.7 100.0 -0.03 -0.98 2006-07 52.17 23.95 1.7 100.0 -0.03 -0.98 1480 Elementary 2003-04 56.99 26.19 1.0 100.0 -0.14 -1.01 2004-05 56.51 24.89 0.9 100.0 -0.20 -0.93 2005-06 56.07 24.79 1.7 100.0 -0.25 -0.99 2006-07 56.07 24.79 1.7 100.0 -0.25 -0.99 346 High 2003-04 35.47 17.21 1.8 93.3 0.45 0.16 2004-05 38.82 18.07 3.8 100.0 0.58 0.61 2005-06 38.63 17.23 2.5 93.7 0.18 -0.35 2006-07 38.63 17.23 2.5 93.7 0.18 -0.35 437 Middle 2003-04 48.58 21.40 3.7 100.0 0.12 -0.69 2004-05 50.00 21.24 3.3 100.0 0.04 -0.64 2005-06 49.69 21.28 3.9 100.0 0.02 -0.82 2006-07 49.69 21.28 3.9 100.0 0.02 -0.82 Percent Minority Students 2264 All Years 2003-04 50.53 28.13 1.4 100.0 0.30 -1.11 2004-05 51.71 28.27 0.7 100.0 0.27 -1.14 2005-06 52.77 28.29 1.0 100.0 0.22 -1.17 2006-07 52.77 28.29 1.0 100.0 0.22 -1.17

PAGE 425

Appendix C: Data Preparation Procedures (Continued) 411 Variable N School Level School Year Mean STD Min Max Skew Kurt 1480 Elementary 2003-04 52.40 28.86 1.4 100.0 0.20 -1.22 2004-05 53.57 28.98 1.6 100.0 0.18 -1.24 2005-06 54.80 28.82 1.0 100.0 0.13 -1.25 2006-07 54.80 28.82 1.0 100.0 0.13 -1.25 347 High 2003-04 44.64 26.06 2.6 100.0 0.59 -0.64 2004-05 45.56 26.51 0.7 99.9 0.54 -0.76 2005-06 45.93 27.04 1.0 99.9 0.49 -0.84 2006-07 45.93 27.04 1.0 99.9 0.49 -0.84 437 Middle 2003-04 48.89 26.45 4.8 100.0 0.41 -0.89 2004-05 50.32 26.38 5.6 99.8 0.35 -0.95 2005-06 51.32 26.51 5.6 99.9 0.32 -0.99 2006-07 51.32 26.51 5.6 99.9 0.32 -0.99 Percent LEP students 2109 All Years 2003-04 8.81 10.71 0.0 63.3 1.96 3.90 2004-05 8.59 10.58 0.0 61.6 2.00 4.12 2005-06 8.86 10.65 0.0 65.7 1.97 4.07 2006-07 8.86 10.65 0.0 65.7 1.97 4.07 1367 Elementary 2003-04 10.92 12.24 0.1 63.3 1.54 1.92 2004-05 10.60 12.09 0.0 61.6 1.59 2.10 2005-06 10.92 12.13 0.1 65.7 1.56 2.12 2006-07 10.92 12.13 0.1 65.7 1.56 2.12 326 High 2003-04 4.49 4.78 0.0 23.1 1.47 1.75 2004-05 4.35 4.66 0.0 26.0 1.53 2.09 2005-06 4.43 4.67 0.0 28.3 1.60 2.73 2006-07 4.43 4.67 0.0 28.3 1.60 2.73 416 Middle 2003-04 5.25 5.46 0.1 32.5 1.70 3.51 2004-05 5.20 5.45 0.1 36.4 1.89 4.84 2005-06 5.49 5.72 0.1 37.4 2.04 5.70 2006-07 5.49 5.72 0.1 37.4 2.04 5.70 Percent Students with Disabilities 2263 All Years 2003-04 15.48 5.42 0.5 40.6 0.69 1.44 2004-05 15.28 5.32 0.4 43.9 0.73 1.99 2005-06 15.28 5.44 0.3 72.6 1.52 9.84 2006-07 15.28 5.44 0.3 72.6 1.52 9.84 1480 Elementary 2003-04 16.02 5.72 1.2 40.6 0.83 1.30 2004-05 15.78 5.61 1.6 43.9 0.92 1.97 2005-06 16.02 5.77 1.9 72.6 1.81 10.73 2006-07 16.02 5.77 1.9 72.6 1.81 10.73 346 High 2003-04 13.26 4.45 0.7 31.1 0.04 0.91 2004-05 13.19 4.48 0.4 27.6 -0.15 0.39 2005-06 13.14 4.47 0.3 30.9 -0.04 0.68 2006-07 13.14 4.47 0.3 30.9 -0.04 0.68 437 Middle 2003-04 15.45 4.60 0.5 27.9 -0.14 0.13 2004-05 15.24 4.50 0.5 28.7 -0.17 0.19 2005-06 14.51 4.31 0.8 28.7 -0.07 0.25 2006-07 14.51 4.31 0.8 28.7 -0.07 0.25

PAGE 426

Appendix C: Data Preparation Procedures (Continued) 412 Variable N School Level School Year Mean STD Min Max Skew Kurt Percent Gifted students 1790 All Years 2003-04 4.97 5.75 0.1 52.3 2.94 12.51 2004-05 4.99 5.86 0.1 54.9 2.98 12.88 2005-06 4.90 5.79 0.0 57.2 3.06 14.12 2006-07 4.90 5.79 0.0 57.2 3.06 14.12 1345 Elementary 2003-04 4.29 5.40 0.1 52.3 3.45 17.74 2004-05 4.27 5.47 0.1 54.9 3.55 18.79 2005-06 4.18 5.55 0.1 57.2 3.73 20.95 2006-07 4.18 5.55 0.1 57.2 3.73 20.95 22 High 2003-04 7.04 9.92 0.4 39.8 2.54 6.35 2004-05 7.60 10.36 0.7 39.6 2.20 4.50 2005-06 4.67 4.99 0.0 37.4 2.75 11.70 2006-07 4.67 4.99 0.0 37.4 2.75 11.70 423 Middle 2003-04 7.01 6.04 0.1 37.3 2.04 5.26 2004-05 7.13 6.19 0.1 39.1 2.09 5.58 2005-06 7.30 6.36 0.1 39.5 2.10 5.77 2006-07 7.30 6.36 0.1 39.5 2.10 5.77 Positive Learning Environment 2264 All Years 2003-04 363.70 28.56 131.4 398.7 -1.55 3.55 2004-05 364.07 26.68 220.3 398.1 -1.38 1.92 2005-06 364.70 26.10 221.3 397.8 -1.39 1.99 2006-07 364.70 26.10 221.3 397.8 -1.39 1.99 1480 Elementary 2003-04 377.43 16.75 241.5 398.7 -2.94 13.71 2004-05 376.86 14.85 272.9 397.9 -2.22 8.45 2005-06 376.57 15.70 252.8 397.8 -2.33 9.01 2006-07 376.57 15.70 252.8 397.8 -2.33 9.01 347 High 2003-04 336.01 26.59 232.4 396.0 -0.52 0.88 2004-05 338.23 26.52 259.0 397.2 -0.48 0.26 2005-06 340.30 24.93 261.7 396.6 -0.22 0.10 2006-07 340.30 24.93 261.7 396.6 -0.22 0.10 437 Middle 2003-04 339.21 29.47 131.4 394.9 -1.52 6.09 2004-05 341.26 27.93 220.3 398.1 -0.83 1.18 2005-06 343.86 28.64 221.3 397.8 -0.95 1.15 2006-07 343.86 28.64 221.3 397.8 -0.95 1.15 Positive Teacher Qualifications 2264 All Years 2003-04 140.03 16.85 61.3 200.7 -0.66 1.69 2004-05 139.13 18.06 59.9 194.4 -0.83 1.62 2005-06 137.52 18.90 40.4 191.6 -0.97 1.77 2006-07 137.52 18.90 40.4 191.6 -0.97 1.77 1480 Elementary 2003-04 139.11 17.60 61.3 187.9 -0.69 1.71 2004-05 137.75 19.13 59.9 192.4 -0.85 1.57 2005-06 136.38 19.90 40.4 191.6 -1.00 1.68 2006-07 136.38 19.90 40.4 191.6 -1.00 1.68 347 High 2003-04 146.39 14.61 79.9 200.7 -0.89 2.87 2004-05 146.26 14.44 90.2 194.4 -0.63 1.50

