USF Libraries
USF Digital Collections

Reading fluency through alternative text

MISSING IMAGE

Material Information

Title:
Reading fluency through alternative text rereading with an interactive sing-to-read program embedded within middle school music classroom
Physical Description:
Book
Creator:
Biggs, Marie
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Singing
Rereading
Fluency
Early adolescents
Alternative text
Dissertations, Academic -- Childhood Education -- Doctoral -- USF   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: Singing exaggerates the language of reading. The students find their voices in the rhythm and bounce of language by using music as an alternative text. A concurrent mixed methods study was conducted to investigate the use of an interactive sing-to-read program Tune Into Reading (Electronic Learning Products, 2006) as an alternative text, embedded within a heterogeneous music classroom. Measured by the Qualitative Reading Inventory-4 (QRI-4) (Leslie & Caldwell, 2006), the fluency, word recognition, comprehension, and instructional reading level of the treatment students were compared to their counterparts who sang as part of the regular music program. Concurrently, this investigation also provided a description of the peers' interactions during the literacy task assigned by the music teacher. The intent of this study was to address the following three research questions.Furthermore, the use of the interactive program provided opportunities for differentiated reading level achievement. Finally, group dynamics highly influenced the early adolescent's motivation, engagement, participation, and successful outcomes in reading fluency.
Thesis:
Dissertation (Ph.D.)--University of South Florida, 2007.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Marie Biggs.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 287 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 - 001988959
oclc - 307532767
usfldc doi - E14-SFE0002184
usfldc handle - e14.2184
System ID:
SFS0026502:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Reading Fluency Through Alternative Text: Rereading With an Interactive Sing-to-Read Program Embedded W ithin a Middle School Music Classroom by Marie Cecile Biggs A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Childhood Education College of Education University of South Florida Major Professor: Susan Homan, Ph.D. James King, Ed.D. Robert Dedrick, Ph.D. Linda Evans, Ph.D. Date of Approval: October 26, 2007 Keywords: singing, rereading,fluency, early adolescents, alternative text, middle school, embedded literacy Copyright 2007, Marie Cecile Biggs

PAGE 2

DEDICATION I would like to dedicate this disser tation to my husband, Allen, for his constant love, patience, and support. I know that it would have been very difficult for me to complete this scholarly jour ney without your continued faith in me. Thank you, Allen, for being such a w onderful husband and companion. I look forward to a long wonderful life with you. I would also like to dedicate this di ssertation to my ch ildren, Nicole and Coleen and my little guy Jaiden. Nicole, thank you so much for your love and support. Your words of encouragement will stay with me forever. Coleen and Jaiden thank you both for reminding me of the importance of my family. I love you all so very much.

PAGE 3

ACKNOWLEDGMENTS I have many people to thank for th eir continued suppor t throughout this scholarly journey. I am most grateful to my committee members, Susan Homan, James King, Robert Dedrick, and Linda Ev ans for their expert guidance on this project. Many thanks go to Dr. Homan fo r the opportunities that she has afforded me over the last few years. To Drs. Ki ng and Dedrick thank you for your insights and suggestions that helped to clarify this manuscript. I would also like to extend a warm thank you to Dr. Evans for de monstrating through her actions that research is about the chil dren. Thank you all. In addition, I would also like to thank Susan Bennett ABD, for her continued support and friendship. I don’t know what I would have done without all your help and support over the last few y ears. I am so very grateful to have you in my life. Thank you

PAGE 4

i TABLE OF CONTENTS LIST OF TABLES.............................................................................................................vi LIST OF FIGURES.........................................................................................................viii ABSTRACT....................................................................................................................... ix CHAPTER ONE: INTRODUCTION..................................................................................1 Statement of the Problem.......................................................................................11 Purpose of the Study..............................................................................................12 Research Questions................................................................................................13 Significance of the Study.......................................................................................14 Limitations of the Study.........................................................................................16 Definition of Terms................................................................................................17 Organization of the Manuscript.............................................................................19 CHAPTER TWO: REVIEW OF LITERATURE..............................................................20 History of the Middle School.................................................................................21 History of the Junior High..........................................................................21 Evolution of the Middle School..................................................................24 Early Adolescent Learner Development.....................................................26 The Middle School Teacher........................................................................32 History of Content Reading ..................................................................................35 Instructional Practices.................................................................................38 Skills and Strategies .................................................................................. .40 Comprehension .......................................................................................... 41 Textbooks............................................................................................... .....42 The Learner and Content Reading..............................................................43 Summary................................................................................................. ....44 Current Complexities of Literacy Learning...........................................................47 The Early Adolescent Literacy Learner ....................................................47 School Structure.........................................................................................50 Mandates and High-Stakes........................................................................51 Summary....................................................................................................56 Recommended Effec tive Strategy Practice for Early Adolescent........................58 Sociocultural Influences............................................................................58

PAGE 5

ii Peer interactions...........................................................................59 Effective Instructi onal Strategies .............................................................61 Direct E xplicit Comprehension Instruction..................................62 Strategy Instruction Embedded in the Content.............................64 Diverse Texts ...........................................................................................66 Technology...................................................................................67 Motivation and Engagement......................................................................70 Fluency................................................................................................ .......72 Repeated Reading........................................................................75 Singing .....................................................................................................78 Current Study.............................................................................................81 Summary....................................................................................................82 CHAPTER THREE: METHOD........................................................................................84 Purpose of the Study and Research Questions.......................................................84 Research Design.....................................................................................................86 Quantitative Phase....................................................................................87 Qualitative Phase......................................................................................89 Mixing the Methods..................................................................................90 Research Context.................................................................................................... 91 School Site................................................................................................91 Music Classroom......................................................................................9 4 Participants........................................................................................................ ......97 Sample Design...........................................................................................97 Sample Characteristics.............................................................................101 Ethical Considerations.........................................................................................107 Instruments........................................................................................................ ...108 Qualitative Reading Inventory-4..............................................................108 Validity and Reliability.................................................................108 Scoring.........................................................................................113 Florida Comprehensive Assessment Test................................................118 Validity and Reliability.................................................................118 Reliability of the Data............................................................................................12 1 Internal Consistency.................................................................................12 1 Interrater Reliability.................................................................................122 Procedures........................................................................................................... ...123 Pre-Study Involvement............................................................................123 How the Students Use the Tune Into Reading.........................................127 Control Group.......................................................................................... 129 Data Collection....................................................................................................... 130 Quasi-Experimental Design Data Collection...........................................130 Interpretive Case Study Design Data Collection....................................131 Data Analysis......................................................................................................... .135 Quasi-Experimental Design Data Analysis.............................................135 Question 1.............................................................................. ...136

PAGE 6

iii Question 2.............................................................................. ...138 Interpretive Case Stu dy Design Data Analysis.......................................139 Credibility of the Data.............................................................................................141 Second Observer......................................................................................14 2 Analysis Check........................................................................................142 Concurrent Triangulation Strategy..........................................................143 Integration of the Data ............................................................................144 Summary............................................................................................................... ..144 CHAPTER FOUR: FINDINGS.......................................................................................145 Question One: Findings for the Treatment and Control Groups..........................145 Results........................................................................................................... ......146 Fluency (WPM) at the Same Reading Level..........................................150 Word Recognition (WR) at the Same Reading Level.............................151 Comprehension (COMP) at the Same Reading Level............................152 Highest Reading Level............................................................................154 Summary of Findings Question 1...........................................................155 Question Two: Findings for the FCAT Level Groups.........................................157 Fluency (WPM) for FCAT Levels 1-5 on...............................................158 Word Recognition (WR) for FCAT Levels 1-5 .....................................163 R eading Comprehension (COMP) for FCAT Levels 1-5.......................167 Highest Reading Level on WPM for FCAT Levels 1-5.........................171 Highest Reading Level on WR for FCAT Levels 1-5............................173 Highest Reading Level on COMP for FCAT Levels 1-5.......................174 Highest Reading Level on RL for FCAT Levels 1-5 .............................175 Summary of Findings for Question 2......................................................176 Question Three: Qualitative Findings for Peer Interactions...............................178 Overview........................................................................................................... ...179 Participants........................................................................................... ....179 The Role of the Researcher..................................................................................180 Theoretical Considerations..................................................................................180 Sociocultural Theory................................................................................180 Peer Interactions...................................................................................... .181 Assessments........................................................................................................ .183 FCAT Reading Level Scores...................................................................184 Fluency: Absence of Prosodic Reading...................................................187 Students Self-Reports: Reading Disposition............................................188 Treatment and Control Data Analysis...................................................................190 Integration of the Data............................................................................194 Treatment Group..................................................................................................19 4 Description of the Computer Lab..........................................................194 Data Analysis...................................................................................................... .195 Group Dynamics....................................................................................196 Safe Risk-Free Environment.....................................................196 Peer Observation.......................................................................197

PAGE 7

iv Peer Hierarchy..........................................................................199 Peer Support..............................................................................200 Motivation................................................................................................201 Extrinsic Motivation...................................................................201 Autonomy...................................................................................202 Intrinsic Motivation....................................................................203 Singability vs. Readability........................................................................204 Students’ Perspectives of Alternative Text.................................205 Disequilibrium............................................................................207 Summary of Results for the Treatment Group..........................................208 Control Group....................................................................................................211 Description of Classroom Routine............................................................211 Data Analysis.................................................................................................... .211 Engagement..............................................................................................213 Alternative Approach to Singing...............................................213 Extrinsic Motivation..................................................................214 Group Formats..........................................................................................2 15 Dominate and Vulnerable Peers.................................................215 Peer Leaders...............................................................................216 Reading Strategies....................................................................................21 8 Fake Rereading..........................................................................218 Disengaged.................................................................................219 Summary of Results for Control Group.....................................................220 Cross Case Analysis.............................................................................................222 Social Systems.........................................................................................222 Peer Positions............................................................................. ..222 Instructional Expectations............................................................223 Alternative Approaches...............................................................224 Reading Strategies......................................................................224 Summary of Cross Case Findings............................................................225 Chapter Summary................................................................................................226 CHAPTER FIVE: DISCUSSION....................................................................................229 Summary of the Study.........................................................................................229 Discussion: Conclusions and Implications.........................................................236 Addre ssing Early Adolescents Diffe ring Fluency Development.............236 Assessing Fluent Readers........................................................................238 The Role of Prosody in Reading Fluency................................................243 Sociocultural Interactions........................................................................250 Contributions of the Study...................................................................................252 Recommendations for Practice............................................................................254 Fluency Instructions..................................................................................254 Assessing Fluent Middle School Readers ................................................256 Sociocul tural Interactions and In fluences on Instruction..........................256 Recommendations for Future Research...............................................................258

PAGE 8

v REFERENCES................................................................................................................260 APPENDICES.................................................................................................................282 Appendix A. Qualitative Reading Inventory-4 ...................................................283 Appendix B. Matrix for Peer Interactions............................................................286 Appendix C. Construct Key ................................................................................287 ABOUT THE AUTHOR...........................................................................End Page

PAGE 9

vi LIST OF TABLES Table 1. Middle School Timeline...................................................................................46 Table 2. Ethnic Enrollment at School Site......................................................................92 Table 3. Percentages of Stude nts’ Classification Variables..........................................103 Table 4. Group Percent of Student FCAT Level Scores...............................................105 Table 5. Summary of Descriptiv e Statistics from QRI-4 Pretests................................106 Table 6. Correlation of the QRI-4 Sc ores and Standardized Tests Scores...................112 Table 7. Determining Instructi onal Reading Level from the QRI-4.............................117 Table 8. Schedule of Pretests for the Treatment and Control Groups..........................131 Table 9. Summary of Descriptive St atistics at the Same Reading Level......................147 Table 10. ANOVA Table for Treatment and Control at the Same Reading Level.........149 Table 11. Changes in Inst ructional Reading Levels.......................................................155 Table 12. Interactions and Effect Sizes..........................................................................156 Table 13. Descriptive St atistics on Fluency (WPM) for FCAT Levels 1-5....................158 Table 14. ANOVA Table on WPM for FCAT Levels 1-5.............................................160 Table 15. Descriptive Statistics on Word Recognition (WR) for FCAT Levels 1-5......163 Table 16. ANOVA Table on WR for FCAT Levels 1-5.................................................165 Table 17. Descriptive Sta tistics on Reading Comprehension for FCAT Level 1-5........167 Table 18. ANOVA Table on COMP for FCAT Levels 1-5............................................169 Table 19. Descriptive St atistics on WPM at the Highest Reading Level.......................172

PAGE 10

vii Table 20. Descriptive St atistics on WR at an Highest Reading Level............................173 Table 21. Descriptive St atistics on COMP at an Highest Reading Level ......................174 Table 22. Descriptive St atistics on RL at an Hi ghest Reading Level.............................175 Table 23. Effect Sizes for WPM, WR, COMP, and RL by FCAT Level Groups........177 Table 24. Constructs from Treatment Group................................................................195 Table 25. Themes from Treatment Group....................................................................196 Table 26. Constructs from Control Group....................................................................212 Table 27. Themes fr om Control Group...........................................................................213

PAGE 11

viii LIST OF FIGURES Figure 1. Diagram of Mixed Method s Concurrent Design and Procedures....................91 Figure 2. Floor Plan of the Music Classroom..................................................................97 Figure 3. Fry Readability Graph....................................................................................126 Figure 4. Qualitative Observation Schedule..................................................................132 Figure 5. An Example of Matrix Used for Peer Interactions.........................................135 Figure 6. Interaction Graph on Fluenc y (WPM) at the Same Reading Level................150 Figure 7. Interaction Graph on Comprehension (COMP) ............................................153 Figure 8. Interaction Graph on WPM for FCAT level At in Reading...........................161 Figure 9. Interaction Graph on COMP by FCAT Levels...............................................170

PAGE 12

ix Reading Fluency Through Alternative Text: Rereading With an Interactive Sing -to-Read Program Embedded Within a Middle School Music Classroom Marie Cecile Biggs ABSTRACT Singing exaggerates the language of readi ng. The students find their voices in the rhythm and bounce of language by using music as an alternative text. A concurrent mixed methods study was conducted to investigate th e use of an intera ctive sing-to-read program Tune Into Reading (Electronic Learning Products, 2 006) as an alternative text, embedded within a heterogeneous mu sic classroom. Measured by the Qualitative Reading Inventory-4 (QRI-4) (Leslie & Caldwell, 2006) the fluency, word recognition, comprehension, and instructional reading leve l of the treatment students were compared to their counterparts who sang as part of the regular music program. Concurrently, this investigation also provided a de scription of the peers’ interac tions during the literacy task assigned by the music teacher. The intent of this study was to addr ess the following three research questions. First, wh at is the difference in read ing outcomes for students who used the singing software ve rses the students who sang as part of their regular music curriculum? Second, are the reading outcomes di fferent when the students were grouped by FCAT reading levels? Third, how do the peer s interact during the literacy task of singing to read? The fi rst two questions addressed the quant itative phase of this study to assess the collective differen ces on the dependent variable s overtime and by group. The

PAGE 13

x qualitative phase in this study used an inte rpretive case study approach to describe peer interactions during the a ssigned literacy task. The study findings suggest that rereadi ng through singing, usi ng the interactive singing program, Tune Into Reading, was more effective regardless of the reading levels for treatment students compared to control students. In addition, prosody appeared to have a direct connection to reading comp rehension. Furthermore, the use of the interactive program provided oppor tunities for differentiated reading level achievement. Finally, group dynamics highly influenced the early adolescent’s motivation, engagement, participation, and successf ul outcomes in reading fluency.