PAGE 427

Appendix C: Data Preparation Procedures (Continued) 413 Variable N School Level School Year Mean STD Min Max Skew Kurt 2005-06 143.20 16.24 80.5 191.6 -0.82 1.23 2006-07 143.20 16.24 80.5 191.6 -0.82 1.23 437 Middle 2003-04 138.12 14.66 81.8 177.4 -0.26 0.73 2004-05 138.15 15.43 89.5 173.7 -0.44 0.25 2005-06 136.87 16.46 61.6 174.3 -0.68 1.41 2006-07 136.87 16.46 61.6 174.3 -0.68 1.41 Percent of Student Computers with Content Software 2264 All Years 2003-04 50.58 22.67 0.0 87.5 -0.03 -0.91 2004-05 54.25 22.51 0.0 87.5 -0.21 -0.89 2005-06 52.16 18.86 0.0 87.5 -0.05 -0.57 2006-07 52.69 18.39 0.0 87.5 -0.07 -0.59 1480 Elementary 2003-04 53.02 22.43 0.0 87.5 -0.19 -0.82 2004-05 56.77 22.33 0.0 87.5 -0.37 -0.75 2005-06 54.42 18.70 0.0 87.5 -0.23 -0.42 2006-07 55.26 17.64 0.0 87.5 -0.26 -0.40 347 High 2003-04 43.63 21.43 3.6 87.5 0.44 -0.63 2004-05 46.79 21.42 5.4 87.5 0.23 -0.94 2005-06 45.07 17.99 5.4 87.5 0.45 -0.07 2006-07 44.53 18.13 7.1 87.5 0.42 -0.30 437 Middle 2003-04 47.85 23.00 0.0 87.5 0.15 -0.92 2004-05 51.65 22.39 0.0 87.5 -0.09 -0.79 2005-06 50.14 18.50 7.1 87.5 0.16 -0.66 2006-07 50.47 18.92 3.6 87.5 0.24 -0.56 Percent of Student Computers with Office/ Production Software 2264 All Years 2003-04 74.91 16.76 0.0 87.5 -1.66 2.70 2004-05 76.73 15.60 0.0 87.5 -1.79 3.21 2005-06 78.15 14.89 0.0 87.5 -2.16 5.08 2006-07 80.06 12.92 0.0 87.5 -2.35 6.88 1480 Elementary 2003-04 72.25 18.10 0.0 87.5 -1.41 1.69 2004-05 74.62 16.81 0.0 87.5 -1.49 1.85 2005-06 75.95 16.59 0.0 87.5 -1.85 3.32 2006-07 78.24 14.28 0.0 87.5 -2.04 5.09 347 High 2003-04 80.98 11.52 10.0 87.5 -2.53 8.16 2004-05 81.52 11.53 0.0 87.5 -3.04 12.87 2005-06 82.61 10.35 10.0 87.5 -3.51 16.42 2006-07 83.73 8.39 12.5 87.5 -3.47 18.36 437 Middle 2003-04 79.09 13.06 10.0 87.5 -2.07 5.12 2004-05 80.07 12.42 0.0 87.5 -2.41 7.72 2005-06 82.06 9.24 37.3 87.5 -1.97 3.91 2006-07 83.32 9.27 27.2 87.5 -2.81 9.04 Percent of Student Computers with Advanced Production Software 2264 All Years 2003-04 22.25 18.30 0.0 87.5 1.14 1.06 2004-05 21.96 18.60 0.0 87.5 1.14 1.07

PAGE 428

Appendix C: Data Preparation Procedures (Continued) 414 Variable N School Level School Year Mean STD Min Max Skew Kurt 2005-06 26.07 21.21 0.0 87.5 0.86 0.08 2006-07 28.24 22.96 0.0 87.5 0.67 -0.51 1480 Elementary 2003-04 22.15 18.52 0.0 87.5 1.05 0.83 2004-05 21.45 18.87 0.0 87.5 1.06 0.78 2005-06 25.67 21.25 0.0 87.5 0.77 -0.11 2006-07 27.18 23.17 0.0 87.5 0.65 -0.60 347 High 2003-04 23.03 17.07 0.0 87.5 1.41 1.82 2004-05 22.91 16.84 0.0 87.5 1.42 2.16 2005-06 27.40 20.33 0.0 87.5 1.13 0.68 2006-07 31.17 21.71 0.0 87.5 0.87 -0.06 437 Middle 2003-04 22.00 18.51 0.0 87.5 1.29 1.38 2004-05 22.91 18.99 0.0 87.5 1.28 1.43 2005-06 26.37 21.75 0.0 87.5 0.99 0.24 2006-07 29.52 22.98 0.0 87.5 0.66 -0.55 Percent of Teachers Who Regularly Use Technology to Delivery Instruction 2264 All Years 2003-04 19.98 12.14 0.0 87.5 1.11 1.75 2004-05 22.30 12.65 0.0 87.5 0.97 1.24 2005-06 23.92 13.79 0.0 87.5 0.83 0.75 2006-07 27.82 14.13 0.0 87.5 0.57 0.27 1480 Elementary 2003-04 18.22 11.68 0.0 87.5 1.27 2.52 2004-05 20.30 12.32 0.0 87.5 1.18 2.16 2005-06 21.51 13.29 0.0 87.5 0.99 1.20 2006-07 25.76 14.26 0.0 87.5 0.75 0.57 347 High 2003-04 24.58 11.43 5.0 68.0 0.76 0.55 2004-05 27.02 12.39 5.0 68.0 0.85 0.63 2005-06 28.29 12.76 5.0 67.8 0.53 -0.24 2006-07 31.98 12.90 0.0 77.4 0.48 0.46 437 Middle 2003-04 22.32 12.85 2.5 78.1 1.10 1.47 2004-05 25.32 12.41 0.0 70.0 0.71 0.19 2005-06 28.60 14.20 2.5 87.5 0.77 0.94 2006-07 31.49 13.13 5.0 78.1 0.35 0.06 Percent of Teachers Who Regularly Use Technology for Administrative Purposes 2264 All Years 2003-04 61.56 17.04 0.0 87.5 -0.74 0.29 2004-05 65.74 15.70 5.0 87.5 -0.86 0.53 2005-06 68.63 14.69 0.0 87.5 -1.07 1.29 2006-07 72.23 12.67 0.0 87.5 -1.06 1.66 1480 Elementary 2003-04 59.25 17.38 0.0 87.5 -0.67 0.14 2004-05 63.46 16.01 5.0 87.5 -0.79 0.37 2005-06 66.70 15.25 0.0 87.5 -1.04 1.16 2006-07 70.73 13.07 0.0 87.5 -1.00 1.59 347 High 2003-04 66.21 15.76 5.0 87.5 -0.90 0.75 2004-05 69.84 13.99 12.5 87.5 -1.10 1.27 2005-06 71.80 12.96 27.7 87.5 -0.97 0.73 2006-07 74.91 11.56 17.5 87.5 -1.25 2.26