PAGE 14

1 CHAPTER ONE: INTRODUCTION Prominent in educational discourse is understanding and meeting the unique and differentiated needs of the early adolescent literacy learner. This is extremely important as these students prepare to meet the challenges of living in an informational age as fluent, active, and independent readers (Alvermann, 2001; Guthrie & Wigfield, 2000; Kamil, 2002). However, less than a decade ago, this population of learners suffered from scant attention to their literacy learning as “policy makers, curri culum developers, and school leaders rallied to address the literacy needs of students in grades K-3” (Elish-Piper & Tatum, 2006, p. 6). As a result, this placed the specialized literacy needs of the early adolescent at a disadvantage. The 2005 Na tional Assessment of Educational Progress (NAEP) reading results have shown improve ment in literacy achievement for the elementary level. These reading improvement s however, have not necessarily translated to early adolescent literacy learners as once developmenta lly, cognitively, contextually, or instructionally assumed (Biancarosa & Snow, 2006). Specifically, the developmental stance that assumes instructiona l practice for this population of literacy learners all have made the cognitive shift from learning to r ead to reading to learn. However, as these students navigate their liter acy learning across various cont ent areas and through diverse and alternative texts, it should not be assume d they are fluent readers and comprehenders prepared to meet the challenges of the new millennium. Early adolescence, typically defined as ages 10-14 (middle school years), is a time of transition and rapid change in the students’ emotional, social, physical, and cognitive

PAGE 15

2 development (Cottle, 2001; Moje, Young, R eadence, & Moore, 2000; Pikulski, 1991). Developmentalists, following the work of G. Stanley Hall (1908), continue the debate that early adolescents are neither children nor fully mature adolescents. Instead, they are caught in the developmental tensions of adolescence (Bean & Brodhagen, 1996). These tensions, which parallel the onset of this developmental stage, can become even more daunting when the early adolescent student enters the contextual environment of the middle school. At this level, more cognitive strategic demands in reading are placed upon the students to comprehend divers e texts (Alvermann & Phelps, 2005). The complex process of comprehending te xt is the ultimate goal of reading. Alexander (1998) believes this is extremely difficult for ea rly adolescents because their cognitive strategic processes in reading are very diverse a nd are under continual development. Even though early adolescen ts are situated within a particular developmental stage, their cognitive abilities in reading vary with the different literacy tasks presented. Jetton and Alex ander (2000) suggest, early adolescent readers’ use of text comprehension strategies range acro ss a developmental continuum, and there is interplay of prior knowledge, experience, a nd strategic processes. Therefore, an adolescent reader may be a competent fluent re ader in one literacy task and yet fall back and need support in another task. Ivey (1999) in her case study of three sixth grade students of varying reading ab ilities, found that middle school readers were complex and multidimensional in their reading. These comp lexities may become more pronounced as the middle school reader enters the context of middle school. Th is may affect their ability to read fluently within and across various content areas.

PAGE 16

3 Content area teachers assume it is their re sponsibility to cover their subject matter in a timely, accurate, and effective manner (Alvermann & Moore, 1991). The cognitive shift from learning to read to reading to le arn is assumed to occur before students leave elementary school. Dole, Duffy, Roehler, and Pearson (1991) found, this assumption supported through the pedagogical lens of the middle school c ontent area teacher. Middle school content area teachers often incorrec tly believe that by the time most of their students enter their classrooms they are flue nt readers. Therefore, they may believe incorporating strategic approaches towards fluency in re ading are not needed for this population of learners. To further complicate this contextual dilemma, middle school content teachers have resisted the recommendation to incorpor ate literacy-related in struction into their curricula (Phelps, 2005). Biancarosa and Snow (2006) suggest that content area teachers should be encouraged to provide literacy skills and strategies that are embedded in their content area. By emphasizing the literacy practices that are sp ecific to their subject area, they can maintain the integrity of the content while providing strategi c literacy instruction to comprehend and to be fluent with the sp ecific concepts being taught (Alfassi, 2004). However, Bulgren, Schumaker, Deshler, Lenz, and Marquis (2002) report, content teachers feel they do not have the time or experience to include explicit literacy instruction into an already cr owded curricula. This may be a result of deeply embedded values, beliefs, and practices, and the need to conform to stringent standards imposed by the No Child Left Behind Act of 2002 (O’Brien, Stewart, & Moje, 1995). The problem looms even greater in this er a of standards-based reform one that calls upon educators to meet these standards, to teach to these sta ndards, and to have

PAGE 17

4 these standards evaluated through annual high -stakes testing (Frenc h, 2003). The results of this yearly assessment can have a drama tic impact on the early adolescent literacy learner with the possibility of retention, class placement, and specifically, instructional practices provided to the student s. Rothstein (2000) questions whether an annual test of a student’s knowledge, at just one point in time can provide an accurate assessment of this population of literacy learners. The score obtai ned from this high-stakes test place the early adolescents below, at, or above their classmates in r eading, and it is assumed that the early adolescent students who may or may not have passed the test will receive the instructional strategies needed to prepare th em to be fluent readers and comprehenders. This narrow focus places the literacy needs of this popul ation of learners at a disadvantage as they prepare to become pr oductive citizens in our larger world (French, 2003; Sackes, 2000). Currently, this cognitive stance integrates developmental and contextual considerations and is supplemented with an appreciation of the socialcultural influences that shape instruc tional practices for these lite racy learners (Phelps, 2005). Specifically, the social interactions (e.g. talk, peer modeling, or social reinforcement) of the early adolescent peer groups, that blends their diverse backgr ounds and experiences during the literacy task (r ereading through singing), occu rring within the cultural environment of the classroom. During early adolescence, the peer group becomes a prominent context for development (Brown, 1990).The school and cl assroom provides opportunities for peers to interact throughout the day. Ryan (2000) reports “peer interactions consume significantly more time in adolescence compared to chil dhood” (p. 107). These inte ractions with peers can concern both academic (e.g., achievement) and nonacademic matters (e.g.,

PAGE 18

5 engagement, motivation, self-efficacy, and inte rest). Ryan (2000) suggests, there are generally three ways that ear ly adolescents experience peer interactions within the context of middle school: through informati on exchange, modeling, and peer pressure. Information exchange occurs when adoles cents have a discussi on with their peers (Berndt, 1999). In an experimental study with eighth-grade students, Berndt, Laychak, and Park (1990) found that when adolescents had to make an academic decision such as attend a rock concert or study for a test, they initially responded differently from one another. However, after discussing this dile mma with their peers, their answers were similar to their peers. This form of inte raction could influence the early adolescent’s choice to partake in the literacy task presented by the teac her if it was used effectively. Modeling is another form of adolescent peer interaction. This interaction refers to individual changes in cognition, beliefs, or affect, which are a result of adolescents observing their peers (Ryan, 2000) Observing a specific behavi or performed, or listening to a peer voice a certain belief, can induce an adolescent to adopt such behaviors or beliefs. Schunk and Zimmerman (1996) reported, peer modeling influences self-efficacy beliefs. In their study, they f ound that early adolescents who verbalized that they had difficulty with a task and then observed their peers have success with the same task then believed they could complete the task. The ear ly adolescent, when f aced with a literacy task, may have success by observing their peers. Peer pressure is a third way that the early adolescent interact s with their peers. Peer pressure takes on the role of so cial reinforcement (Ryan, 2000). Brown, Lohr, and Eicher (1986) found that beliefs and behaviors that are discouraged by the groups are not likely to be displayed, whereas beliefs and behavior s that are positively

PAGE 19

6 received by the group are more likely to surface. Therefore, participation in the literacy tasks that the peer group positively received th rough this social interaction, could have a positive effect on the group’s beliefs and deci sions to participate by the group members. The field of reading has moved far beyond the view that literacy is the ability to read and write across various content areas alone (Bean, 2000). Instead, the concept of content reading has been broadened to reflect the integration of communication processes (reading, writing, talking, listening, and viewing) as the students engage in text–related learning (Alfassi, 2004; Lenz & Deshler, 2003; Vacca & Vacca, 2002). There is the assumption that the linear textbook is necessa ry for teaching and learning the content specifics (Wade & Moje, 2000). It is this assumption that influences instructional delivery and perceptions of fluent, active, and independent readers (Alvermann, 2002). However, Phelps (2005) reports, alternative texts that focus on new literacies through digital media have had a great influence on the early adolescent’s inst ructional practices. The computer offers students more cont rol in terms of support, pace, and active processing of text (Kamil, 2002). The use of t echnology as an alternat ive text links real world experiences and interests, and provide s opportunities for alternative text reading with the early adolescent literacy learner. Le u (2000) reports on the positive effects for middle school readers when print and visual texts (e.g., hypermedia, the internet, and interactive CD-ROMS) are ut ilized. Reading diverse text s across and within various content areas can be further complicated if early adolescent students do not have the background knowledge, experiences, and strategies for reading a variety of texts fluently. Fluency is a necessary aspect of successful reading as it allows readers to read with speed, accuracy, and proper expression (N ational Reading Panel, 2000; Rasinski,

PAGE 20

7 2004). For years, teachers thought, if students could learn to decode words accurately, they would be successful in reading printed te xt. The assumption is often made that early adolescents are at a satisfact ory level of fluency in reading. However, according to Alvermann and Phelps (2005), this is not alwa ys the case, specifica lly with content area materials. While it is true that accuracy in students’ ability to decode words is important for fluency, decoding needs to be automatic However, automatic decoding for fluent reading is not sufficient. Rasinski (2004) poi nts out the need to connect accuracy and automaticity to reading prosody. Reading prosody is the point where fluenc y connects fluent decoding directly to comprehension (Rasinski, 2004). The prosody co mponents of reading fluency address the use of phrasing and expression (Dow hower, 1987, 1991; Schreiber, 1980, 1987, 1991; Schreiber & Read, 1980).When readers adjust appropriate volume, tone, emphasis, phrasing, and other elements while readi ng aloud, they are providing evidence of comprehending text (Rasinski, 2004). In this se nse, fluency is a multifaceted event with reading comprehension as the goal. Through guided and repeated reading, bot h prosody and decoding (automaticity and accuracy) in word recognition are deve loped. Samuels (1979) defines repeated reading as a fluency-building st rategy that consists of timed rereading of a short passage several times (at least 3 times), checking fo r accuracy ( word recognition), automaticity (words per minute) and with prosody (expres sion) The steps for an effective fluency instructional model are: (a) provide a model for students e xpressive fluent reading, (b) give the students a passage to read (approximately 150 words) 3 times at their instructional reading level (word recognition with 90-95 % accuracy), and (c) have the

PAGE 21

8 students orally read the pa ssage assessing for accuracy automaticity, and expression (Rasinski, 2004). The National Reading Panel (2000) found su fficient evidence that guided oral reading done through repeated reading w ill have a positive impact on fluency and comprehension across a range of grades levels and in a variety of general and special education classrooms. Rasinski (2004) contends that reading fluency is a “bridge between two major components of readingwording decoding and comprehension. At one end of the bridge, fluency connects to accuracy and automaticity in decoding. At the other end, fluency connects to comprehension through pros ody, or expressive inte rpretation” (p. 1). Repeated reading is most authentic when the practiced material is eventually performed orally, such as plays, poetry recita tion, or in this study singing lyrics to songs (Rasinki, 2004; Stayter & Allington, 1991). This form of repeated exposure through singing assists the reader with fluency through prosodic reading. The reader uses appropriate volume, rhythm, pitch, tone and phrasing (prosody), while singing the song lyrics, and therefore, they giving evidence of actively constructing meaning from the passage (Rasinki, 2004). Singing as an alternative text can build reading fluency and comprehension and can be naturally embedded within the music content classroom. Butzlaff (2000) contends that there are similar characterist ics with singing instruction and the reading process: (a) music text and written text involve formal written notations that are read left to right, (b) the sensitivity to phonological distinctions and word recognition requires a sensi tivity to pitch and tonal distin ctions in both reading and singing, (c) when students learn the lyrics to songs they ar e engaging in reading, and (d) learning song lyrics is often repetitive, so th at rereading of text o ccurs through singing.

PAGE 22

9 Hall, Boone, Grashel, and Watkins (1997) suggest that students should sing independently, on pitch, and with rhythm. While most singing in the music classroom is done in groups, minimal time is spent w ith students singing individually, making it difficult to assist each student to develop these specific faculties. Along with singing independently, Levinowitz ( 1989) suggests that student s would sing songs more accurately with copies of individual text than without. However, singing in the music classroom is usually performed as a whol e group with one song and one group text. Usually the text is displayed on an overhead or chart, regardless of the variety of instructional reading levels of the student body. Currentl y, the use of an individual computer program could address these concerns. Individualized computer assisted training in the music classroom is a recent additional tool teachers can employ for st udents to learn to sing and acquire songs individually. In a study analyzing 150 empiri cal articles on comput er applications in music learning, Webster (2002) reported generally positive results with singing performance and pitch accur acy; however, studies on song acq uisition with software for students in the middle school setting are spar se, especially studies relating singing to reading. One report on the computer program Carry-A-Tune (Educational Learning Products, 2004) is in publication to date. This was a pilot study to examine the use of the sing-to-read software program with remedi al reading middle school students (Biggs, Homan, Dedrick, Minick, & Rasinski, in press). Carry-A-Tune is an individual computer product originally developed to improve singing. The program uses a vocal range anal yzer that tracks th e singer’s pitch and rhythm, comparing it to the correct pitch of the song. E ach student uses a microphoned

PAGE 23

10 headset linked to the computer to sing along repeatedly and to record their singing. There is a great need to investigate the effects of this and other computer singing programs, especially if the potential exists that they coul d be as helpful to music teachers as it seems to be for reading instructors and their student s. The current study inve stigated the use of an individualized interac tive sing-toread program Tune Into Reading (TIR) (Electronic Learning Products, 2006), adapted from Carry-A-Tune, as an alternative text embedded in the middle school music classroom curriculum. Tune Into Reading, not unlike it predecessor Carry-A-Tune, has several unique features that can be used to meet the sp ecialized needs of this population of literacy learners. In both programs each student uses an individual soundproof microphoned headset for listening, singing, and recording. This provides real time pitch recognition and feedback to the user. The inclusion of pitch recognition is important because Lamb and Gregory (1993) found that p itch discrimination is signifi cantly correlated (.77) with reading ability. In Tune Into Reading as was the case with Carry-A-Tune the scoring mechanism (pitch accuracy scores 0-100) ac commodates each individual’s vocal range, and contains a portfolio sign-in menu that a ligns with the custom vocal range of each participant. However, Tune Into Reading gene rates reports that print pitch scores for the individual student and/or the class, wher eas Carry-A-Tune did not. In addition unlike Carry-A-Tune Tune Into Reading provides individual folders for each participant. As soon as the participant sign into the program and clicks on the My Lesson folder they have access to the songs that are at their in structional reading level. Also, while both programs had songs analyzed for readability levels, Tune Into Reading has over 200 hundred songs, whereas Carry-A-Tune had only 24 songs. The songs range from first to

PAGE 24

11 tenth grade readability levels. This wide ra nge of available reading levels will provide opportunities for the students to build fluenc y through repeated reading by singing songs at their individual instructional reading level. The literature on reading fluency often fo cuses on the beginning reader’s initial stage of literacy acquisition or on the older adolescent reader who has difficulty learning to read. This focus has placed reading fluenc y in a deficit view, rather than creating a direct link to comprehension (Clay, 1985). St ayter and Allington (1991) suggest that “we have failed to consider some of the broader ramifications of an emphasis on fluency, especially with older and more develope d readers” (pp.143-144). Especially when fluency instruction could support both the st ruggling and more developed reader’s, as they transition to the context of middle sc hool, navigating their literacy learning, across various content areas and though divers e and sometimes difficult texts. Statement of the Problem This study examined how the use of sing-to-read program Tune Into Reading as an alternative text might support literacy learning of early adolescents and thereby, improve their fluency(word per minute), word recognition (accuracy in oral reading), comprehension (implicit and explicit questions after reading), and instructional reading level (combined sores of accuracy and comprehension). A majority of early adolescents need opportunities and instru ctional support to read varied and diverse materials in order to build their experiences, fluency, and range as readers (Kuhn & Stahl, 2003). Literacy learni ng should take into account developmental issues, as well as thoughtful and critical lite racy expressions that embrace the multiple literacies that these students bring to school within and across various content areas

PAGE 25

12 (Kamil, 2000). The problem is that little is known about this population of literacy learners and about how to provi de literacy instruction that wi ll address this change while, at the same time, providing support for thei r social and academic needs (Alvermann & Phelps, 2005). In order to gain a perspe ctive on the impact that these assumptions have on middle school readers, it is appropriate to examine these student s within a music classroom, to investigate singing as a form of repeated reading to improve fluency. This study investigated a population of middle schoo l students who are in a music classroom as part of their assigned yearly elective cycl e. Examining this sample will provide better insights into the area lacking in the availabl e literature – the possibility of providing effective literacy instructi on through alternative text embedded in music content area instruction. Purpose of the Study The purpose of this study was to investigat e the use of an inte ractive sing-to-read program Tune Into Reading as an alternative text, embedded within a heterogeneous music classroom. This investig ation also provided a descriptio n of the peers’ interactions during the literacy task assigned by the mu sic teacher. This study used a concurrent mixed methods design. The intent of the st udy was to address the following research questions: Quantitative Research Questions 1. To what extent is the reading perfor mance of word rec ognition, fluency, comprehension, and instructional reading level, as measured by the QRI-4, of

PAGE 26

13 students using the Tune Into Reading program different from their regular music curriculum counterparts? 2. To what extent does the Tune Into Reading program differently impact the reading scores of students who are “below, at, or above” grade level as determined by the Florida Comprehensive Assessment Test (FCAT) reading scores? Qualitative Reading Question 1. How do middle school readers interact with their p eers, within the context of their music classroom? The first quantitative research questi on addressed the readers’ use of the interactive sing-to-read program Tune Into Reading as an alternative text. Prior to the treatment, I administered a pretest using the QR I-4. Scores from the pretest were used to ensure that the students in the control and e xperimental groups were not different in their performance in fluency (measured by words per minute), word recognition (measured by oral reading accuracy), comprehension (measured by implicit and explicit questions after the reading), and instructiona l reading level (measured by combined sores of accuracy and comprehension) before implementation. Af ter the implementation of the interactive singtoread program, Tune Into Reading I administered a posttest using the QRI-4 and compared the posttest scores with the pretes t scores to determine if students in the experimental group had gained significantly ove r their counterparts in the control group. The second quantitative research questi on investigated whether there is an interaction effect of the repeated reading methods us ing the sing-to-read program, Tune Into Reading, as an alternative text on the reading performance of the students when they were grouped as “below, at, or above” gr ade level as determined by the Florida

PAGE 27

14 Comprehensive Assessment Test (FCAT) 2006 in reading. The results in reading achievement level scores (achievement leve ls 1-5), according to the state of Florida Department of Education, are reported as: (a ) students who scored a Level 1 or 2, are considered below proficiency in meeting grad e level benchmarks, (b) students scoring a Level 3 are considered at grad e level, and (c) students who scored at a Level 4 or 5 are considered above grade level (FCAT Briefing Book 2005). Concurrently, the qualitative observations were used to probe for significant themes by describing aspects of peer inter actions ( e.g., peer talk, peer modeling, and peer social reinforcement) among students who are singing using th e interactive program Tune Into Reading, versus the peer interactions among students who are singing in the traditional music class. Significance of the Study Currently, although there appear to be emerging themes and important information being investigated about the c ontextual conditions, developmental needs and instructional practices, concer ning reading in the content ar eas, the knowledge base for early adolescent liter acy learners is still limited (Alvermann, 2002; Bean, 2000; Kamil, 2002; Moore, 1996). The National Reading Panel (2000) identified fluency as one of the five critical components of reading (P ikulski & Chard, 2005). Fluency in reading, however, is often thought of as a deficit, remedial tool for word accuracy and automaticity, rather than a direct link to comprehension (Clay, 1985; Stayter & Allington, 1991; Rasinki, 2004). Repeated reading is th e methodology that is mo st appropriate to develop fluency and comprehensi on so that early adolescents can navigate their literacy learning strategically across various content areas. Howe ver, little is known about

PAGE 28

15 repeated reading to build fluency with early adolescent literacy learners of varying reading abilities. Even though the assumption is often made that many early adolescents are at a satisfactory level of fluency in reading, this is not always the case, and specifically, it is not the case with conten t area materials (Alvermann & Phelps, 2005). The standards based reform movement with hi gh-stakes testing has also contributed to the assumption that a middle school student is a fluent reader. The current study will add to the knowledge base important information pertaining to fluency instruction through repeated reading for a range of literacy learners. Along with this cognitive stance and its overlap with the development stage and contextual conditions is an appreciation of the socialcultural influences that shape instructional practices for th is population of literacy lear ner (Phelps, 2005). Specifically, the social interactions (e.g., ta lk, modeling, and social reinfo rcement) of the peer group, blending each member’s diverse background a nd experiences during the literacy task (repeated reading thr ough singing), and occur within th e cultural environment of the classroom. Through observations and descriptions of peer interactions, more information will be provided to the field concerning these in teractions during specific the literacy task presented. In 2004, to help address this population of literacy learners, a panel of five nationally known educational researchers me t with representatives of the Carnegie Corporation of New York and the Alliance for Excellent Education. The focus was to draw up a set of recommendations on how to meet the needs of adolescent literacy learners while propelling the field forwar d (Biancarosa & Snow, 2006). A list of 15

PAGE 29

16 elements were reported and then divided in to two sections: instructional improvements and infrastructure improvements. The instructional elements consisted of: a) direct, exp licit comprehension instruction, b) effectiv e literacy instruction embedded in content, c) motivation and selfdirected learning, d) text base d collaborative learning, e) st rategic tutoring, f) diverse texts, g) intensive writing, h) technology, and i) formative on-going teacher assessments (Biancarosa & Snow, 2006). Researchers were urged to re-conceptualize how they perform research with early ad olescent literacy learners. I nvestigations should combine different elements so that important info rmation about the early adolescent can be determined. This study utilized five of these elements. It investigated early adolescent literacy gains when instruction is embedde d in the music content area. Also, the interactive sing-to-read program Tune Into Reading is delivered through a technological format with a diverse and interesting te xt, which may be motivating and engaging (Guthrie & Wigfield 2002). Most important explicit comprehensi on instruction through rereading to enhance comp rehension was addressed. Limitations of the Study The following list is provided to acknowledge and clarify the limitations of this study that impact the generalizability of the findings: 1. Random sampling of individual students was not an option in this study, and therefore possessed a threat to the ex ternal validity. This limited the generalizability of the findings. To address this threat, random assignment by classes were made. In addition, analys is was conducted to match sample

PAGE 30

17 characteristics including demographics, and reading performance prior to analysis and treatment. 2. Complete FCAT level scores in reading we re unavailable for all participants (four students were missing scores from the trea tment group and four from the control. The researcher acknowledges that missing data might have limited the findings for question two. However, there was an e qual distribution of pe rcentages in each group stratified as Below grade level (FCAT level 1 and 2), At grade level, (FCAT level 3), and Above grade level (FCAT level 4 and 5). 3. The study duration was only seven weeks, ha d it been longer it might of netted different results. 4. The characteristics of the samples were predominantly low SES White males in eighth grade. This limits the findings fo r other sample characteristics. Definition of Terms The following is a list of the terms and ope rational definitions that will be used throughout the study: 1. Active Reader: Readers who engage in an active search for meaning using multiple strategies as they m onitor their understand ing of what they have read (Pearson & Fielding, 1991). 2. Alternative Texts: Various textual formats that are us ed to supplement the linear text or replace the textbook in the conten t areas. Most often they are digital in nature. In this study the alternative texts refer to the inte ractive sing to read program (Alvermann & Phelps, 2005).

PAGE 31

18 3. Early Adolescent Literacy Learner: typically defined as ages 10-14 years (middle school), is a time of transition and rapid ch ange in the students’ emotional, social, physical, and cognitive devel opment (Cottle, 2001; Moje, Young, Readence, & Moore, 2000; Pikulski, 1991). 4. Embedded Literacy in the Content: Literacy embedded in th e content addresses two directions for instructional implementation (Biancarosa & Snow, 2006) First, within the Language Arts classroom these principl es are not discrete skills or techniques instead the emphasis should be how to teach the strategy or skill using other contentarea materials. Second, content area teacher s should encourage literacy skills and strategies that emphasize the reading and writ ing practices that are specific to their subject area (Alfassi, 2004). 5. Fluent Reader: A reader who reads with accura cy, automatic recall, and voice expression, volume and pitch (Rasinski, 2004) 6. Independent Reader: A reader who requires less in the way of structured learning support (Pearson & Gallagher, 1983). 7. Literacy Tasks : Assigned task related to reading and writing given to the students by the teacher. 8. Socialcultural Influences in Literacy: A sociocultural approach to literacy instruction is multidisciplinary and occupies the fields of history, anthropology, linguistics, psychology, and sociology. Sociocultural approaches emphasize the interdependence of social and individual processes in the construction of knowle dge. When viewing literacy development from a sociocultural ap proach, literacy arises from the child’s participation in social activities in which th ere are real reasons to use written language

PAGE 32

19 (Englert & Palinscar, 1 991). In this study the social interactions (e .g., talk, modeling, and social-reinforcement) of the peer group, blending diverse background and experiences during the literacy task (r ereading through singing), o ccurring within cultural environment of the classroom. Organization of the Manuscript This manuscript has been organized into five chapters. Chapter One identified the problem and places it in the context for the study. The research questions, limitations, and definitions are also included. Chapter Two review ed the literature relevant to the research questions. Research strands include (a) Historical Revi ew of the Middle School Movement: The Context, the Learner, and R eading Instruction (b) Current Contextual Conditions: Influence of Standards and Mandates with Literacy Development (c) Effective Practice and Instructional De livery for the Early A dolescent Literacy Learner. Chapter Three presen ted the methods that were us ed to conduct this study. It outlined the research questions, research c ontext, and the participants. In addition it described the design of the study; including ethical considerations, instruments, and procedures. The final sections explained reliability measures and the manner in which the data was collected, analyzed, and interprete d. Chapter Four summarized the findings of the study. The descriptive statistics and findi ngs derived from the data analysis are reported. Chapter Five presented the conclusions of the study, the re sulting implications of the study results, and the recommendati ons for classroom and future research.

PAGE 33

20 CHAPTER TWO: LITERATURE REVIEW The National Assessment of Educational Progress (NAEP) (2005) reports over 73 % of eighth grade students perform below or at a basic level in their reading achievement. Consistent with NAEP results Bi ancarosa and Snow (2006), in their report to the Carnegie Corporation, contend that ove r 70% of adolescents struggle with their reading in some manner and therefore require instruction that is differentiated and strategic. This is alarming as few gain the literacy knowledge n eeded to successfully engage in higher-level problem solving required for an information transforming economy (Donahue, Voelkl, Campbell, & Mazzeo, 1999). In addition, although emerging themes appear and important info rmation in connection to reading in the content classrooms, while addressing the c ontextual conditions, developmental needs, and instructional practices, the knowledge base for early adolescent li teracy learners is limited (Alvermann, 2002; Bean, 200 0; Kamil, 2002; Moore, 1996). How do we prepare our early adolescents to be fluent, active, and independent readers, who meet the literacy demands and ch allenges of living in an informational age? Although this issue poses current complexitie s for adolescent literacy learners, the dilemma of how best to meet the unique need s of the early adolescen t, typically defined as ages 10-14 (middle school years), has been a historical debate for over 100 years. In order to understand the gaps in the literature and how best to currently meet the literacy needs of the early adolescent, I will provide a brief review of the historical background.

PAGE 34

21 This historical context will be helpful to inform current practices. Therefore, this literature review will chronologically address the history of the middle school movement, its overlap with the unique developmental n eeds of its learners and teachers, and the parallels historically with reading in the content areas. This will be followed by a review of the complexity of current practices of content reading embedded in the middle school content areas and the influences of mandates and standards. The final section of this review examines studies and thoughts about effective strategic practices to meet the needs of the higher literacy demands for the future. Brief Historical Review of th e U.S. Middle School Movement: The Context, the Learner, and the Para llel of Content Area Reading Instruction To understand the challenges in today’ s middle school, teaching and instruction can not be separated from the social and institu tional context in which it occurs. To gain a perspective about the context it is important to understand its history (Brodhagen & Bean, 1996). This historical lens allows us to gain an understanding of the current contextual, developmental, and instructional conditions afforded to this population of literacy learners, and suggest how the field s hould move to address their needs. The History of the Junior High School The prominent configuration of education in the 1900’s consisted of eight years of primary school and four years of secondary school. Instructional focus for the early adolescent (grades seven and eight) consisted of a review of the first six years of schooling (Brimm, 1969). There were claims th at the early adolescent’s time was wasted in school with this narrow focus, which resu lted in political and societal pressures to reconfigure the elementary schools (Cuban, 1992).