PAGE 429

Appendix C: Data Preparation Procedures (Continued) 415 Variable N School Level School Year Mean STD Min Max Skew Kurt 437 Middle 2003-04 65.71 15.25 14.9 87.5 -0.84 0.58 2004-05 70.22 14.28 12.5 87.5 -0.94 0.78 2005-06 72.64 12.70 17.4 87.5 -1.02 1.12 2006-07 75.18 11.22 22.8 87.5 -1.05 1.15 Frequency Students Use Content Delivery Software 2264 All Years 2003-04 5.91 1.62 0.0 8.0 -0.64 -0.01 2004-05 5.12 1.82 0.0 8.0 -0.32 -0.48 2005-06 5.43 1.95 0.0 8.0 -0.54 -0.39 2006-07 5.53 2.01 0.0 8.0 -0.66 -0.23 1480 Elementary 2003-04 6.14 1.60 0.0 8.0 -0.82 0.33 2004-05 5.43 1.67 0.0 8.0 -0.41 -0.23 2005-06 5.57 1.87 0.0 8.0 -0.62 -0.16 2006-07 5.63 1.92 0.0 8.0 -0.69 -0.09 347 High 2003-04 5.34 1.62 1.0 8.0 -0.32 -0.21 2004-05 4.62 2.04 0.0 8.0 -0.08 -0.82 2005-06 5.29 2.13 0.0 8.0 -0.47 -0.67 2006-07 5.38 2.23 0.0 8.0 -0.62 -0.51 437 Middle 2003-04 5.60 1.56 0.0 8.0 -0.44 -0.07 2004-05 4.46 1.84 0.0 8.0 0.00 -0.63 2005-06 5.07 2.00 0.0 8.0 -0.30 -0.74 2006-07 5.30 2.08 0.0 8.0 -0.54 -0.48 Frequency Students Use Production Tool Software 2264 All Years 2003-04 6.49 2.09 0.0 12.0 -0.18 -0.06 2004-05 4.54 2.33 0.0 12.0 0.46 -0.02 2005-06 5.16 2.76 0.0 12.0 0.32 -0.56 2006-07 5.24 2.82 0.0 12.0 0.31 -0.60 1480 Elementary 2003-04 6.15 2.14 0.0 12.0 -0.08 -0.08 2004-05 4.03 2.12 0.0 11.0 0.46 0.03 2005-06 4.38 2.48 0.0 12.0 0.53 -0.08 2006-07 4.40 2.52 0.0 12.0 0.53 -0.11 347 High 2003-04 7.54 1.61 3.0 12.0 -0.04 0.26 2004-05 6.06 2.37 0.0 12.0 0.18 -0.24 2005-06 7.35 2.38 1.0 12.0 -0.38 -0.21 2006-07 7.57 2.41 0.0 12.0 -0.30 -0.19 437 Middle 2003-04 6.81 1.94 1.0 12.0 -0.09 -0.28 2004-05 5.06 2.31 0.0 12.0 0.45 -0.04 2005-06 6.05 2.75 0.0 12.0 0.15 -0.53 2006-07 6.25 2.69 0.0 12.0 0.08 -0.65 Level of Human Tech Support 2264 All Years 2003-04 6.48 2.77 0.0 12.0 0.06 -0.70 2004-05 6.62 2.84 0.0 12.0 0.04 -0.90 2005-06 7.14 2.76 2.0 14.0 0.02 -1.10 2006-07 7.94 2.58 1.0 13.0 -0.01 -0.96 1480 Elementary 2003-04 6.30 2.70 0.0 12.0 0.09 -0.66 2004-05 6.29 2.78 0.0 12.0 0.15 -0.85 2005-06 6.88 2.77 2.0 14.0 0.15 -1.09

PAGE 430

Appendix C: Data Preparation Procedures (Continued) 416 Variable N School Level School Year Mean STD Min Max Skew Kurt 2006-07 7.56 2.53 1.0 13.0 0.14 -0.92 347 High 2003-04 6.88 2.79 1.0 12.0 0.09 -0.79 2004-05 7.45 2.82 1.0 12.0 -0.20 -0.83 2005-06 7.75 2.65 3.0 13.0 -0.22 -0.89 2006-07 8.71 2.43 4.0 13.0 -0.29 -0.73 437 Middle 2003-04 6.78 2.90 1.0 12.0 -0.11 -0.74 2004-05 7.05 2.85 0.0 12.0 -0.17 -0.86 2005-06 7.57 2.69 2.0 12.0 -0.23 -1.00 2006-07 8.61 2.60 1.0 13.0 -0.29 -0.84 Level of Hardware/ Internet Dependability 2264 All Years 2003-04 6.01 1.15 0.0 8.0 -1.34 3.24 2004-05 6.25 1.17 0.0 8.0 -1.43 2.70 2005-06 6.13 1.47 0.0 8.0 -1.23 1.05 2006-07 6.35 1.37 0.0 8.0 -1.42 1.89 1480 Elementary 2003-04 5.98 1.18 0.0 8.0 -1.29 2.98 2004-05 6.21 1.16 0.0 8.0 -1.32 2.17 2005-06 6.03 1.52 0.0 8.0 -1.14 0.74 2006-07 6.31 1.41 0.0 8.0 -1.43 1.83 347 High 2003-04 6.14 1.00 1.0 8.0 -1.39 4.44 2004-05 6.32 1.18 1.0 8.0 -1.43 2.43 2005-06 6.29 1.35 1.0 8.0 -1.46 2.15 2006-07 6.41 1.29 1.0 8.0 -1.28 1.48 437 Middle 2003-04 6.00 1.18 0.0 8.0 -1.41 3.29 2004-05 6.32 1.19 0.0 8.0 -1.84 4.84 2005-06 6.32 1.38 0.0 8.0 -1.36 1.58 2006-07 6.44 1.28 1.0 8.0 -1.45 2.18 Table C 24. Descriptive Statistics of Predicto r Variables for Absences Outcome Variable N School Level School Year Mean STD Min Max Skew Kurt Percent Students on Free or Reduced Lunch Program 2311 All Schools 2003-04 52.24 25.36 1.0 100.0 0.10 -0.95 2004-05 52.64 24.10 0.9 100.0 0.01 -0.90 2005-06 52.23 23.88 1.7 100.0 -0.04 -0.97 1517 Elementary 2003-04 57.11 26.21 1.0 100.0 -0.14 -1.01 2004-05 56.57 24.85 0.9 100.0 -0.21 -0.93 2005-06 56.08 24.70 1.7 100.0 -0.25 -0.98 348 High 2003-04 35.50 17.17 1.8 93.3 0.45 0.17 2004-05 38.84 18.02 3.8 100.0 0.58 0.62 2005-06 38.66 17.19 2.5 93.7 0.17 -0.35 446 Middle 2003-04 48.75 21.35 3.7 100.0 0.12 -0.66 2004-05 50.10 21.19 3.3 100.0 0.05 -0.61

PAGE 431

Appendix C: Data Preparation Procedures (Continued) 417 Variable N School Level School Year Mean STD Min Max Skew Kurt 2005-06 49.75 21.20 3.9 100.0 0.02 -0.80 Percent Minority Students 2311 All Schools 2003-04 50.28 28.24 0.0 100.0 0.30 -1.12 2004-05 51.44 28.40 0.0 100.0 0.27 -1.14 2005-06 52.50 28.43 0.0 100.0 0.23 -1.17 1516 Elementary 2003-04 52.04 28.99 0.0 100.0 0.20 -1.22 2004-05 53.19 29.14 0.0 100.0 0.18 -1.24 2005-06 54.43 28.98 0.0 100.0 0.13 -1.25 349 High 2003-04 44.74 26.10 2.6 100.0 0.59 -0.65 2004-05 45.66 26.55 0.7 99.9 0.54 -0.77 2005-06 46.03 27.08 1.0 99.9 0.49 -0.85 446 Middle 2003-04 48.65 26.65 4.8 100.0 0.41 -0.89 2004-05 50.03 26.61 4.6 99.8 0.36 -0.96 2005-06 51.03 26.74 5.6 99.9 0.32 -1.00 Percent LEP students 2145 All Schools 2003-04 8.74 10.69 0.0 63.3 1.97 3.95 2004-05 8.53 10.55 0.0 61.6 2.01 4.17 2005-06 8.80 10.64 0.0 65.7 1.98 4.12 1394 Elementary 2003-04 10.81 12.22 0.1 63.3 1.56 1.96 2004-05 10.51 12.06 0.0 61.6 1.60 2.13 2005-06 10.82 12.11 0.1 65.7 1.57 2.16 328 High 2003-04 4.51 4.77 0.0 23.1 1.47 1.74 2004-05 4.36 4.65 0.0 26.0 1.53 2.10 2005-06 4.43 4.66 0.0 28.3 1.60 2.75 423 Middle 2003-04 5.23 5.44 0.1 32.5 1.70 3.52 2004-05 5.17 5.43 0.1 36.4 1.89 4.87 2005-06 5.46 5.70 0.1 37.4 2.04 5.74 Percent Students with Disabilities 2311 All Schools 2003-04 15.53 5.43 0.5 40.6 0.68 1.40 2004-05 15.31 5.33 0.4 43.9 0.73 1.98 2005-06 15.32 5.45 0.3 72.6 1.50 9.62 1517 Elementary 2003-04 16.08 5.73 1.2 40.6 0.80 1.24 2004-05 15.83 5.62 1.6 43.9 0.91 1.94 2005-06 16.05 5.78 1.9 72.6 1.78 10.45 348 High 2003-04 13.24 4.45 0.7 31.1 0.04 0.89 2004-05 13.19 4.48 0.4 27.6 -0.15 0.39 2005-06 13.15 4.48 0.3 30.9 -0.04 0.64 446 Middle 2003-04 15.44 4.56 0.5 27.9 -0.13 0.18 2004-05 15.23 4.47 0.5 28.7 -0.16 0.21 2005-06 14.51 4.29 0.8 28.7 -0.07 0.26 Percent Gifted students 1824 All Schools 2003-04 4.97 5.76 0.1 52.3 2.91 12.23 2004-05 4.99 5.87 0.1 54.9 2.95 12.63 2005-06 4.90 5.79 0.0 57.2 3.03 13.85 1370 Elementary 2003-04 4.30 5.42 0.1 52.3 3.41 17.27