PAGE 35

22 Educational researchers (Beane, 2001; Brough, 1995; Cuban, 1992; Spring, 1986; Van Til, Vars, & Lounsbury, 1961) agreed th at the suggested reconfiguration of the schools came from societal, political, and academic pressures: (a) influx of immigrations and burdened enrollment at the elementary level, (b) industriali zed period, (training a workforce) and high drop-out rate by 8th grade-which resulted in a workforce of unskilled workers, and (c) preparation for the acad emic rigor of high school and college. Along with the societal and political issu es that impacted the reconfiguration of the elementary schools, there was also a developmental movement taking hold. The National Education Association (NEA) was one of the groups taking a developmental stance for school reconfiguration. The NEA ( 1899) argued for a reconfiguration of the elementary schools and a need to start secondary school at 7th grade rather then 9th grade. In their position statement they argued: [T] he transition from elementary to the secondary period may be natural and easy by changing gradually from the one-teacher regimen to the system of special teachers, thus avoiding the violent shock now commonly felt on entering high school. (p.10) This reform effort was led by an NE A committee member Charles Elliot, then president of Harvard College. In his positi on statement he argued that a better college preparation could be achieved for the ear ly adolescents by extending the secondary school programs downward (Brimm, 1969). However, in 1917 the Smith Hughes Act spearheaded by the National Society for the Pr omotion of Industrial Education, proposed curricula programs that focused on improving the workforce of skilled laborers, specifically agriculture (Brimm, 1969). This two-track system met the societal and

PAGE 36

23 political pressures to train a workforce, while providing rigorous academics for the college bound students earlier, and to ease ove rcrowding conditions at the elementary level. The developmental position taken by the NEA was consistent with the work of the influential psychologist G. Stanley Ha ll (1908), who argued for years that early adolescents were in a unique stage of deve lopment and they should be separated in the context of their schooling from their predece ssors and successors. Hall (1908) contended that early adolescents, if placed in the elementary school, would have a negative influence on the younger children, and if placed in the secondary school, would be negatively influenced by older adolescents (Beane & Brodhagen, 1996). It was therefore recommended by The Committee on the Economy of Time and the Commission on the Reorganization of Secondary Education (1918) that the reorganization of schooling for the adolescent be divided into junior a nd senior high levels for secondary school (Juvonen, Nhuan Le, Kaganoff, Augustine, & Constant, 2004). Although NEA developmental position contri buted to the new configuration of the junior high school, educational historians report that the motivation for this new institutionalized structure was created fo r multiple purposes. Beane (2001) and Cuban (1992) contend societal and political pressures had the strongest influence on the reorganization of the junior high school as a result of the conv erging interests of humanists, societal efficiency advocates, and stage-related developmentalists. Specifically, the issues were related to overcrowding of the elementary school and tracking of students for th e vocational path (workforce) or academic path

PAGE 37

24 (college). Lounsbury (1984) contends this period from 1890-1920 was a struggle between academics and vocations. This tracking path tran slated to the instructional practices and curriculum delivery afforded to the early a dolescent. The students directed towards the vocational path received a very different in structional program of survey academic courses and life skills, as compared to thei r counterparts on the academic path who were afforded coursework with academic rigor (Beane & Brodhagen, 1996). However, in spite of the needs of rapidly change society and the premise that the junior high school would help facilitate this change, only one -third of the students made it to 9th grade, from the early 1900’s to the late 1950’ s (Van Til, et al., 1961). The Evolution of the Middle School Even though the junior high school reconf iguration was not a success (e.g., due to the large numbers of students dr opping out of school), enrollme nt at the elementary level continued to increase. Therefore more junior high schools were built, specifically for space purposes (Alexander & George, 1981). The 1950’s brought about discussion not only pertaining to the uniqueness of the stude nts but also how the instructional programs for this population should be matched to their ne eds. In their analysis of the literature on instructional practice for the early adolescent, Gruhn and D ouglas (1956) synthesized the following goals for the junior high school: integration of skills, interests, and attitudes exploration of inte rests and abilities differentiation of educational opport unities based on student background, interest, and aptitudes

PAGE 38

25 socialization experiences that prom ote adjustment, guidance in decision making articulation that assists youths in ma king the transition from an educational program designed for preadolescents to a program designed for adolescents. (p.12) However, after many theoretical discussions about the unique needs and instructional programs that shoul d be developed for the early a dolescent at the junior high level, the translation of theory to organi zational and instructiona l practices was very similar to that of the senior high school in the 1960’s. Bough’s ( 1995) research reports, that there was “an emphasis on content rather then exploration, departmentalized rather then integration, and adherence to a rigid sc hedule” (p. 38). The junior high’s curriculum and organization assumed similar characterist ics as the senior high school. Brimm (1969) contends “The very name, “junior high school,” was pointed to as a serious obstacle in the development of a special program for the early adolescent” (p. 8). These challenges created obstacles for the reformers to meet th eir goals of: (a) schooli ng that addresses the unique developmental needs for the early a dolescent students, and (b) preparation for their future, whether it would be work or college. A growing concern and dissatisfaction ex isted during this time period for the contextual conditions afforded to the ear ly adolescent. While the secondary school enrollment dropped, the elementary leve l of school expanded. The 1960’s brought another wave of political talk to change the junior high to middle school (Cuban, 1962). The goal was to match instructiona l practices to meet the needs of these young learners.

PAGE 39

26 However, Alexander (1968), in his surv ey research, with a stratified random sample of 110 reorganized middle school pr incipals, found that 58 % of the respondents reported that middle schools were developed to eliminate overcrowding of the elementary school, while 42 % said programs were needed to meet the developmental needs of the early adolescent. Ten years la ter Brooks and Edwards (1978 ) conducted a replication of Alexander’s study and found that 42 % of the principals suggested the same reason to eliminate overcrowding, whereas 58 % reported to have a program designed to meet the developmental needs of the ea rly adolescent. Cuban (1992) c ontends it is evident “that the mix of stated motives echoes the variety of reasons given by prom oters of junior high schools at the turn of the cen tury” (p. 243). Specifically, the reconfiguration of the school context because of over crowdi ng conditions at the elementary level, and developmental needs of the early adolescent learner. The Early Adolescent Learners Developmental Needs Research on the developmental character istics of the early adolescent learner was crucial to the reconfigurati on of the junior high and late r to the middle school. Hall’s (1908) work in his book, Adolescence: Its Psychology and Its Relations to Physiology, Anthropology, Sociology, Sex, Crime, Religion and Education, portrayed early adolescence as a period of turmoil and stress. He contends this period is a result of the biological and psychological ch anges that occur. He argue d schooling for these students should be separated from their predecessors and successors because they had unique developmental needs brought on by puberty:

PAGE 40

27 This child is driven from his paradise and must enter upon a long viaticum of ascent, must conquer a higher kingdom of man for himself, break out a new sphere, and evolve a more modern stor y to his psychophysical nature. (p. 71) Tanner’s (1962) research was not unlike Ha ll’s, showing a decline in the average age of puberty for the early adolescent. He found early adolescents were experiencing puberty earlier, approximately 4 months ear lier each decade from the 1900-1960’s. These results were used to help justify the rec onfiguration that resulted in the move of 6th graders to the middle school and 9th graders to the high school level. Eichhorn (1966) coined the term “transescence” as the deve lopmental stage of early adolescents. He defines it as: The stage of development, which begins prior to the onset of puberty and extends through the early stages of adolescence. Since puberty does not occur for all precisely at the same chronological age in human development, the transescent designa tion is based on the many physical, social, emotional, and intellectual changes in body chemistry that appear prior to the time, which th e body gains a practical degree of stabilization over these comple x pubescent changes. (pp. 3-4) Eichhorn (1973), Havighurst, (1972), a nd Tanner (1962), all argued that students should be grouped according to their developm ental stages and not their chronological age. However, cognitive developmenta lists did not take this stance. The cognitive development of the early adolescent during the middle school movement was defined using a Piagetia n framework (Beane & Brodhagen, 1996). The developmental theory of cognition proposed by Piaget (Inhelder & Piaget, 1958) was on

PAGE 41

28 the emergence of formal logical structures and was not specifically related to the uniqueness of the adolescent. According to th is cognitive framework the early adolescent was at the concrete operational stage, a fo rmal operational stage of development, or between the two (Beane & Brodhagen, 1996). In particular, their thinking was shifting from concrete understanding to more abst ract and higher-order reasoning. Like his predecessor (Hall, 1908), Piaget too was concer ned with the developmental unique stage fit of the early adolescent, and should r eceive instruction appropriate with their developmental stage. However, Piaget’s theo ry was deficient concern for a broader array of biological, emotional, social, and societ al concerns engaged in other theorists’ discussions (Beane & Brodhagen, 1996). A new paradigm in the 1980’s middle school reform movement impacted this population of learners. A call to society’s lack of atte ntion was brought to focus by Lipsitz’s (1980) book Growing up Forgotten which stated that the early adolescent was generally underserved and that education should address the “whole child.” The focus should not only include an understanding of the development stage for the early adolescent, but also an unders tanding of the social relations hips and affective conditions that influenced this population of learners. Johnson, Markle, and Stingley’s (1982) research investigated how peer acceptance was related to academic achieveme nt. Greenberg, Siegel, and Leitch (1982) studied adolescents’ attachme nts to their parents and p eers using a newly developed psychometric instrument Inventory of Adolescent Attachments that measures self-esteem and life satisfaction of relationships with parents and peers Using a hierarchical regression model with 213 ear ly adolescents (ages 10-14) the researchers contend

PAGE 42

29 attachments were more powerful with parents than peers in measures of well-being and self-esteem. Another line of studies focused on the a ffective issues of the early adolescent literacy learners. Mager (1968) and Rosenshine (1980) provided data that suggested the students’ attitude is directly related to learning and that sc hool climate impacts students’ attitude. The shift in focus moved to not only understanding the early adolescent’s physical and psychological developmental need s but also how these needs matched the learning environment provided for this population. Recognition of the need to understand the whole child was explored in Alexander and George’s (1981) book The Exemplary Middle School. The authors contend that the reconfiguration of the middle school had very little to do with academic achievement of the early adolescent. Instead what should be used to guide student achievement are the characteristics of an exemplary middle school model. The researchers offer 12 characteristics for this model: 1. Statement of school philosophy and goals 2. System for planning and evaluating de signed for middle school and involving all stakeholders (school administrato rs, teachers, parents, and students) 3. Curriculum plan that provides instruc tion that builds continuous progress and meets the differentiated needs of the population 4. Guidance and relationship with adults 5. Interdisciplinary planning, teaching, and evaluation 6. Flexible grouping for instruction 7. Block scheduling to provide flexible and efficient use of time

PAGE 43

30 8. Varied programs 9. Instruction which utilizes a balanced vari ety of effective strategies to achieve continuous progress of each learner to meet instructional objectives 10. Strong leadership, and professional development 11. Plan for evaluation for both the students and the school 12. All stakeholders working to meet the needs of the early adolescent learner. (pp.18-19) Along with the line of research that addressed the developmental needs, contextual conditions, and a match to effectiv e instructional practices was a concern with the transition to a middle school during the onset of puberty. This was thought to be disruptive for the ea rly adolescent. Simmons and Blyth (1987) conducted a comparison study across two different school configurations of 7th grade students. One group of 7th graders transitioned at the beginning of their 7th grade year to a middle school and a second group of 7th graders remained in a K-8 school. Using a short-te rmed longitudinal design, indices of selfconcept, social adjustment, school attitudes, as well as academic achievement, the researcher’s assessed 160 adolescents both pr ior to and during the transition of middle school. They found the students who transiti oned to the new school configuration had lower self-esteem, lower grades, and more nega tive attitude towards school. Eccles, Lord, and Midgley (1991), replicated this study by using the National Education Longitudinal Study (NELS: 88, 1988) data to compare 8th grader students who attended a K-8 school and 8th graders who were students in other school configurations (junior high/ middle school).

PAGE 44

31 The researchers documented that the tran sition to junior hi gh or middle school was marked by a general decline in the st udents’ motivation, attitude about school, perception of ability, and academic achievement The researchers proposed that it was not a good fit between the developmental needs of the adolescent and the environmental change. In their study Eccles et al. (1991) argu ed that early adolescent were facing changes (social, emotional, physical, ps ychological, and cognitive) and the school environment provided did not fit their needs. Instead of providing for the developmental needs of the students (e.g., wanting more au tonomy), they were given less choice and had more restrictions placed on them. As a resu lt of the poor match between developmental needs and the transition into middle school, the students showed decreased motivation, self-esteem, and academic performance (Juvonen et al., 2004). However, the effects of middle school transitions have varied across studies. While some researchers such as Simmons and Blyth (1987) and Eccles, Lord, and Midgley (1991), argued that, a ne gative effect existed with these transitions to the middle school for the adolescent, other researchers illustrated the adjustment had no adverse effects on these students (e.g., Crockett, Pete rsen, Garber, Schulenbe rg, & Ebata, 1989; Hirsch & Rapkin, 1987). Regardless of these alternative reports the Carnegie Council focused their recommendations on th e negative developmental fit. The Carnegie Council on Adolescent Deve lopmental (1989) presented a powerful vision for the middle school and their learners in Turning Points: Preparing American Youth for the 21st Century. The Carnegie report (1989) concluded: Middle grade schools-junior high, in termediate, or middle schools-are

PAGE 45

32 potentially society’s most powerful force to recapture millions of youth adrift. Yet too often they exa cerbate the problems the youth face. A volatile mismatch exists between the organization and curriculum of the middle grades schools, and th e intellectual, emotional, and interpersonal needs of young adolescents. (p.32) In this report the Carnegie Council (1989) identified five overarching goals the early adolescent student should attain on l eaving the middle school. They should be: (1) an intellectual caring person, (2) a person en route to a lifetim e of meaningful work, (3) a good citizen, (4) a caring indivi dual, and (5) a healthy person In order to achieve these goals the council made eight recommendations: dividing large middle schools into smaller communities of learning students should all be taught a core of common knowledge ensure success for all students empower teachers and administrators prepare teachers to teach the middle grades improve academic performance through better health and fitness connect schools with communities These students need an understanding of their unique developmental needs and instructional practice to match their needs. It was these tensions that complicated the lives of the middle school content area teacher (Brodhagen & Bean, 1996). The Middle School Teacher The contextual debate on whether the middle school should be more like the elementary classroom that emphasized a (child centered approach to teaching) or the high

PAGE 46

33 school that emphasized (disci plinary rigor) placed the mi ddle school teacher in the tensions of the contextual configurations a nd developmental needs of the early adolescent (Brodhagen & Bean, 1996).Along with general contextual and developmental issues for the middle school teacher a concern was wh at should be taught and how instruction should be delivered. Specifically, this left the middle school teacher in a state of ambiguity, questioning whether they were cont ent specific professionals, child centered developmentalists, or somewhere in between. In his study of organizationa l design and instructional f eatures McPartland (1987) drew data from a sample of 433 schools in the Pennsylvania Education Quality Assessment. The purpose of the study was to examine effects of instruction that was accomplished through a selfcontained classroom setting and instruction that was departmentalized, while looking at: (a) student-teacher rela tionship and (b) quality of subject matter instruction. McPartland conc luded self-contained classrooms were conducive to student-teacher relationships however, depart mentalization instruction provided higher quality instru ction. He recommended a balance of instructional features that combine both a personal relationship w ith students and mastery of the teacher content, to benefit the ea rly adolescent learner. Becker’s (1987) research inve stigated whether different grade level configurations (elementary or middle) affect learning for the students with different abilities and especially socio-economic levels. From a sa mple of 8, 000 sixth graders in Pennsylvania he determined that the elementary school se tting and instructional approach benefited students from a lower social economic backgr ounds because of different experiences and background knowledge related to school instruction.