PAGE 432

Appendix C: Data Preparation Procedures (Continued) 418 Variable N School Level School Year Mean STD Min Max Skew Kurt 2004-05 4.27 5.49 0.1 54.9 3.50 18.25 2005-06 4.19 5.56 0.1 57.2 3.69 20.51 22 High 2003-04 7.04 9.92 0.4 39.8 2.54 6.35 2004-05 7.60 10.36 0.7 39.6 2.20 4.50 2005-06 4.71 5.03 0.0 37.4 2.70 11.15 432 Middle 2003-04 7.00 6.05 0.1 37.3 2.04 5.18 2004-05 7.10 6.18 0.1 39.1 2.09 5.55 2005-06 7.28 6.34 0.1 39.5 2.10 5.76 Positive Learning Environment 2312 All Schools 2003-04 272.20 24.75 62.7 300 -1.64 4.03 2004-05 273.46 22.89 131.2 299.4 -1.46 2.23 2005-06 274.15 22.11 156.8 299.5 -1.53 2.55 1517 Elementary 2003-04 283.77 15.28 158.9 300 -3.42 17.44 2004-05 284.45 13.12 184.9 299.4 -2.90 12.88 2005-06 284.07 13.64 163.9 299.5 -2.97 13.42 349 High 2003-04 249.96 22.76 165.9 296.8 -0.61 0.58 2004-05 252.40 22.10 183.6 297.2 -0.59 0.30 2005-06 255.55 20.32 188.2 297.7 -0.32 0.03 446 Middle 2003-04 250.25 25.56 62.7 296.7 -1.56 6.69 2004-05 252.58 23.59 131.2 298.1 -0.95 1.79 2005-06 254.93 24.61 156.8 297.8 -1.02 1.39 Positive Teacher Qualifications 2312 All Schools 2003-04 140.00 16.81 61.3 200.7 -0.66 1.67 2004-05 139.09 18.11 59.9 194.4 -0.85 1.69 2005-06 137.42 18.99 40.4 191.6 -0.96 1.71 1517 Elementary 2003-04 139.08 17.55 61.3 187.9 -0.68 1.69 2004-05 137.68 19.20 59.9 192.4 -0.87 1.62 2005-06 136.27 20.05 40.4 191.6 -0.98 1.59 349 High 2003-04 146.37 14.58 79.9 200.7 -0.89 2.87 2004-05 146.21 14.46 90.2 194.4 -0.63 1.47 2005-06 143.11 16.23 80.5 191.6 -0.81 1.20 446 Middle 2003-04 138.12 14.63 81.8 177.4 -0.26 0.71 2004-05 138.28 15.37 89.5 173.7 -0.45 0.27 2005-06 136.91 16.36 61.6 174.3 -0.68 1.45 Percent of Student Computers with Content Software 2312 All Schools 2003-04 50.60 22.63 0.0 87.5 -0.03 -0.90 2004-05 54.25 22.46 0.0 87.5 -0.21 -0.88 2005-06 52.13 18.84 0.0 87.5 -0.04 -0.57 1517 Elementary 2003-04 53.03 22.32 0.0 87.5 -0.18 -0.80 2004-05 56.61 22.27 0.0 87.5 -0.36 -0.76 2005-06 54.29 18.65 0.0 87.5 -0.21 -0.42 349 High 2003-04 43.39 21.61 1.8 87.5 0.42 -0.62 2004-05 46.89 21.40 5.4 87.5 0.22 -0.94 2005-06 45.19 18.02 5.4 87.5 0.44 -0.09 446 Middle 2003-04 47.95 23.03 0.0 87.5 0.14 -0.92 2004-05 51.95 22.47 0.0 87.5 -0.09 -0.80

PAGE 433

Appendix C: Data Preparation Procedures (Continued) 419 Variable N School Level School Year Mean STD Min Max Skew Kurt 2005-06 50.20 18.65 7.1 87.5 0.18 -0.67 Percent of Student Computers with Office/ Production Software 2312 All Schools 2003-04 74.89 16.80 0.0 87.5 -1.66 2.68 2004-05 76.71 15.63 0.0 87.5 -1.79 3.19 2005-06 78.13 14.96 0.0 87.5 -2.16 5.03 1517 Elementary 2003-04 72.22 18.16 0.0 87.5 -1.40 1.66 2004-05 74.59 16.86 0.0 87.5 -1.49 1.83 2005-06 75.93 16.66 0.0 87.5 -1.84 3.26 349 High 2003-04 80.97 11.51 10.0 87.5 -2.52 8.14 2004-05 81.56 11.50 0.0 87.5 -3.05 12.94 2005-06 82.63 10.32 10.0 87.5 -3.52 16.52 446 Middle 2003-04 79.21 12.96 10.0 87.5 -2.09 5.27 2004-05 80.14 12.34 0.0 87.5 -2.42 7.86 2005-06 82.12 9.18 37.3 87.5 -1.98 3.99 Percent of Student Computers with Advanced Production Software 2312 All Schools 2003-04 22.21 18.36 0.0 87.5 1.14 1.05 2004-05 21.89 18.68 0.0 87.5 1.14 1.04 2005-06 26.00 21.28 0.0 87.5 0.87 0.08 1517 Elementary 2003-04 22.05 18.56 0.0 87.5 1.06 0.84 2004-05 21.29 18.92 0.0 87.5 1.06 0.75 2005-06 25.54 21.33 0.0 87.5 0.78 -0.11 349 High 2003-04 22.99 17.07 0.0 87.5 1.40 1.81 2004-05 22.99 16.84 0.0 87.5 1.41 2.12 2005-06 27.63 20.56 0.0 87.5 1.13 0.66 446 Middle 2003-04 22.14 18.68 0.0 87.5 1.29 1.32 2004-05 23.05 19.20 0.0 87.5 1.27 1.37 2005-06 26.26 21.62 0.0 87.5 0.99 0.28 Percent of Teachers Who Regularly Use Technology to Delivery Instruction 2312 All Schools 2003-04 20.01 12.18 0.0 87.5 1.11 1.76 2004-05 22.30 12.72 0.0 87.5 1.01 1.44 2005-06 23.90 13.84 0.0 87.5 0.85 0.86 1517 Elementary 2003-04 18.26 11.74 0.0 87.5 1.28 2.59 2004-05 20.29 12.40 0.0 87.5 1.23 2.50 2005-06 21.49 13.36 0.0 87.5 1.04 1.44 349 High 2003-04 24.56 11.40 5.0 68.0 0.77 0.57 2004-05 27.02 12.36 5.0 68.0 0.85 0.65 2005-06 28.32 12.74 5.0 67.8 0.52 -0.24 446 Middle 2003-04 22.40 12.90 2.5 78.1 1.07 1.34 2004-05 25.40 12.50 0.0 70.0 0.73 0.26 2005-06 28.62 14.21 2.5 87.5 0.76 0.89 Percent of Teachers Who Regularly Use Technology for Administrative Purposes 2312 All Schools 2003-04 65.76 15.66 5.0 87.5 -0.86 0.54 2004-05 68.57 14.72 0.0 87.5 -1.07 1.30 2005-06 61.59 17.08 0.0 87.5 -0.74 0.27 1517 Elementary 2003-04 63.50 15.96 5.0 87.5 -0.78 0.37

PAGE 434

Appendix C: Data Preparation Procedures (Continued) 420 Variable N School Level School Year Mean STD Min Max Skew Kurt 2004-05 66.67 15.24 0.0 87.5 -1.05 1.21 2005-06 59.24 17.43 0.0 87.5 -0.67 0.12 349 High 2003-04 66.28 15.75 5.0 87.5 -0.90 0.76 2004-05 69.84 13.95 12.5 87.5 -1.10 1.30 2005-06 71.77 12.95 27.7 87.5 -0.97 0.71 446 Middle 2003-04 65.91 15.20 14.9 87.5 -0.86 0.62 2004-05 70.25 14.24 12.5 87.5 -0.93 0.77 2005-06 72.53 12.90 17.4 87.5 -1.01 1.03 Frequency Students Use Content Delivery Software 2312 All Schools 2003-04 5.12 1.81 0.0 8.0 -0.32 -0.48 2004-05 5.43 1.95 0.0 8.0 -0.53 -0.38 2005-06 5.92 1.62 0.0 8.0 -0.65 0.01 1517 Elementary 2003-04 5.43 1.67 0.0 8.0 -0.41 -0.23 2004-05 5.57 1.87 0.0 8.0 -0.61 -0.16 2005-06 6.15 1.59 0.0 8.0 -0.83 0.35 349 High 2003-04 5.32 1.63 1.0 8.0 -0.33 -0.20 2004-05 4.62 2.04 0.0 8.0 -0.07 -0.81 2005-06 5.28 2.14 0.0 8.0 -0.47 -0.65 446 Middle 2003-04 5.60 1.56 0.0 8.0 -0.43 -0.06 2004-05 4.46 1.83 0.0 8.0 0.01 -0.63 2005-06 5.08 2.00 0.0 8.0 -0.29 -0.74 Frequency Students Use Production Tool Software 2312 All Schools 2003-04 4.52 2.33 0.0 12.0 0.46 -0.02 2004-05 5.13 2.77 0.0 12.0 0.34 -0.56 2005-06 6.47 2.10 0.0 12.0 -0.18 -0.04 1517 Elementary 2003-04 4.01 2.13 0.0 11.0 0.46 0.01 2004-05 4.36 2.48 0.0 12.0 0.54 -0.08 2005-06 6.12 2.14 0.0 12.0 -0.09 -0.08 349 High 2003-04 7.52 1.63 3.0 12.0 -0.05 0.23 2004-05 6.06 2.36 0.0 12.0 0.18 -0.22 2005-06 7.34 2.39 1.0 12.0 -0.39 -0.22 446 Middle 2003-04 6.82 1.93 1.0 12.0 -0.09 -0.26 2004-05 5.05 2.30 0.0 12.0 0.46 -0.03 2005-06 6.04 2.77 0.0 12.0 0.17 -0.54 Level of Human Tech Support 2312 All Schools 2003-04 6.57 2.85 0.0 12.0 0.05 -0.91 2004-05 7.12 2.77 2.0 14.0 0.04 -1.10 2005-06 6.46 2.77 0.0 12.0 0.07 -0.71 1517 Elementary 2003-04 6.24 2.79 0.0 12.0 0.16 -0.85 2004-05 6.84 2.78 2.0 14.0 0.17 -1.09 2005-06 6.27 2.70 0.0 12.0 0.10 -0.66 349 High 2003-04 6.87 2.79 1.0 12.0 0.10 -0.79 2004-05 7.46 2.82 1.0 12.0 -0.21 -0.83 2005-06 7.78 2.66 3.0 13.0 -0.22 -0.88 446 Middle 2003-04 6.75 2.92 1.0 12.0 -0.10 -0.78 2004-05 7.01 2.85 0.0 12.0 -0.14 -0.88

PAGE 435

Appendix C: Data Preparation Procedures (Continued) 421 Variable N School Level School Year Mean STD Min Max Skew Kurt 2005-06 7.56 2.71 2.0 14.0 -0.20 -1.00 Level of Hardware/ Internet Dependability 2312 All Schools 2003-04 6.25 1.16 0.0 8.0 -1.43 2.72 2004-05 6.13 1.47 0.0 8.0 -1.23 1.07 2005-06 6.01 1.15 0.0 8.0 -1.33 3.22 1517 Elementary 2003-04 6.22 1.15 0.0 8.0 -1.31 2.18 2004-05 6.04 1.52 0.0 8.0 -1.15 0.77 2005-06 5.99 1.18 0.0 8.0 -1.28 2.97 349 High 2003-04 6.14 1.00 1.0 8.0 -1.38 4.30 2004-05 6.32 1.18 1.0 8.0 -1.44 2.45 2005-06 6.29 1.35 1.0 8.0 -1.45 2.09 446 Middle 2003-04 6.01 1.17 0.0 8.0 -1.40 3.33 2004-05 6.33 1.18 0.0 8.0 -1.85 4.92 2005-06 6.33 1.38 0.0 8.0 -1.36 1.62 Table C 25. Descriptive Statistics of Predictor Variables for Student Conduct Outcome Variable N School Level School Year Mean STD Min Max Skew Kurt Percent Students on Free or Reduced Lunch Program 2311 All Schools 2003-04 52.2425.361.0100.0 0.10 -0.95 2004-05 52.6424.100.9100.0 0.01 -0.90 2005-06 52.2323.881.7100.0 -0.04 -0.97 1517 Elementary 2003-04 57.1126.211.0100.0 -0.14 -1.01 2004-05 56.5724.850.9100.0 -0.21 -0.93 2005-06 56.0824.701.7100.0 -0.25 -0.98 348 High 2003-04 35.5017.171.893.3 0.45 0.17 2004-05 38.8418.023.8100.0 0.58 0.62 2005-06 38.6617.192.593.7 0.17 -0.35 446 Middle 2003-04 48.7521.353.7100.0 0.12 -0.66 2004-05 50.1021.193.3100.0 0.05 -0.61 2005-06 49.7521.203.9100.0 0.02 -0.80 Percent Minority Students 2311 All Schools 2003-04 50.2828.240.0100.0 0.30 -1.12 2004-05 51.4428.400.0100.0 0.27 -1.14 2005-06 52.5028.430.0100.0 0.23 -1.17 1516 Elementary 2003-04 52.0428.990.0100.0 0.20 -1.22 2004-05 53.1929.140.0100.0 0.18 -1.24 2005-06 54.4328.980.0100.0 0.13 -1.25 349 High 2003-04 44.7426.102.6100.0 0.59 -0.65 2004-05 45.6626.550.799.9 0.54 -0.77 2005-06 46.0327.081.099.9 0.49 -0.85 446 Middle 2003-04 48.6526.654.8100.0 0.41 -0.89 2004-05 50.0326.614.699.8 0.36 -0.96