PAGE 47

34 Researchers of curriculu m instruction, Lounsbury and Vars (1978), Hodgkinson (1986), and Slavin (1988) identified impr ovement in learning when cooperative techniques were in place. Mo llified (1988) stressed the need to balance learning needs for the early adolescent, and to provide pr ofessional development for teachers. Mac Iver and Epstein (1993) research ed Elementary and Middle Schools (CREMS), through the Johns Hopkins Research Center. They conducted a survey with principals of 2,400 schools in the United Stat es which included seventh-grade students. A total of 1,753 (73 %) provided information about their schools. 1,344 returned the surveys by mail, and 409 completed surveys by telephone. The telephone interviews were conducted through a random subsample of a ll nonrespondents to the mail survey. The researchers used multiple regression analyses to identify significant consequences of instructional practices by middle school teachers for their students. The focus of the study was to investig ate instructional de livery (strategic approach in reading), teache rstudent relationships, organi zational instructional formats (interdisciplinary team s, or departmentalization), and remediation for students. They found responsive practice (strong teacher to student relationships), and support for students who struggle (extra period during th e school day) are most beneficial. In addition, instructional organization through inte rdisciplinary teams was shown to be more responsive to the needs of the early adolescent rather than departme ntalized organization. However, even with the compliance with responsive practice, most middle grade instructional delivery emphasized drill and pr actice and infrequently used interactive instructional approaches or cooperative le arning. This practice especially impacted

PAGE 48

35 strategic processes that woul d be used to understand conten t material, particularly in reading and comprehending c ontent subject materials. Historical Parallels of R eading in the Content Areas Historically, content area read ing origins transpired, as a result of a readers’ need for strategies when they engage in certain subj ect areas, with many different types of texts for different purposes. Specifically, content area reading instruction is designed to deliver those strategies, so that students develop r eading-to-learn strategi es across and within various content areas (Moore, Readence, & Rickleman, 1983). In a historical review Moor e et al. (1986) presented an historical overview of this field by presenting the origins of content reading and a disc ussion on how best to deliver instruction. The researchers assert that the historical review is of “public discourses…tracing the prominent opinions and research findings, that were reported in journal articles, conf erence proceedings, and textbooks ” (Moore et al., 1983, p.420). Instruction in general du ring the early 1900’s consisted of rote learning. Students were responsible for memorizing and then reci ting information back to show evidence of learning. This changed, however, with the turn of the century as new goals for reading instruction were influenced by humanists, developmentalists, and scientific determinists (Moore et al., 1983). Humanists were concerned for the de velopment of the whole child, and the schools were charged with ensuring that le arning should be meaningful and a student should be an independent thinker. The Pr ogressive Movement was derived from the humanistic stance, contributed to mean ingful reading (Moore et al., 1983).

PAGE 49

36 Dewey (1910) and James (1923) two co mpelling factors in the progressive movement contributed to moving the educati on field forward for meaningful reading. Dewey (1910) criticized rote l earning and argued that learni ng should be connected to a child’s experiences, interest, and proble m solving abilities. In his classic work, How We Think (1910) Dewey presented the theoretical de velopment of reflective thought and how that should transfer to practice. James’ s (1923) work was concerned with the child knowing factual information, but not being able to make inferences about the information read. It was this meaningful, inferential le arning and independent th inking that carried clear implications for the readi ng process (Moore et al., 1983). Developmentalism also became influen tial to content area reading history. Identifying the needs of the early adol escent through child st udy, psychologists (e.g., Gesell, 1915; Hall, 1908) informed the read ing field pertaining to growth and development patterns of the early adol escent. Reading educators (e.g., Gray, 1939) translated this practi ce to reading. Gray’s (1939) resear ch focused on reading strategies for purposeful reading. Gray noted: instead of assuming that pupils enter the higher grades with fully developed and adequate reading habits, an essential step on the part of all teachers is to ascertain the level at which their pupils can re ad with ease and understanding… This may be different for each student, but is n ecessary for teachers to identify the developmental level of each student. (p.7) His findings contend reading strategy instru ction should be for all students beyond the elementary grades, and instruction should be differentiated to meet the literacy needs of all students.

PAGE 50

37 Along with the progressive movement and developmentalism, the Scientific Determinists called for scientific empirical support of reading. Scientific Determinists looked for one absolute truth about the read ing using the process of the scientific methods. However, there was debate betw een social efficacy groups and reading researchers about interpretations. Social e fficacy groups sought to identify the most effective ways to measure students’ academic ability in reading. They argued for the use of standardized testing measurements as a way to determine the students’ reading achievement (Callahan, 1962). Test developers Binet (1904), Rice (1913), and Thorndike (1917) investigated standardized testing in struments measuring r eading comprehension. In this way tests could be administered and scored under a consistent set of procedures, and this would make it possible to compare results across individuals and schools. These instruments measured reading comprehension without the benefit of direct instruction. Resnick and Resnick (1977) contend that students need support in comprehending text, and assessing comprehension without explicit instruction does not accurately measure what the student understands. Huey (1908) and Thorndike (1917) exam ples of pioneers helped lay the foundation in reading theory and practice. Huey ’s (1908) form of inquiry in the literacy field explored the psychological influences of reading comprehension particularly how children’s personal background lite racy experiences influence th eir reading development. Discussions between the researchers consisted of oral language acquired both in and out of school, playing with sounds and words, and developing schema about the complex process of reading. Thorndike (1917) explor ed and investigated the complexity of comprehending text, cognition in reading and how internalizing r eading moves through

PAGE 51

38 questioning from oral to silent gui ded reading. In hi s classic work Reading as Reasoning: A Study of Mistakes in Paragraph Reading, Thorndike (1917) conducted a quantitative study of 200 sixth grade student s to understand th e reasoning process in reading. He found readers need a predetermined purpose for reading. Thorndike argued for oral reading to be replaced with silent reading, a nd to have students ask themselves questions while reading, answer questions after reading, and summarize material that they read. Thorndike (1917) concluded that “Perhaps it is in their outside reading of stories and in their study of geography, history, and the like, that many school childr en learn to read” (p. 282). Studies also looked at the correlation between student academic achievement and reading. Smith (1919) compared subject matte r achievement in math with grades in English and found high correlation among the m easures. He concluded reading ability was related to school achievement. Along w ith Smith’s work, Wagner’s (1938) work measured reading skills in nine areas of subject matter achievement for ninth grade students. She found ability in composite read ing comprehension was related strongly to composite ninth-grade achievement in other content areas. Instructional Practices Historically along with the conceptual unde rstanding of conten t area reading there were issues relating to how in structional delivery should occur. In essence there were two forms of instructional formats existed: dir ect, skill-centered instru ction and functional, content center instruction (Her ber, 1970). Direct instruct ion strategies to understand content occurs when teachers identify a set of skills and present them to students regardless of the content tasks. Reading researchers and educat ors Gray (1919), Gates

PAGE 52

39 (1935) and McKee (1934) gave argument s that by providing systematic reading instruction students were sure to receive instruction in all skills that were deemed important. This placed content specific l earning second to reading skills. Another argument for direct instruction was secondary -school educators only assumed that they were responsible for content specifics a nd by requiring reading skills taught across disciplines, students would acquire the process for understanding the content. Functional content centered instruction occurs when conten t teachers identify reading skills that are a prerequisite for co mprehending content material. These skills are then presented along with the subject matter to be taught. This format of instructional delivery was endorsed by early progressivi sts (e.g., Parker, 1894; Thorne-Thomsen, 1901). These researchers claimed reading would be enhanced through the study of various content subjects. Therefore, reading and specific skills and strategies should be embedded in the cont ent instruction. Moore et al. (1983) contend there were two main histor ical arguments for contentcentered instruction: motivation and transfer. Motivation assumed the affective aspects of reading; if students were in terested and understood the pur pose for reading the content material, they would improve their reading. Mo tivation in reading can be defined as the cluster of personal goals, values, and beliefs th at an individual posse sses and applies in a literacy situation (Guthrie & Wigfield, 2000).Transfer of read ing skills and strategies concerned the ability to use sp ecific skills learned in one co ntent area and transfer it to another. In Teaching Reading in the Content Areas Herber (1970) addressed determining whether early adolescent would be best served by reading instructi on in separate reading periods or during the presentation of content material. Herber (1970) believed functional

PAGE 53

40 instruction is the preferred method, where cont ent teachers address reading abilities while teaching the content specific subject matter. This contention received empirical support from a series of investigations (e.g., Herber & Barron, 1973; Herber & Riley, 1979; Herber & Sanders, 1969; Herber & Vacca, 1977). Skills and Strategies Relate d To Specific Content Areas Along with general instructional procedur es related to cont ent area reading, a historical question arose whet her there should be content specific reading skills and strategies or generic reading skills and strategies taught ac ross all content areas remains of interest. Judd and Buswell’ s (1922) studies involved an eye movement analysis over seven different content areas. They found di fferent types of text materials require different strategies. They measured the num ber of eye fixations per line, duration of fixations, and the number of regressions that differed according to text being read. They recommended that across various content areas there should be diffe rent reading skills and strategies to access the c ontent material being read. Vocabulary frequency counts, and difficulty with subject matter technical and vocabulary within various cont ent areas, were also of conc ern historically. Thorndike’s (1921) work sought to scientif ically measure the frequenc y of vocabulary uses from sources that students would have to read. Lists of words we re generated and these words were tested for accuracy and automaticy by timed tests (15 minutes) (Dolch, 1928). Along with the number of commonly occurri ng high frequency words, studies were conducted on technical vocabulary w ithin various content areas. Pressley (1923) collected 200 school te xts of various subjects, and then subjectively chose words she felt were words that appeared frequently and that required

PAGE 54

41 understanding for the conten t area. Although the methodology was flawed because of subjectiveness of word choice, other studies in content specific fields: gave evidence supporting for the need to understand technical vocabulary within various content areas to comprehend the subject matter: history Barr and Gifford (1929), math Buswell and John (1931), and science Curtis (1938) (Moore et al., 1983). Comprehension Another line of research investigated how the early adolescent comprehends various content materials. Ritter and Lofl and (1924) studied the correlation between comprehension questions that were answered after reading conten t specific expository passages (e.g., science passage from text) a nd comprehension of ge neral narrative reading (e.g., passage from book or language arts te xt) tasks. They found correlations varied among different grades and individuals. They interpreted the findings as meaning reading competencies were to be learned within the context of the content area to be comprehensive. McAllister (1930, 1932) conducted a qualita tive study assessing content reading materials and classroom tasks. He used observa tions of subjective anal ysis with students’ written reports and interviews. McAllister concluded differences in students’ comprehension within various content areas b ecause of the type of reading activities and the support given to students to complete th e tasks. However, he also suggested there should be considerations for more generic r eading skills and strategies that could be taught and modified to meet the specific content material. Generic treatment of readi ng skills and strategies is based on the premise that one common set of skills and strategies can be used in various content areas and be adjusted

PAGE 55

42 to meet the needs of each content area (Moore et al., 1983). Strategies such as comprehension monitoring, fluency, and questioni ng can be used in all content specific classrooms and be modified to meet the spec ific content demands. This way the integrity of the content remains and cont ent teachers are using reading sk ills and strategies to assist their students. Textbooks In the early 1900’s students’ textbooks were McGuffy Readers, selections in pose and poetry for the reading classrooms. Furthe rmore, carried over from the 1800’s were messages including religious and moral themes (Moore et al., 1983). Most of the texts were narrative in nature. This created probl ems, however, when the early adolescent had to read expository texts, because of the different strategies and skills needed to comprehend this textual format. Content area teachers often used a single textbook to teach content specifics. The content teachers had difficulty using the content textbooks to meet the needs of their students because of th e difficult language the new text utilized. In addition, teachers were not trained eff ectively in pedagogy (Beane & Brodhagen, 1996). In the beginning of the 20th century the need to supplement the textbook was addressed. In 1927, Good wrote The Supplementary Reading Assignment, which reported suggested practices to use supplemental books along with the classroom text. Kilpatrick (1919) and Whipple (1920) presented vari ous units of study that used thematic approaches with supplemental materials. Th e use of supplemental materials in various content areas was difficult because of time, management, and cost (Moore et al., 1983). Also, the various levels and different needs of the readers in the classes added to the complexity of a single text for the classroom.

PAGE 56

43 The Learner and Content Area Reading Gray (1937), Kottmeyer (1944) and W itty (1948) contended reading instruction for the early adolescent in the secondary school setting during th e early 1900’s was a general remedial pull-out model, of large scale testing, and in struction in special classes. The problem was that content area teachers then had to provide reading instruction that would transfer to meet the needs of these le arners in their classroom. These concerns for remedial programs helped focus content teach ers on reading in the content area and the need for not only students who struggled but also average readers who needed support in reading content material (Moore et al., 1983). School testing demonstrated reading development did not stop in the elementary grades, and reading abilities were seen to have no upper levels. Therefore, differentiated reading instruction was determined to be important in the middle grades (Bond & Bond, 1941). During the time of the Carnegie Repor t of 1989, which, urged a development fit for school instructional pract ices for the early adolescent learner, Alvermann and Moore’s (1991) review of th e secondary school reading pr actices gave insight into reading in the content areas. The dominant in structional activity in the secondary school content reading practice was a combinati on of lecture, textbook assignment, and classroom recitation (Holton, 1982). Dolan, Harri son, and Gardner (1979) noted that half of all classroom reading occurs in short bursts of less than 15 seconds in any one minute. Usually, according to the researchers, these reading bursts were combined with speaking, listening, or writing activities. Textbooks or teacher lectur es were the primary source of information. Conley (1986), and Mitman, Mergendoller, and St. Clair (1987) concluded this was due to a lack

PAGE 57

44 of teachers’ knowledge and self-confidence about reading practices, including how to integrate reading skills within content info rmation. Wiley (1977) reported a single set of textbooks was used to relate content materi al over the course of the school year. The teacher could assess content material by aski ng questions requiring verbatim responses from the texts. The purpose was to contro l classroom discourse to specific factual information extracted from the single text. Secondary teachers relied on a single text book, due to concerns about time and resources (Alvermann & Moore, 1991). The time to prepare for projects and pulling supplemental material, due to the changing of classes with multiple students, and the need to cover content materials made re lying on a single textbook easier (Dillion, 1983). Content coverage and price often determ ined adoption of specific textbooks. Unfortunately, the comprehensive coverage of content often left the early adolescent unable to comprehend the mate rial needed (Broudy, 1975; Coser, Kadushin, & Powell, 1982). Pearson and Fielding (1991) contend that a sequence is necessary to be most beneficial in order to build r eading comprehension. They suggest: The optimal context for independent cont extual practice may be one in which practice is preceded by instru ction, it is carried out on appropriate materials, is monitored to insure students actually are engaged, and it followed by response of feedback to what is being read. (p. 850) Summary The last one hundred years have shifte d the contextual focus on the early adolescent learner from the “wasted grades ” in the elementary schools to discipline oriented practices of the high school. The develo pmental nature of the learner also shifted

PAGE 58

45 from psychophysical nature to understanding th e unique qualities of th e early adolescent learner. The middle school and its philosophy brought attention to unique needs of this population of learners. There was a call for th e developmental fit for the learner, within the school context, delivered through instructiona l practices concerni ng reading in the content area. Unfortunately, until the 1980s research on the contextual issues of the middle school, the overlap of the needs of early adolescents, and how best to provide instructional practice, was of remarkably low quality (Johnston, 1984). The low quality was attributed to weak design and met hodology, and as claimed by Wiles and Thompson (1975) in their analys is of the research on the middle school, research by proponents and opponents of the middle school movement mere ly studied and reported outcomes that confirmed their subjective positions. The late eighties and nineties witnessed many changes in the early adolescen t literacy learner and their needs within the context of middle school however, they are still faced with many new and yet similar complexities as their historical predecessors. Table 1 provides a timeline of implications from the middle school movement and how this histori cally parallels the evolution of content reading.