PAGE 436

Appendix C: Data Preparation Procedures (Continued) 422 Variable N School Level School Year Mean STD Min Max Skew Kurt 2005-06 51.0326.745.699.9 0.32 -1.00 Percent LEP students 2145 All Schools 2003-04 8.7410.690.063.3 1.97 3.95 2004-05 8.5310.550.061.6 2.01 4.17 2005-06 8.8010.640.065.7 1.98 4.12 1394 Elementary 2003-04 10.8112.220.163.3 1.56 1.96 2004-05 10.5112.060.061.6 1.60 2.13 2005-06 10.8212.110.165.7 1.57 2.16 328 High 2003-04 4.514.770.023.1 1.47 1.74 2004-05 4.364.650.026.0 1.53 2.10 2005-06 4.434.660.028.3 1.60 2.75 423 Middle 2003-04 5.235.440.132.5 1.70 3.52 2004-05 5.175.430.136.4 1.89 4.87 2005-06 5.465.700.137.4 2.04 5.74 Percent Students with Disabilities 2311 All Schools 2003-04 15.535.430.540.6 0.68 1.40 2004-05 15.315.330.443.9 0.73 1.98 2005-06 15.325.450.372.6 1.50 9.62 1517 Elementary 2003-04 16.085.731.240.6 0.80 1.24 2004-05 15.835.621.643.9 0.91 1.94 2005-06 16.055.781.972.6 1.78 10.45 348 High 2003-04 13.244.450.731.1 0.04 0.89 2004-05 13.194.480.427.6 -0.15 0.39 2005-06 13.154.480.330.9 -0.04 0.64 446 Middle 2003-04 15.444.560.527.9 -0.13 0.18 2004-05 15.234.470.528.7 -0.16 0.21 2005-06 14.514.290.828.7 -0.07 0.26 Percent Gifted students 1824 All Schools 2003-04 4.975.760.152.3 2.91 12.23 2004-05 4.995.870.154.9 2.95 12.63 2005-06 4.905.790.057.2 3.03 13.85 1370 Elementary 2003-04 4.305.420.152.3 3.41 17.27 2004-05 4.275.490.154.9 3.50 18.25 2005-06 4.195.560.157.2 3.69 20.51 22 High 2003-04 7.049.920.439.8 2.54 6.35 2004-05 7.6010.360.739.6 2.20 4.50 2005-06 4.715.030.037.4 2.70 11.15 432 Middle 2003-04 7.006.050.137.3 2.04 5.18 2004-05 7.106.180.139.1 2.09 5.55 2005-06 7.286.340.139.5 2.10 5.76 Positive Learning Environment 2312 All Schools 2003-04 185.397.06147.3199.2 -1.20 2.49 2004-05 183.818.2480.9198.5 -1.83 11.75 2005-06 183.618.23134.6199 -1.10 1.94 1517 Elementary 2003-04 187.664.90162.6199.2 -0.59 0.68

PAGE 437

Appendix C: Data Preparation Procedures (Continued) 423 Variable N School Level School Year Mean STD Min Max Skew Kurt 2004-05 185.806.7980.9197.9 -3.19 38.18 2005-06 185.726.67134.6197.9 -1.29 3.65 349 High 2003-04 178.658.94151.2198.4 -0.44 0.20 2004-05 177.559.73150.8198.4 -0.41 -0.01 2005-06 176.599.95137.7198.5 -0.40 0.54 446 Middle 2003-04 182.957.47147.3198.2 -0.92 2.31 2004-05 181.918.58144.8198.5 -0.94 1.85 2005-06 181.948.10147.7199 -0.63 0.71 Positive Teacher Qualifications 2312 All Schools 2003-04 140.0016.8161.3200.7 -0.66 1.67 2004-05 139.0918.1159.9194.4 -0.85 1.69 2005-06 137.4218.9940.4191.6 -0.96 1.71 1517 Elementary 2003-04 139.0817.5561.3187.9 -0.68 1.69 2004-05 137.6819.2059.9192.4 -0.87 1.62 2005-06 136.2720.0540.4191.6 -0.98 1.59 349 High 2003-04 146.3714.5879.9200.7 -0.89 2.87 2004-05 146.2114.4690.2194.4 -0.63 1.47 2005-06 143.1116.2380.5191.6 -0.81 1.20 446 Middle 2003-04 138.1214.6381.8177.4 -0.26 0.71 2004-05 138.2815.3789.5173.7 -0.45 0.27 2005-06 136.9116.3661.6174.3 -0.68 1.45 Percent of Student Computers with Content Software 2312 All Schools 2003-04 50.6022.630.087.5 -0.03 -0.90 2004-05 54.2522.460.087.5 -0.21 -0.88 2005-06 52.1318.840.087.5 -0.04 -0.57 1517 Elementary 2003-04 53.0322.320.087.5 -0.18 -0.80 2004-05 56.6122.270.087.5 -0.36 -0.76 2005-06 54.2918.650.087.5 -0.21 -0.42 349 High 2003-04 43.3921.611.887.5 0.42 -0.62 2004-05 46.8921.405.487.5 0.22 -0.94 2005-06 45.1918.025.487.5 0.44 -0.09 446 Middle 2003-04 47.9523.030.087.5 0.14 -0.92 2004-05 51.9522.470.087.5 -0.09 -0.80 2005-06 50.2018.657.187.5 0.18 -0.67 Percent of Student Computers with Office/ Production Software 2312 All Schools 2003-04 74.8916.800.087.5 -1.66 2.68 2004-05 76.7115.630.087.5 -1.79 3.19 2005-06 78.1314.960.087.5 -2.16 5.03 1517 Elementary 2003-04 72.2218.160.087.5 -1.40 1.66 2004-05 74.5916.860.087.5 -1.49 1.83 2005-06 75.9316.660.087.5 -1.84 3.26 349 High 2003-04 80.9711.5110.087.5 -2.52 8.14 2004-05 81.5611.500.087.5 -3.05 12.94 2005-06 82.6310.3210.087.5 -3.52 16.52 446 Middle 2003-04 79.2112.9610.087.5 -2.09 5.27 2004-05 80.1412.340.087.5 -2.42 7.86

PAGE 438

Appendix C: Data Preparation Procedures (Continued) 424 Variable N School Level School Year Mean STD Min Max Skew Kurt 2005-06 82.129.1837.387.5 -1.98 3.99 Percent of Student Computers with Advanced Production Software 2312 All Schools 2003-04 22.2118.360.087.5 1.14 1.05 2004-05 21.8918.680.087.5 1.14 1.04 2005-06 26.0021.280.087.5 0.87 0.08 1517 Elementary 2003-04 22.0518.560.087.5 1.06 0.84 2004-05 21.2918.920.087.5 1.06 0.75 2005-06 25.5421.330.087.5 0.78 -0.11 349 High 2003-04 22.9917.070.087.5 1.40 1.81 2004-05 22.9916.840.087.5 1.41 2.12 2005-06 27.6320.560.087.5 1.13 0.66 446 Middle 2003-04 22.1418.680.087.5 1.29 1.32 2004-05 23.0519.200.087.5 1.27 1.37 2005-06 26.2621.620.087.5 0.99 0.28 Percent of Teachers Who Regularly Use Technology to Delivery Instruction 2312 All Schools 2003-04 20.0112.180.087.5 1.11 1.76 2004-05 22.3012.720.087.5 1.01 1.44 2005-06 23.9013.840.087.5 0.85 0.86 1517 Elementary 2003-04 18.2611.740.087.5 1.28 2.59 2004-05 20.2912.400.087.5 1.23 2.50 2005-06 21.4913.360.087.5 1.04 1.44 349 High 2003-04 24.5611.405.068.0 0.77 0.57 2004-05 27.0212.365.068.0 0.85 0.65 2005-06 28.3212.745.067.8 0.52 -0.24 446 Middle 2003-04 22.4012.902.578.1 1.07 1.34 2004-05 25.4012.500.070.0 0.73 0.26 2005-06 28.6214.212.587.5 0.76 0.89 Percent of Teachers Who Regularly Use Technology for Administrative Purposes 2312 All Schools 2003-04 61.5917.080.087.5 -0.74 0.27 2004-05 65.7615.665.087.5 -0.86 0.54 2005-06 68.5714.720.087.5 -1.07 1.30 1517 Elementary 2003-04 59.2417.430.087.5 -0.67 0.12 2004-05 63.5015.965.087.5 -0.78 0.37 2005-06 66.6715.240.087.5 -1.05 1.21 349 High 2003-04 66.2815.755.087.5 -0.90 0.76 2004-05 69.8413.9512.587.5 -1.10 1.30 2005-06 71.7712.9527.787.5 -0.97 0.71 446 Middle 2003-04 65.9115.2014.987.5 -0.86 0.62 2004-05 70.2514.2412.587.5 -0.93 0.77 2005-06 72.5312.9017.487.5 -1.01 1.03 Frequency Students Use Content Delivery Software 2312 All Schools 2003-04 5.921.620.08.0 -0.65 0.01 2004-05 5.121.810.08.0 -0.32 -0.48 2005-06 5.431.950.08.0 -0.53 -0.38 1517 Elementary 2003-04 6.151.590.08.0 -0.83 0.35