PAGE 59

46 Table 1 Middle School Timeline Period Middle School Content Reading 1900’s National Education Association Progressive Movement Dewey (1908) called for the reconfigurati on of James (1923) learning should be secondary school for 7th and 8th connected to child's experiences, and grade, taking a developmental should provide opportuniti es for critical stance. thinking. 1908 G. Stanley Hallbook on Research explored the psychological adolescence unique stage of infl uences on reading comprehension. development. Huey (1908) and Thorndike (1917) 1917 Smith Hughes Actbeginning of Testing instruments to measure vocational curricula programs. read ing comprehension developed. 1920 The development of the Junior Reading in the content area should High School by the Commission provides di rect systematic instruction. on the Reorganization of Secondary-school teachers assumed Secondary Education. they were responsible for content. 1956 Gruhn & Douglas-dissatisfaction Disseminati on on theories related to with instructional programs for the diversity of reading levels and needs early adolescent. Need to match for the early adolescent in the instruction to the students. various content classes. 1960 Junior high has too many Research on motivation and transfer similarities to senior high. The of reading skills. Also, whether evolution of the middle school. reading should be taught in a separate class or in the content class. 1981 Alexander & George’s book Herber’s work addressed functional The Exemplary Middle School. instructionaddresses reading abilities with content materials. 1989 Carnegie Council concluded that P earson & Fielding work on understanding of the deve lopmental reading comprehension found that needs of the early adolescent and a practi ce with text is preceded by match of those needs to practice. instruction.

PAGE 60

47 Current Complexities of Literacy Learning The Early Adolescent Literacy Learner As previously discussed, historically th e focus was to understand the unique stage of development for this population of lear ners (e.g., Hall 1908), along with matching this developmental stage to the context of sc hool (Alexander & George, 1981). However, through this historical lens there was very little depth in understanding the unique and complex issues for the learner, particularly involving their lite racy learning needs. Alexander (1998) developed and tested her “ Model of Domain Learning” for adolescent readers. She suggests that the early adol escent readers range across a developmental continuum in their reading with various texts. She believes that reading development can be traced in the evolution and interplay of three fundame ntal factors: prior knowledge, interest, and strategic proce ssing. Her research suggests that there are stages that the reader goes through in their reading: 1. Acclimation: occurs when the reader is on unfamiliar terrain and this requires considerable strategic effort. 2. Competence: occurs when the reader is starting to efficiently process and becomes more fluent in their reading. 3. Expertise: occurs when the reader is comprehensive, fluent, creative, and analytical. Alexander (1998) cautions, however that st ages are not grade or age specific, and that a reader may be competent or an expert fl uent reader in one kind of literacy task, and in turn may drop back to acclimation during another literacy task. This specifically

PAGE 61

48 occurs when there is a lack of prior knowledge, interests, or strategic processing. This complexity was supported in Ivey’s ( 1999) cas e study with three sixt h grade students of varying reading abilities. In her multicase study, Ivey spent five mont hs with three sixth grade students who had different levels of success with their r eading. She found these students were complex and multidimensional as readers. All were motiv ated to read texts they found interesting and had selfselected. Their disposition to read was dependent on the instructional environment in which their reading occurred. Ivey also noted that care must be taken about generalizations or labels placed on readers. The student deemed a “struggling reader” was able to read fluently and comprehensively when texts were at her instructional level, and he r listening comprehension wa s also strong. The “average reader” was unmotivated to read and therefore, although fluent with words, lacked strong comprehension strategies. However, when this student was able to self-select books of interest this changed. The “capable reader” would read whenever it was requested but was troubled by not understanding the purpose fo r reading some school sanctioned texts. Therefore, reading occurred only when requi red, or when this reader understood the purpose of the reading assignment. Ivey cauti ons that labels or categories given to classify readers offers limited information about who the reader is and the complexities of individual experiences. Although Ivey’s study supports her argument that the early adolescent is a complex and multi-dimensional reader, this study occurred in a single classroom with three students. However, O’Brien (2001) also cautions us not to be too facile in our assessment of adolescents’ literacy abilities.

PAGE 62

49 Drawing on his work with “at-risk” stude nts in a high school literacy lab, he argues that the students’ full literacy compet ence is not apparent solely by the narrow structure of school-sanctioned l iteracy. Instead he contends that in his research the students displayed sophisticated literacy skills as they are combined with art, sound, and print in their multimedia productions. O’Brie n argues that we must recognize these students who are labeled as “at risk” can be “artistic, crea tive, innovative, and daring at using a variety of popular media… [T]hey ar e skillful and creative at constructing and interpreting a range of media texts… using a variety of symbols and signs for conveying and communicating” (p.3). Along with understanding the diversity of th e early adolescent literacy learner, their cognitive abilities ar e also under continual development (Phelps, 2005). Kuhn, Black, Keselman, and Kaplan (2000) study addressed instru ctional practices in the content area of science. The researchers di d not conduct a reading study per se but they contend middle school student’s cognitive development is aided by both direct instruction and by practice. Kuhn et al., experimental study consiste d of middle school students in a multimedia science experiment project of si x weeks. They found the treatment students outperformed the control students on the proj ect final assessment. The experimental group received explicit direct instructions and practice on how to complete the tasks, whereas the control groups we re given instructions on how to complete the tasks. The experimental group outperformed the counter parts on this task along with similar multivariable transfer follow-up tasks. The researchers argue that the early adolescent cognitive skills can be aided by both direct instruction and practice.

PAGE 63

50 Understanding the full range of adolescent li teracies and the role of literacy in adolescent development is important. Th is suggests the early adolescents need opportunities in school to explore both multip le texts and multiple literacies and to receive instruction and oppor tunities for practice and suppor t from peers and adults. However, Cuban (1992) argues the contextual conditions afforded for the middle school students have not dramatically changed sin ce the reconfiguration of the junior high school. School Structure O’Brien, Stewart, & Moje (1995) contend th e infusion of content literacy into the middle school curriculum and school organizatio n has changed very little over the last one hundred years. The institutional organiza tion is formed around an approved formal curriculum divided by disciplines and is c ontrolled by time and space through the context of school. Talbert and Bascia (1990) claim th at this organization is framed around: (a) Six or seven class periods, about 50 minutes each, (b) Approved knowledge base divided into subject areas, and (c) Thr ee or four elective classes that are mandated to meet core curriculum requirements. The success of the curriculum is gauged by content coverage, and the amount of seat time a student accrues (O’Brien et al., 1995). Furthermore, given the unique indi vidual differences among early adolescent literacy learners, curriculum delivery is of ten a one-size-fits-all practice (Alvermann, 2001; Ivey, 1999; Moore, 2000). Therefore, the inte gration of content literacy to meet the diverse needs for this population is challe nged through the contex tual structure and curriculum delivery. Regardless of the compli cations of the school structure, diverse needs of the early adolescent learner, and the pedagogical lens of the middle school

PAGE 64

51 teacher, one of the significant complexities fo r this population is the political pressure from the reform movement with mandates and highstakes testing. Mandates and HighStakes Testing Historically, political and societal infl uences have impacted the educational process for the early adolescent learner for the last one hundred years (Cuban, 1992).The current effect of mandates and accountability th rough high-stakes testing as a result of the 2001 the No Child Left Behind Act (NCLB) creates the current academic dilemma for this population of literacy learners (RAND, 2005). In itially, prior to the advent of NCLB in 1997, Congress directed the Direct or of the National Institute of Child Health and Human Development (NICHD), in consultation with the Secretary of Education, to convene a national panel to assess the effectiveness of diffe rent approaches used to teach children to read. The National Reading Pa nel (NRP) issued a report in 2000 that responded to a congressional mandate to help parents, teacher s, and policymakers identify key skills and instructional methods central to reading. Using these findings as a foundation for literacy instruction and implementation, NCLB establ ished the Reading First initiative program under Title I, Part B, Subpart I of the Elem entary and Secondary Education Act (ESEA). The goal of this initiative is to ensure that all children in America are reading at or above grade level by the end of the third grade (Uni ted States Department of Education, 2001). The Reading First initiative focus was directed to reading improvement in instruction for grades kindergarten through third grade. As a result of this initiative, less focus was directed to the literacy learni ng needs of middle school students (Alvermann, 2001; Biancarosa & Snow, 2006; Guthrie & Wigfield, 2000; Kamil, 2002). McCombs, Kirby, Barney, Darilek, and Ma gee (2005) contend in their RAND report to the Carnegie

PAGE 65

52 Corporation despite the reading progress made by primary grad e students, this is not the situation for the early adolescent. deLeon, (2002) reports “ many children are not moving beyond basic decoding skills—deciphe ring and/or sounding out—to fluency and comprehension, even as they advance to the fourth grade and classes in history, mathematics, and science” (p. 1). McCombs et al., (2005) claim there is a need for continual instruction in readi ng beyond the third grade. Howe ver, teaching reading in the secondary schools to adolescents is an “orphaned responsibility” (deLeon, 2002). In their study Amrein and Berliner (2002) suggest the Heis enberg Uncertainty Principle: “The more important that any quantitative social indicator becomes in social decision making, the more likely it will distor t and corrupt the social processes it is intended to monitor”(p.5), applies to high-stak e testing currently occurring in the schools. This principle, the research ers suggest, warns us that at taching serious consequences (e.g., high school graduation, re tention, class remediation) to a high-stake testing environment may have serious personal and educational consequences. The purpose of Amrein and Berliner’s st udy was to investigat e whether the highstakes testing program promotes the intended transfer of learning. A sample of eighteen states that had the most severe consequences because of testing results was used in this study. The effects of high-stakes tests on learning (general domain knowledge) as compared to training (narrow focus) were measured by examining indicators of student achievement with other standardized test s. The four different measures were: the ACT, administered by the American College Testing Program; the SAT, the Scholastic Achievement Test, administrated by the College Board;

PAGE 66

53 the NAEP, the National Assessment of Educational Progress, under the direction of the National Center for Education Statistics; and the AP exams, the Advanced Placement examination scores, administrated by the College Board. (p.20) This study, according to Amrein and Ber liner was to clarify the relationship between the scores obtained on a high-stake s test and the domain knowledge the test scores represents. The researchers used an archival time–series research design, to examine the state-by-state and year-to-y ear data on each transfer measure. The independent variables were before and after scores of high-stake testing for high school graduation. The dependent variables were and scores from year to year, (ACT, SAT, NAEP, and AP) before and after the implementa tion of the high-stake test. National trend lines were used as nonequivalent comparis ons group along side th e state trend lines. Also, correlations looked at pa rticipation rates in each stat e after high-school graduation tests. Amrein and Berliner (2002) found the ACT data indicated 67% of the states that used high-school graduation exams posted decreases in ACT performance. These decreases were unrelated to participation rates, and on average, achievement on the ACT decreased. The SAT data indicated that 56% of the states usin g the high-school graduation exam posted decreases in SA T performance after the exams were implemented, however, these decreases were related to SAT participation. Nationally SAT participation showed a decrease of 61% in the states that used high school exit exams. Therefore, the researchers argue if th ese participation rates serve as indicators of

PAGE 67

54 testing, the belief that high-stakes testing pol icies will prepare more students or motivate them to attend college, is not supported. The NAEP data had limitations Interpretation of data for the high school exams and the relationship with math and reading data for the fourth and eighth grade students is weak. The AP data however, showed high school graduation exams did not improve achievement for students as presented by th e number of students passing the various exams. When participation rates were cont rolled the percent of students who passed the AP examinations decreased. Amrein and Berlin er (2002) overall contend “ there is no compelling evidence from a set of states wi th high-stakes testing polices that those policies result in transf er to the broader domains of knowledge and skill for which highstakes test scores must be indicators” (p.54). In addition, the RAND Corporation was authorized by the Carnegie Corporation to inve stigate the data results of state instituted high-stakes testing and scores from the NAEP, regarding the state of literacy achievement for adolescents. McCombs, Kirby, Barney, Darilek, and Magee (2005) we re commissioned by the Carnegie Corporation to investigate the curr ent state of adolescen t literacy learning. Using data from the 2003 NAEP report, the rese archers examined the results in reading achievement at the national level (NAEP), as compared to individual state reported achievement, for students who had reached prof iciency in national l iteracy standards. McCombs et al., suggest that we need to be cautious, because there are differences in rigor of the state level test s and the testing at the nati onal level. Specifically, when defining what it means to be proficient in reading. McCombs et al., suggest:

PAGE 68

55 One important caveat to keep in mind is that, although we present data on the similarities and differences in the results of state assessments and the state NAEP, data from these two assessments are not directly comparable, because of the differences in the tests themselv es and in the definitions of proficiency levels in the NAEP and state performance standards. While one could argue that state and national literacy goals shoul d be reasonably similar, in reality there is debate about whether NAEP achievement standa rds are too challenging. Indeed, Linn (2003) points out that the proficiency st andard on the NAEP is an ambitious one, intended to encourage greater effo rt. The National Assessment Governing Board (NAGB), which sets the standards fo r the NAEP, notes that the proficiency level on the NAEP indicates that students reaching this level have demonstrated competency over challenging subject matter. (p.4) McCombs et al., findings suggest there are several concerns to meet the NCLB (2013) goal for proficiency. They are as follows: 1. Fewer than half the students meet state proficiency standards, less than half of the student’s meet NAEP national proficiency literacy standard. 2. Overall, the pass rates on the middle school states assessments ranged from 21% to 88 %. However, between only 10% and 43 % of 8th graders scored at the proficient level of the NAEP Read ing Assessment. The average pass rate of the 8th graders on the NAEP assessment was 32 %. 3. There is a wide disparity in reading achievement for the subgroups of students (disaggregated by race/ethnicity and poverty status). At the 8th grade level we see a difference of 26–28 percentage point s between the prof iciency rates of

PAGE 69

56 white and African American students; 22–26 percentage points between white and Hispanic students; and 22–24 per centage points between economically advantaged and economically disadvantaged students. 4. At both grade levels (4th and 8th) students with limited English proficiency and students with disabilities tra iled well behind their peers. In conclusion McComb et al. (2005) contend that theses finding are very disturbing, for our adolescent literacy learning, as they prepare for meeting the high demands of literacy needs for the new millennium. The researchers recommend that: It is clear that simply mandating sta ndards and assessments is not going to guarantee success. Unless we, as a nation, are prepared to focus attention and resources on the issue of adolescent lit eracy, our schools are likely to continue producing students who lack skills and w ho are ill-prepared to deal with the demands of post-secondary educa tion and the workplace (p. 85). Summary Currently, although there appears to be emerging themes and important information being investigated about the deve lopmental needs, contextual conditions, and instructional practices, the knowledge base for early adoles cent literacy lear ners is still very much under-studied (Alvermann, 2002; Bean, 2000; Kamil, 2002; Moore, 1996). The research reviewed states some of the current complexities for this population of literacy learners. Historicall y, the early adolescent’s lite racy needs shifted from a superficial understanding of their development (physical emotional) to a complex appreciation of the multidimensional nature of this literacy learner. However, the

PAGE 70

57 pedagogical lens of the middle school teache r has not, it appears, addressed the unique needs for this population of learners. Along with the instructional practices, the contextual conditions afforded to these students contribute to the complexities to m eet their literacy needs. Curriculum delivery and contextual organization ha ve continued to mimic histor ically the junior high. The structure of the divided depart mental domains and the use of a single text for instruction is continued practice for this population. The dominance of the accountability m easures mandated by NCLB (2001) and evaluated through high-stakes te sting has also added to th e current complexities. It appears these tests have provi ded contributing factors as to how content literacy is provided to the early adolescent within the cont ent specific classrooms. This is a result of sanctions applied to the teacher, school, distri ct, and the state as: (a) denied diploma, (b) retention of students, (c) re mediation mandates from the scor es students attained, and (d) rewards and punishments for all stakeholders invo lved in this tests. This influences the instructional delivery for the ea rly adolescent in lite racy at many levels of their schooling. There are however, suggestions and recomme ndations for effective strategic practices practitioners, researchers, a nd all community members need to think about in order to prepare the early adol escent literacy learner for the future.

PAGE 71

58 Recommended Effective Strategic Practi ces for the Early Adolescent Learner Sociocultural Influences A shift in the field of adolescent literacy in the last 10 years has occurred (Phelps, 2005). The research on the political and social climate afforded to this population of literacy learners has shifted hi storically from the “wasted gr ades” of the early 1900’s, to the developmental fit match of the 1980’s, to an apprecia tion of the sociocultural influences on literacy practices at the current times. CookGumprez, (1986), and Scriber and Cole (1981) suggest the sociocultural theories of literacy occur as literacy is used in specific contexts for specific purposes, and is socially constructed and constituted. The act of literacy is embedded in a network of social relations. Moje (1996) suggests that in the secondary content classroom the social context that shapes literacy practices is uniquely complex. Teachers and students in secondary classrooms move from class to cla ss, teacher to teacher, and with a subgroup of peers. Teachers and student s construct meaning about l iteracy and learning events based on values, beliefs, knowledge, depending on the contextual situation. Additionally, teachers and students bring meaning to these interactions through their past beliefs, values, and knowledge during social interactio ns (Moje, 1996). St udies that are guided by broad theories as a social construction have focused on how so cial interactions influence literacy l earning (e.g., Myers, 1992). Moje’s (1996) two year ethnographic study focused on how and why a content area chemistry teacher and her students engage d in literacy activities. Moje contends literacy in this classroom was practiced as a tool for organi zing thinking and learning in the context of the classroom built on relationshi ps with the teacher and students. Also, the

PAGE 72

59 researcher explains that within this study the literacy strategies used were domain specific and did not transfer to other domains. Moje sp eculates the use of literacy strategies in content area classes should be domain specifi c and socially supported by the teachers and students in the classroom. Furthermore, strate gies should be shown as how they could be used in other content areas. Moje also sugge sts that more research should investigate classroom interactions and how they play a part in shaping literacy practices. Englert and Palinscar (1991) define their sociocultura l approach to literacy instruction as the interdepende nce of social and individual processes in the construction of knowledge. When viewing literacy devel opment from a sociocultural approach, literacy arises from the child’s participation in social activities in which there are real reasons to use written language. Ryan’s (2000) work investigates the research on peer groups’ interactions, as a cont ext for adolescent achieveme nt, motivation, engagement, and socialization. Peer Interactions In her analysis on the resear ch of peer group socializati on for the early adolescent Ryan (2000) theorizes peers generally interact three ways with one another. During early adolescence, the peer group becomes a pr ominent context for development (Brown, 1990). The school and classroom provide opportu nities for peers to interact throughout the day. Ryan (2000) reports “peer interac tions consume significantly more time in adolescence compared to childhood” (p. 107). Th ese interactions with peers can concern both academic (e.g., achievement) and nonacademic matters (e.g., engagement, motivation, self-efficacy, and interest). Ryan (2000) suggests three ways that early

PAGE 73

60 adolescents generally experience peer intera ctions within the context of middle school: through information exchange, modeling, and peer pressure. Information exchange occurs when adoles cents have a discussi on with their peers (Berndt, 1999). In an experimental study with eighth-grade students, Berndt, Laychak, and Park (1990) found that when adolescents had to make an academic decision, such as go to a rock concert or study for a test, th ey initially responded differently from one another. However, after discussing this di lemma with their peers, their answers were similar to their peers. This form of inte raction could influence the early adolescent’s choice to partake in the literacy task presented by the teac her if it was used effectively. Modeling is another form of adolescent peer interaction. This interaction refers to individual changes in cognition, beliefs, or affect, which are a result of adolescents observing their peers (Ryan, 2000) Observing a specific behavi or a peer performs or listening to a peer voice a certain belief can induce an adolescen t to change their stance or adopt their peers’ behaviors or beliefs Schunk and Zimmerman (1996) reported peer modeling influences self-efficacy beliefs. In their study, they found that early adolescents who verbalized that they had difficulty with a task and then observed their peers have success with the same task then believed they could complete the task. The early adolescent, when faced with a literacy task, may have succe ss by observing their peers. Peer pressure is the third way that the ear ly adolescent interacts with their peers. Peer pressure takes on the role of so cial reinforcement (Ryan, 2000). Brown, Lohr, and Eicher (1986) found that beliefs and behaviors that are discouraged by the groups are not likely to be displayed whereas beliefs and behavior s that are positively received by the group are more likely to surface. Therefore, participation in the literacy

PAGE 74

61 tasks that the peer group positively received through this interaction could have a positive effect on the group’s beliefs and decisions to participate by the group members. Peer pressure may also play a role in how the peer group influences motivation. Brown, Lohr, and McClenahan (1986) repor t that peer pressure regarding school involvement, is significantly correlated with self-reported behaviors and attitudes regarding school. Ryan (2000) recommends further research on peer interactions within a domain specific classroom may fill in the gaps in the literature. The recommendations from the research of Moje (1996) and Ryan (2000) are used to frame this study’s qualitative component. Ryan’s theory on the th ree general categories of peer interactions will frame the interpretive case study, along wi th Moje’s recommendations that research on interactions within the setting of the content classr oom should be studied to inform practice as to how literacy learning could be shaped. Effective Instructional Strategies In 2004, to help address the issue of adoles cent literacy learners, a panel of five nationally known educational researchers me t with representatives of the Carnegie Corporation of New York and the Alliance for Excellent Education. The focus was to draw up a set of recommendations on how to meet the needs of adolescent literacy learners while propelling the field forw ard (Biancarosa & Snow, 2006). A list of 15 elements were reported and then divided in to two sections: instructional improvements and infrastructure improvements. The instructional elements consisted of: a) direct, exp licit comprehension instruction, b) effectiv e literacy instruction embedded in content, c) motivation and selfdirected learning, d) text base d collaborative learning, e) st rategic tutoring, f) diverse

PAGE 75

62 texts, g) intensive writing, h) technology, and i) formative assessment (Biancarosa & Snow, 2006). Researchers were urged to re-c onceptualize how they perform research with early adolescent literacy learners. Inves tigations should combine different elements so important information about the early adol escent can be determined. The current study utilizes five of these elements. Biancarosa a nd Snow (2006) urge that we must meet these challenges because: Literacy demands have increased and cha nged as the technologi cal capabilities of our society have expanded and been made widely available; concomitantly, the need for flexible, self-regulated individuals who can respond to rapidly changing contexts have also increased. The goal in improving adolescent literacy should not simply be to graduate more students from slightly improved schools, but rather to envision what improvements will be necessary to pr epare tomorrow’s youth for the challenges they will f ace twenty and thirty years from now. America’s schools need to produce literate citizens who are prepared to compete in the global economy and who have skills to purs ue their own learning well beyond high school. (p. 9) Direct Explicit Comprehension Instruction There is an enormous amount of resear ch on reading comprehension. Specially, Dunkin’s (1978-1979) work is pivotal for unde rstanding the need to address reading comprehension for middle school students. Durkin’s monumental work in reading comprehension was in search of how teachers in the field assist chil dren in developing a more critical and deeper understanding about what they read. A request for proposals from the National Institute of Education (NIE) for studies in reading comprehension led

PAGE 76

63 Durkin to undertake this study. The NIE assumed reading comprehension could be taught, was being taught, and ye t instruction in comprehensio n was not as effective as it should be. Durkin, a veteran observer of the classr oom was struck by the second assumption. In her frequent visits to the classroom she had witnessed almost no comprehension instruction being taught. This may be because studies in comprehension instruction were never the focus of previous research, and obs ervations were centered at primary grades. To address this Durkin went in search of the literature to define comprehension and placed her focus on observations in middle a nd upper elementary grades looking not only at the reading block but also in the content area of social studies. Durkin conducted observational studies fo r four years in an elementary fourth grade reading classroom, and in grades three to six during a social studies class period. She reported that comprehension instruction consisted primarily of answering questions, completing workbook pages, or taking tests. Researchers however, questioned Durkin’s criteria for determining what constituted instruction (e.g., Hodges; Heap, 1982). Pearson and Fielding (1991) contend however, this wo rk that motivated other researchers to pursue the meaning of comprehension instruct ion. The researchers suggested the first and most important issue was to recognize the co mplex process of reading comprehension is not a passive process, but an active one. Pearson’s (1985) work on e xplicit instruction for comprehension was an example of research motivated by Durkin’s defin ition. He and his colleagues provided a model that teachers could use to support their stude nts and demonstrate how strategies would build comprehension. The Gradual Release of Responsibility Model of Instruction

PAGE 77

64 (Pearson & Gallagher, 1983) suggests teachers model an instruction strategy and have students practice that strategy with guidance followed by independent practice. The role of the teacher is to model, guide and releas e responsibility to their students. This is accomplished through teacher modeling of their cognitive processes, then assisting and scaffolding students to share their cognitive pr ocesses, and finally re leasing responsibility to the individual learne r. Using this model the teacher f acilitates, models, and coaches the learner not to provide individualized instru ction but to monitor progress individually. This form is aligned with Vygotskian (1978) principle of moving students when they are directed from an adult, to the point where they can take control of their own learning. Therefore, instruction is scaffolded, through su pport of the teacher to help students carry out the literacy task (Langer, 1984). Strategy Instruction Embedded in the Content Where this instruction ta kes place and how it assists the students to understand the material in the content area is important. In order to address these concerns and to meet the literacy needs of early adolescent, it is important to investigate how literacy is embedded into the content areas (Snow & Bi ancarosa, 2006). Literacy embedded in the content addresses two directions for instru ctional implementation (Snow & Biancarosa, 2006). First, within the Language Arts classroom these princi ples are not discrete skills or techniques, instead the emphasis should be to teach the strategy or skill using other contentarea materials. Second, content area teachers should enc ourage literacy skills and strategies that emphasize the reading and writing practices that are specific to their subject area (Alfassi, 2004). Alfassi’s (2004) research investigated lit eracy that was embedded in content.

PAGE 78

65 In her study Alfassi conducted two sequential experimental studies, over the course of a school year. The studies were interrelated, examining the e fficacy of two models of reading strategy instruction (Reciprocal Read ing Model and Direct Explanation Model). The studies were conducted in a Midwestern hi gh school with profic ient readers. The first study was in an intact heterogeneous fr eshman English language arts classroom, with 49 students. The experimental group consisted of 29 students, whereas, the control group included 20 students. Teachers of the trea tment group were involved in a six-hour strategy training session. Eight expository passages from the stude nt’s textbooks were used. Fry readability was conducted on all passages (Fry, 1977). In addition, 10 comprehension questions, created by the researcher, requiring short an swers following the reading, were completed without the use of the text. Questions were both explicit and implicit (Pearson & Johnson, 1978). Two independent raters (reading specia lists) read the questions and classified them. The internal consistency of the quest ions as measured by Cronbach’s alpha ranged from .71 to .85. At the end of treatment th e teacher gave a Gates-MacGinite Reading Comprehension Test (2000). Th is standardized test was us ed to investigate transfer effects from strategy in struction, to reading co mprehension application. Alfassi contends Study1,demonstrated th at using authentic texts and strategy instruction within th e language arts class resulted in significantly better results, F (2.44)=4.08, p< .05 than their counterparts who we re just exposed to literacy strategies, without the benefits of explicit instruction. In Study 2 the sample participants were 277 sophomore students in four different content classes (science, arts, social studi es, and math). Each of the four classes

PAGE 79

66 combined specific strategy inst ruction with conten t specific instruction. The researcher investigated differential effect s of combined strategy instruc tion to answer different types of questions (explicit, and implicit). Text -driven questions (explicit) related to information in the test, and knowledge-drive n questions (implicit) information gleaned from the test. The results showed afte r the intervention there was a significant improvement on implicit questions, F (1,276) = 12.84, p<. 001. The findings suggested students improved comprehension especially with implicit questions and with explicit strategy instruction. Overall, Al fassi claims that in order fo r readers to construct meaning from text explicit in struction embedded in the conten t area can support all readers. Diverse Texts Along with strategies to comprehend text, it is important to have texts the early adolescent is able to read. Too often texts in the content classroom are too difficult for students to understand. Diversity in text selection for the content classes addresses two issues: (a) interest, and (b) readability fo r the students to understand and access the materials taught. In their studies Worthy, Moorman, and Turner (1999) and Ivey and Broaddus (2001) investigate middle school students’ in terests, engagement and motivation for reading. Worthy et al. (1999) c onducted a two-part survey st udy of reading preferences with 12 sixth-grade language arts teache rs and 426 of their students, from an economically and ethnically diverse distri ct in Texas. They found a gap between students’ preferred reading ma terials and what they were given in school. In addition, when students were interviewed, they were readily able to give the names of their

PAGE 80

67 favorite books, or authors. Wort hy et al., considered this evidence that students’ attitudes toward reading are not as negative as assumed. Ivey and Broaddus (2001) surveyed 1,765 sixth-grade students, in 23 diverse classrooms, located in the northeastern and Mi d-Atlantic States. The purpose of the study was to describe the early adolescents’ motiva tion to read. The researchers found that time to read books, and teacher read alouds are what appeared to motivate these students. While other researchers (e.g., Allington, 1977) have studied the benefits of time for reading to improve reading, th is study found that students fe lt independent reading was a time to make sense out of what they read. Along with affective issues related to reading, the need to read texts at the student’s instructional reading level is im portant. Biancarosa and Snow (2006) suggest that too often students become frustrated wh en the book is too hard for them to read. Given the wide range of reading abilitie s at the middle school level (e.g., Ivey, 1999), texts must be accessible for this diverse populat ion, and meet the vari ous interest levels of the students. Therefore, middle school c ontent classrooms should have diverse texts, especially high-interest texts of varying read ing levels. The current study uses alternative genre of songs as diverse texts. The 200 songs have been analyzed for an instructional reading level. Technology Most middle school content area instru ction in reading is textbook centered, which presents a formidable task for early adolescents in their r eading. Alvermann (2003) suggests that this may be because the students are not able to gain the necessary background knowledge and specialized vocabulary because they are infrequently able to

PAGE 81

68 read their textbooks. The early adolescent who may have difficulty with the linear textbook, is often more adept in media text, wh ich also motivates and engages them while connecting them to realworld interactions (Alvermann, 2003). The computer offers students more c ontrol in terms of s upport, pace and active processing of text (Kamil, 2002). The use of technology as an alternat ive text, links real world experiences and interests and provi des a sound base for its use with early adolescent readers. The National Reading Pane l (NRP) (2000) reports th at there is little empirical research on the topic of the relationship of hypermed ia that supports literacy learning and instruction for middle school read ers. However, there is promising evidence from the synthesized work by (Leu, 2000) on th e effectiveness of literacy instruction for this audience. Leu (2000) reports on the posi tive effects for middle school readers when print and visual texts (e.g., hypermedia, th e internet, and inter active CD-ROMS) are utilized. A meta-analysis of the effects of technology and reading for middle school learners was conducted by Pearson, Ferdi g, Blomeyer, and Moran (2005). They were commissioned by the North Central Regiona l Educational Laboratory (NCREL) Center for Technology to investigate experimental a nd quasi-experimental studies over the last decade in literacy and technology. The purpose of the study was to investigate technology tools used with middle school students addre ssing the reading areas of: (a) strategy use, (b) metacognition, (c) reading motivation, (d ) reading engagement, and (e) reading comprehension. However, Pearson et al. found little experimental research for reading and technology use in the middle grades.