PAGE 439

Appendix C: Data Preparation Procedures (Continued) 425 Variable N School Level School Year Mean STD Min Max Skew Kurt 2004-05 5.431.670.08.0 -0.41 -0.23 2005-06 5.571.870.08.0 -0.61 -0.16 349 High 2003-04 5.321.631.08.0 -0.33 -0.20 2004-05 4.622.040.08.0 -0.07 -0.81 2005-06 5.282.140.08.0 -0.47 -0.65 446 Middle 2003-04 5.601.560.08.0 -0.43 -0.06 2004-05 4.461.830.08.0 0.01 -0.63 2005-06 5.082.000.08.0 -0.29 -0.74 Frequency Students Use Production Tool Software 2312 All Schools 2003-04 6.472.100.012.0 -0.18 -0.04 2004-05 4.522.330.012.0 0.46 -0.02 2005-06 5.132.770.012.0 0.34 -0.56 1517 Elementary 2003-04 6.122.140.012.0 -0.09 -0.08 2004-05 4.012.130.011.0 0.46 0.01 2005-06 4.362.480.012.0 0.54 -0.08 349 High 2003-04 7.521.633.012.0 -0.05 0.23 2004-05 6.062.360.012.0 0.18 -0.22 2005-06 7.342.391.012.0 -0.39 -0.22 446 Middle 2003-04 6.821.931.012.0 -0.09 -0.26 2004-05 5.052.300.012.0 0.46 -0.03 2005-06 6.042.770.012.0 0.17 -0.54 Level of Human Tech Support 2312 All Schools 2003-04 6.462.770.012.0 0.07 -0.71 2004-05 6.572.850.012.0 0.05 -0.91 2005-06 7.122.772.014.0 0.04 -1.10 1517 Elementary 2003-04 6.272.700.012.0 0.10 -0.66 2004-05 6.242.790.012.0 0.16 -0.85 2005-06 6.842.782.014.0 0.17 -1.09 349 High 2003-04 6.872.791.012.0 0.10 -0.79 2004-05 7.462.821.012.0 -0.21 -0.83 2005-06 7.782.663.013.0 -0.22 -0.88 446 Middle 2003-04 6.752.921.012.0 -0.10 -0.78 2004-05 7.012.850.012.0 -0.14 -0.88 2005-06 7.562.712.014.0 -0.20 -1.00 Level of Hardware/ Internet Dependability 2312 All Schools 2003-04 6.011.150.08.0 -1.33 3.22 2004-05 6.251.160.08.0 -1.43 2.72 2005-06 6.131.470.08.0 -1.23 1.07 1517 Elementary 2003-04 5.991.180.08.0 -1.28 2.97 2004-05 6.221.150.08.0 -1.31 2.18 2005-06 6.041.520.08.0 -1.15 0.77 349 High 2003-04 6.141.001.08.0 -1.38 4.30 2004-05 6.321.181.08.0 -1.44 2.45

PAGE 440

Appendix C: Data Preparation Procedures (Continued) 426 Variable N School Level School Year Mean STD Min Max Skew Kurt 2005-06 6.291.351.08.0 -1.45 2.09 446 Middle 2003-04 6.011.170.08.0 -1.40 3.33 2004-05 6.331.180.08.0 -1.85 4.92 2005-06 6.331.380.08.0 -1.36 1.62 Correlations of Technology Indicators with Predictor Variables Table C 26. Correlations and P-values of Predictor Variables for Learning Environment and Technology Indicators for FCAT Reading Outcome Variable 1 2 3 4 5 6 7 8 9 10 1 Positive Learning Environment 2 0.10 Positive Teacher Qualifications <.0001 3 0.13 0.04 Percent of Student Computers with Content Software <.0001 <.0001 4 -0.05 -0.02 0.20 Percent of Student Computers with Office/ Production Software <.0001 0.0329 <.0001 5 0.08 0.11 0.31 0.32 Percent of Student Computers with Advanced Production Software <.0001 <.0001 <.0001 <.0001 6 -0.02 0.07 0.27 0.24 0.35 Percent of Teachers Who Regularly Use Technology to Deliver Instruction 0.0581 <.0001 <.0001 <.0001 <.0001 7 0.04 0.04 0.26 0.33 0.27 0.56 Percent of Teachers Who Regularly Use Technology for Administrative Purposes <.0001 0.0005 <.0001 <.0001 <.0001 <.0001 8 0.04 0.01 0.30 0.02 0.06 0.08 0.10 Frequency Students Use Content Delivery Software 0.0006 0.2158 <.0001 0.0427 <.0001 <.0001 <.0001 9 -0.11 0.13 0.13 0.22 0.26 0.35 0.27 0.25 Frequency Students Use Production Tool Software <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 10 -0.08 0.10 0.11 0.11 0.14 0.20 0.19 0.10 0.12 Level of Human Tech Support <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 11 -0.05 0.01 0.06 0.08 0.03 0.05 0.08 0.03 0.01 0.08 Level of Hardware/ Internet Dependability <.0001 0.5997 <.0001 <.0001 0.0014 <.0001 <.0001 0.0128 0.2939 <.0001

PAGE 441

Appendix C: Data Preparation Procedures (Continued) 427 Table C 27. Correlations and P-values of Predictor Variables for Learning Environment and Technology Indicators for FCAT Math Outcome Variable 1 2 3 4 5 6 7 8 9 10 1 Positive Learning Environment 2 0.10 Positive Teacher Qualifications <.0001 3 0.13 0.04 Percent of Student Computers with Content Software <.0001 <.0001 4 -0.05 -0.02 0.20 Percent of Student Computers with Office/ Production Software <.0001 0.0235 <.0001 5 0.09 0.11 0.31 0.32 Percent of Student Computers with Advanced Production Software <.0001 <.0001 <.0001 <.0001 6 -0.02 0.07 0.27 0.24 0.35 Percent of Teachers Who Regularly Use Technology to Deliver Instruction 0.0602 <.0001 <.0001 <.0001 <.0001 7 0.04 0.04 0.26 0.33 0.27 0.56 Percent of Teachers Who Regularly Use Technology for Administrative Purposes 0.0001 0.0003 <.0001 <.0001 <.0001 <.0001 8 0.04 0.01 0.29 0.02 0.05 0.08 0.10 Frequency Students Use Content Delivery Software 0.0006 0.238 <.0001 0.0574 <.0001 <.0001 <.0001 9 -0.11 0.13 0.13 0.22 0.26 0.35 0.27 0.24 Frequency Students Use Production Tool Software <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 10 -0.08 0.10 0.11 0.11 0.14 0.20 0.19 0.10 0.13 Level of Human Tech Support <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 11 -0.05 0.00 0.06 0.08 0.03 0.05 0.08 0.03 0.01 0.08 Level of Hardware/ Internet Dependability <.0001 0.7154 <.0001 <.0001 0.001 <.0001 <.0001 0.0067 0.3345 <.0001

PAGE 442

Appendix C: Data Preparation Procedures (Continued) 428 Table C 28. Correlations and P-values of Predictor Variables for Learning Environment and Technology Indicators for FCAT Writing Outcome Variable 1 2 3 4 5 6 7 8 9 10 1 Positive Learning Environment 2 0.11 Positive Teacher Qualifications <.0001 3 0.14 0.04 Percent of Student Computers with Content Software <.0001 <.0001 4 -0.05 -0.02 0.20 Percent of Student Computers with Office/ Production Software <.0001 0.0407 <.0001 5 0.09 0.12 0.31 0.32 Percent of Student Computers with Advanced Production Software <.0001 <.0001 <.0001 <.0001 6 -0.02 0.07 0.27 0.24 0.35 Percent of Teachers Who Regularly Use Technology to Deliver Instruction 0.0731 <.0001 <.0001 <.0001 <.0001 7 0.04 0.04 0.26 0.33 0.26 0.56 Percent of Teachers Who Regularly Use Technology for Administrative Purposes <.0001 0.0002 <.0001 <.0001 <.0001 <.0001 8 0.03 0.01 0.30 0.02 0.06 0.08 0.11 Frequency Students Use Content Delivery Software 0.0016 0.1941 <.0001 0.0237 <.0001 <.0001 <.0001 9 -0.11 0.13 0.13 0.22 0.26 0.35 0.27 0.25 Frequency Students Use Production Tool Software <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 10 -0.08 0.10 0.10 0.11 0.13 0.20 0.19 0.10 0.12 Level of Human Tech Support <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 11 -0.05 0.01 0.06 0.08 0.03 0.06 0.08 0.03 0.01 0.08 Level of Hardware/ Internet Dependability <.0001 0.5539 <.0001 <.0001 0.0012 <.0001 <.0001 0.0112 0.2295 <.0001

PAGE 443

Appendix C: Data Preparation Procedures (Continued) 429 Table C 29. Correlations and P-values of Predictor Variables for Learning Environment and Technology Indicators for Absences Outcome Variable 1 2 3 4 5 6 7 8 9 10 1 Positive Learning Environment 2 0.09 Positive Teacher Qualifications <.0001 3 0.12 0.03 Percent of Student Computers with Content Software <.0001 0.0298 4 -0.05 -0.03 0.19 Percent of Student Computers with Office/ Production Software <.0001 0.0202 <.0001 5 0.10 0.09 0.31 0.33 Percent of Student Computers with Advanced Production Software <.0001 <.0001 <.0001 <.0001 6 -0.02 0.07 0.28 0.25 0.35 Percent of Teachers Who Regularly Use Technology to Deliver Instruction 0.1122 <.0001 <.0001 <.0001 <.0001 7 0.04 0.04 0.27 0.34 0.28 0.56 Percent of Teachers Who Regularly Use Technology for Administrative Purposes 0.0003 0.0005 <.0001 <.0001 <.0001 <.0001 8 0.06 0.00 0.29 0.00 0.05 0.06 0.09 Frequency Students Use Content Delivery Software <.0001 0.7296 <.0001 0.8327 <.0001 <.0001 <.0001 9 -0.10 0.12 0.13 0.22 0.27 0.33 0.27 0.22 Frequency Students Use Production Tool Software <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 10 -0.06 0.09 0.11 0.09 0.12 0.17 0.16 0.09 0.11 Level of Human Tech Support <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 11 -0.07 -0.01 0.04 0.07 0.03 0.05 0.06 0.02 0.01 0.07 Level of Hardware/ Internet Dependability <.0001 0.3145 2E-04 <.0001 0.0154 <.0001 <.0001 0.0794 0.5377 <.0001

PAGE 444

Appendix C: Data Preparation Procedures (Continued) 430 Table C 30. Correlations and P-values of Predictor Variables for Learning Environment and Technology Indicators for Conduct Outcome Variable 1 2 3 4 5 6 7 8 9 10 1 Positive Learning Environment 2 0.12 Positive Teacher Qualifications <.0001 3 0.07 0.03 Percent of Student Computers with Content Software <.0001 0.0298 4 -0.01 -0.03 0.19 Percent of Student Computers with Office/ Production Software 0.3578 0.0202 <.0001 5 0.05 0.09 0.31 0.33 Percent of Student Computers with Advanced Production Software <.0001 <.0001 <.0001 <.0001 6 0.02 0.07 0.28 0.25 0.35 Percent of Teachers Who Regularly Use Technology to Deliver Instruction 0.2091 <.0001 <.0001 <.0001 <.0001 7 0.07 0.04 0.27 0.34 0.28 0.56 Percent of Teachers Who Regularly Use Technology for Administrative Purposes <.0001 0.0005 <.0001 <.0001 <.0001 <.0001 8 0.01 0.00 0.29 0.00 0.05 0.06 0.09 Frequency Students Use Content Delivery Software 0.2925 0.7296 <.0001 0.8327 <.0001 <.0001 <.0001 9 0.00 0.12 0.13 0.22 0.27 0.33 0.27 0.22 Frequency Students Use Production Tool Software 0.7315 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 10 -0.04 0.09 0.11 0.09 0.12 0.17 0.16 0.09 0.11 Level of Human Tech Support 0.0002 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 11 -0.02 -0.01 0.04 0.07 0.03 0.05 0.06 0.02 0.01 0.07 Level of Hardware/ Internet Dependability 0.0605 0.3145 0.0002 <.0001 0.0154 <.0001 <.0001 0.0794 0.5377 <.0001

PAGE 445

431 Appendix D: Permissions

PAGE 446

Appendix D: Permissions 432

PAGE 447

Appendix D: Permissions 433

PAGE 448

About the Author Tina Newstein Hohlfeld has devoted her career to the empowerment of all children. In this capacity, she has worked as a regu lar education and special educatio n teacher in early childhood, elementary, middle school, and high school settings. Also, she has supported the education of young children, parents, and teachers as a childcare director an administrator in an early intervention center, an adjunct professor, and a consultant supporting the implementation of quality child care programs. Through out all of these experiences the integration of techno logy has been crucial to success of the project or program and children. In her latest career spiral at the University of South Florida, she has learned how to conduct research about how technology integration su pports student achievement. Ms. Hohlfeld earned a Bachelors of Science degree in Humanities and Tech nology at Drexel University and a Masters of Education degree in Elementary and Speci al Education at Lehigh University.


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001970779
003 fts
005 20081202091240.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 081202s2007 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002333
035
(OCoLC)276841333
040
FHM
c FHM
049
FHMM
090
LB1028.3 (ONLINE)
1 100
Hohlfeld, Tina N.
4 245
The relationship between technology integration and achievement using multi-level modeling
h [electronic resource] /
by Tina N. Hohlfeld.
260
[Tampa, Fla.] :
b University of South Florida,
2007.
3 520
ABSTRACT: The purpose of this longitudinal study was to examine the relationship between technology integration indicators and school level achievement. Four years of school level secondary data from publicly available databases maintained by the Florida Department of Education were combined for all public elementary, middle, and high schools in the state. This study examined approximately 2300 schools that participated each year in the Florida Innovates Survey about technology integration between 2003-04 and 2006-07. Complexity theory supported the use of multi-level modeling to examine the relationships between technology integration and outcomes. Three achievement outcomes (reading, mathematics, and writing) and two mediating behavioral outcomes (attendance and misconduct) were investigated. Moderating variables controlled in the model included school level, demographics, and learning environment.^ After data preparation, all composite variables were developed using factor analysis. Models were progressively built with significant variables at each level retained in subsequent levels of the study. A total of 94 models were estimated with maximum likelihood estimation using SAS 9.1.3 statistical software. The integration of technology is only one of the many factors that impact student learning within the classroom environment. Results supported previous research about the relationship between the moderating variables and school level achievement and confirmed the need to include moderating variables in the model. After controlling for all the other moderating variables, technology integration had a significant relationship with mean school achievement.^ Although the percent of teachers who regularly use technology for administrative purposes was consistently significant in the models for four out of five outcomes studied, the interactions with time, time2, and time3, resulted in curvilinear trends with inconsistent results. These inconsistent significant findings make drawing conclusions about the integration of technology within Florida's public schools difficult. Furthermore, the small changes observed in mean school achievement over the span of this study support the concept that time is a critical factor for school level learning and change. Therefore, continued analyses of the longitudinal trends for Florida schools in the relationship between technology integration variables and school achievement, while controlling for moderating variables, are recommended.
502
Dissertation (Ph.D.)--University of South Florida, 2007.
504
Includes bibliographical references.
516
Text (Electronic dissertation) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 433 pages.
Includes vita.
590
Adviser: Ann E. Barron, Ed.D.
653
Florida.
Longitudinal research.
Public schools.
Secondary data.
Student outcomes.
0 690
Dissertations, Academic
z USF
x Instructional Technology
Doctoral.
773
t USF Electronic Theses and Dissertations.
856
u http://digital.lib.usf.edu/?e14.2333