PAGE 82

69 The research that does exist according to the researchers focuses on comprehension with a slight emphasis on metacogniton. The researchers acknowledge that even though the empirical knowledge is weak, there are many excellent theoretical arguments grounded in best practice. Many offer compelling cases that support the use of technology to enhance literacy learning. Although this analysis yielded no strong claims for practice, it did have several recommendati ons for further research. The researchers recommend that future studies investig ating the use of literacy learning through technology for middle school students consider: 1. more experimental and quasi-experimenta l studies using some sort of correlated design (pretests used as covariat es for posttest or repeated measures). 2. balance issues of focus on control and prec ision for five weeks or more, longer studies might have maturation eff ects or other confounding variables. 3. smaller sample sizes more manageable then larger samples. There might be a trade-off between statistical power and experimental precision, however, it may be easier for researchers to maintain a high degree of fidelity to treatment in smaller studies because of the greater manageability prospects. 4. follow the Complementarity Principle: (a) start with a small descriptive study, then (b) a formative experiment that narrow the range of relevant variable, followed by (c) carefully controlled randomized experi ments, and finally (d) conduct a full scale experimental study. 5. more studies that explore th e relationship between commerci al products developed to address the literacy ne eds for the middle school. Little research has investigated commercial technology products used for improving literacy acquisition at the

PAGE 83

70 middle school level. (pp. 19-23) The current concurrent mixed methods study uses a quasiexperimental design for the quantitative phase inve stigating a commercial inte ractive sing-to-read program Tune Into Reading (TIR) (Electronic Learning Produc ts, 2006), with middle school students in a music classroom. The alternativ e text format is an individual computer program, originally developed to improve singin g, which uses a vocal range analyzer that tracks the singer’s pitch and rhythm, comparing it to the correct pi tch of the song. Each student uses a headset with a microphone, linke d to the computer to sing along repeatedly and to record their singing. As suggested by Pearson et al (2005) there is a need to understand this alternative text format and its relationship to lit eracy learning for the middle school student. Motivation and Engagement Motivation in reading can be defined as th e cluster of personal goals, values, and beliefs that an individual possesses and applies in a lite racy situation (Guthrie & Wigfield, 2000). Central to most theories on motivation is a student’s sense of selfefficacy, a belief in how competently he/she will perform a specific task (Bandura, 1997). Providing early adolescents with clear goals for a comprehension task and giving them feedback on their progress can lead to in creased self-efficacy and greater use of comprehension strategies (Schunk & Rice, 1993). In a longitudinal study of sixth and eighth grade students Wenzel (1996) investigated the social and academic constr ucts of motivation and how that affected academic achievement. A sample of 506 students in grades 6 and 8, participated in this study. All participation was voluntary and 92% of the population was white.

PAGE 84

71 Multiple instruments were used to collect the data. They were Motivational Strategies for Learning Mastery of Goal Orientation and end of the year grades for English class. Wenzel found both the sixth gr ade and eighth grade students’ social goal pursuit correlated significantly and positiv ely with academic motivation in reading related values, reading self-efficacy, and generalized goal or ientation. Pursuit of social goals also related to academic outcomes for both sixth and eight grade students. In addition, social motivation was interrelate d to academic motivation as well as performance. Academic motivation was not a predictor of students’ efforts however. Wentzel concluded that if students see themse lves as successful, dependable, wanting to learn new things, and get things done, th ey are in fact more successful. Self-regulated behavior according to Zimme rman (2000) refers to students who are metacognitve, motivational, and behaviorally active in their learni ng. Learners, in other words, who have self-regulated strategies, believe they can perform efficaciously and set various and numerous goals for themselves within a social cognitive view of selfregulation. In their theory of self-determination, Deci and Ryan (1985) investigated the basic need for competence, claiming that intrinsic motivation is maintained when students feel competent in what they are doing. In the theories of motiva tion through engagement, the focus has been on intrinsic and extrinsic motivation. Intrinsic motivation broadly means that students engage in an activity such as r eading, out of curiosity, pursuit of interest, expr essing a preference for challenging text, and demonstrati ng a disposition to read. Extrinsic motivation relates to engagement for students in an activity such as reading, towards the physical outcome of a reward or grades. The most highly interna lized level of motivational development is

PAGE 85

72 intrinsic motivation (Guthrie & Davis, 2003). At this point, the reader will engage in literacy activities for their own enjoyment, regardless of the reward or a grade. This suggests an early adolescent reader w ho is engaged in thei r reading would be more motivated to read. In an extensive revi ew of how instruction influences students’ engagement, Guthrie and Wigfield (2000) conclu ded that the level of student engagement in reading influences student outcomes. Basically, to provide support for reading engagement for middle school readers, Guthri e and Wigfield (2000) suggest the use of their instructional model of engagement. In this model, Guthrie and Wigfield (2000) suggest six characteristics of classroom instruction that in fluence reading engagement a nd motivation: (a) identify a knowledge goal of the lesson and announce it to the students; (b) provide real-world experiences related to the goa l; (c) provide autonomy suppor t to attain knowledge and learning of these goals; (d) use interesting text s for instruction that is relevant to the learning and knowledge goals being studied ; (e) provide instruction of cognitive strategies that empowers students to succeed in reading these texts; and (f) provide opportunities for social collaboration of the students during teaching. The current study provided real world expe rience with the use of the computer program. Autonomy with choice of songs was pr ovided to the students in a diverse and interesting textual format. Fluency Biancarosa and Snow (2006) claim part of what makes teaching effective literacy strategies so difficult is the wide range of needs and experiences that present challenges for the early adolescent learner. Some reader s at this level still have difficulty with

PAGE 86

73 fluently reading the words accurately and w ith automaticity this could hamper their understanding of the various texts in the c ontent area. Whereas, other early adolescent readers read accurately and quickly, enough for comprehension to take place however, they lack the ability to recall strategies to help them comprehend what they read. Still, others have learned the strategies but have not practiced them sufficiently. This is because they have only used them a limited amount of time, with a limited amount of different texts. Fluency has been identified by The Nati onal Reading Panel (2 000) as one of the five critical components of reading (Pikulski & Chard, 2005) As part of the NRP’s review process two salient areas of fluency reading studies emerge d, guided oral reading and silent reading. Guided or al reading studies included su ch approaches as repeated, impress, paired, shared, and a ssisted reading. Silent readi ng studies provided the student participants with time to read by him or herself. Chall’s (1996) model of reading devel opment suggests readers go through stages in their reading, and each stage emphasizes a pa rticular aspect of th e reading. process. According to this theoretical model the r eader moves from: (a) early and emergent development with words, (b) through formal instruction, (c) building fluency for words, (d) then developing automaticity of word r eading, and (e) finally placing emphasis on using reading to learn instead of learning to read to interp ret and synthesize meaning. This model can be interpreted as having th e reader move from familiarity with the sound symbol relationship to automaticity with words to evaluate and synthesize text. However, as previously noted by Alexander (1998) in her Model of Domain Learning, she contends the early adolescent shifts in fluency in reading depending on the literacy

PAGE 87

74 task. Alexander cautions that the stages of fl uency are not grade or age specific, and that a reader may be competent or an expert fluent reader in one kind of literacy task however, they may drop back to acclimati on (needed support) during another literacy task. This specifically occurs when there is a lack of prior knowledge, interests, of strategic processing. Topping (2006) concur s with Alexander when he contends: Fluency is not an entity, a benchmarkable competence, or a static condition. Fluency is adaptive, contextdependent process th at can operate at a number of layers or levels (this is also true of comprehension). Even expert readers will show dysfluency when confr onted with a text on an unfamiliar topic that provides challenges beyond their indepe ndent reading level. Fluency is of little value in itself-it value lies in what it enables. (p. 106) Topping suggests that there are a number of factors that in teract with each other in the area of reading fluency. To demonstrate this interaction he created a model of fluency entitled, The Deep Processing Fluency (DPF) Model (Topping, 2006, pp. 106-129). Topping (2006) claims the relevant factors of reading fluency are arranged into four sequential sectors: (1) predis posing factors ( entr y skills and conditions that facilitate fluency, e.g., text difficulty, engagement vocabulary, memory, motivation, and selfefficacy), (2) surface fluency (speed of accu rate and automatic word recognition), (3) strategic fluency (control of speed of read ing to yield comprehens ion and expression at the optimal level required for specific purpose), and (4) deep fluency (control of speed of reading to maximize comprehension, expres sion and deep reflection for specific purposes, enhancing explicit awar eness an self-regulation of these processes ) (p.107).

PAGE 88

75 Topping also suggests an effective met hod to promote reading fluency is through repeated reading. He cautions however, that all methods are relative to text difficulty for the individual students because most student s are “surface fluent” or word callers at readability level that are too difficult. Some teachers according to Topping advocate having students read and reread texts below their independent reading level, or just assess reading fluency for speed and word recall (surf ace fluency). These practices contribute to construing reading fluency in rather a “linear way” (Topping, 2006, p. 117) however repeated reading is seen a multidimensional event. Repeated Reading: Accuracy, Automaticity, and Prosody Repeated reading was often seen as a wa y to improve word recognition, accuracy and speed for beginning readers or older st ruggling readers (La Berge & Samuels, 1974). In 1979, Samuels tested the theory of automatic information processing in reading. Theoretically it was assumed that if a child could read a passage with accuracy and automatic reading recall (sp eed) they could then concentrate on comprehending what they read in text. To test this theory, Samu els conducted a study w ith a group of mentally challenged beginning readers by having them read and reread short passages (150 words) a number of times until they were able to read the passage with a rate of 85 words per minute (wpm). Initially the children would have a copy of text at their reading level and listen while the passage was read aloud and mode led with the correct pacing, pitch, tone, emphasis, and volume. Then the children woul d go back to their seats and practice. When they felt they were ready they came to Samu el’s and read aloud the passage. Samuel’s would time the children’s reading and chart thei r progress. When they were able to read

PAGE 89

76 the passage with a rate of 85 wpm they c ould move to the next passage. Samuels found however this was very time consuming proce ss, then in 1985 O’Shea, Sindelar, & O’Shea found in their study with third gr ade readers that students only needed to reread text four times to get the benefits of fluent reading. Along with the rereading O’Shea et al., also tested for reading comprehension in their study by having students retell what they remembered after they read the passage. This study helped connect reading fluency (decoding) to reading comprehension. Fluency connection to improve comprehensi on for readers of all ages and abilities has been established (Dowhower, 1987; Ko skien & Blum; Schreiber, 1980). The explanation according to Schriber (1980) is th e lack of prosodic information in printed text specifically; the pitch, stress, volume, a nd tone that help listeners obtain meaning from spoken language. Schriber suggests this could be compensated through repeated reading, which imitates speech. The prosody co mponents of reading fluency address the use of phrasing and expression (Dow hower, 1987, 1991; Schreiber, 1980, 1987, 1991; Schreiber & Read, 1980). When readers adju st appropriate volume, tone, emphasis, phrasing, and other elements when readi ng aloud, they are providing evidence of comprehending text. In this sense fluency, can be seen as a mu ltifaceted event with reading comprehension as the goal. Taylor, Wade, and Yekovich’s (1985) study with 45 struggling readers and 45 of their more proficient counterparts were in distinguishable in passa ge recall after their rereading intervention. Two recall scores were obtained, free recall and cumulative recall that included probes and direct questions. They found that practice through rereading texts was most effective to increase recall.

PAGE 90

77 O’Shea, Sindelar, & O’Shea (1985) found av erage third grade readers who either needed word accuracy or speed and others who needed support in comprehension each met their goals through repeat ed reading. Whereas, Dowhowe r’s (1987) research found that accurate but slow re aders improved both within and between passages in their comprehension when rereading, especially wh en rereading several different passages at their instructional level. For years, teachers thought if students could learn to decode words accurately, they would be successful in reading printed te xt (Rasinski, 2004).While it is true that accuracy in a students’ ability to decode words is important for fluency, as Samuels believed in the 1970’s, decoding needs to be automatic. However, this is still not sufficient. Rasinski (2004) points out the need to connect accuracy and automaticity to reading prosody. Stayter and Allington (1991) s uggest that “we have failed to consider some of the broader ramifications of an emphasis on fl uency, especially with older and more developed readers” (pp.143-144). In their case study with a class of seventh grade students Stayter and Allington (1991) report that fluency instruction enriched the meaning of text. This study investigated a class of 25 heterogeneous seventh graders over five days as they reread a nd rehearsed short dramas. Interv iews were conducted after the students performed for their class. The part icipants all came away with a different understanding of themselves as readers. As noted by the resear cher one student said: The first time I read to know what the words are. Then I read to know what the words say and late r as I read I thought about how to say the words…As I got to know the character better, I put more

PAGE 91

78 feeling in my voice. (p. 145) Texts performed orally are ideal for re peated and prosodic reading (Rasinski, 2004). McGuire (2004) contends that the “r hythms and meter of spoken language are much like the lyrical rhythms and melodies of music” (p. 1). In her autobiographical narrative about her personal struggle to overcome her read ing disability she uses music as the central metaphor to format the study. Rasi nski, Homan, and Biggs (in press) report that “Singing lyrics to songs is a form of reading that is nearly ideal for fluency instruction. Songs are meant to be sung (read) orally and th ey are meant to sung (read) repeatedly” (p.14). This form of repeated exposure through singing as a vehicle for reading, as in the case of the current study, can build reading fluency and comprehension and can be naturally embedded within the music content classroom. Singing Butzlaff (2000) contends there are sim ilar characteristics with singing and reading: (a) music text and writt en text involve formal written notations that are read left to right, (b) the sens itivity to phonological distinctions and word recognition require a sensitivity to pitch and tonal distinctions in both reading and singi ng, (c) when students learn the lyrics to songs they are engaging in reading, and (d) learning song lyrics are often repetitive, so that rereadi ng of text occurs through singing. Music Learning Theory is an explanation on when and how music is learned. This theory’s primary objective is the developmen t of students’ tonal and rhythm audiation (Gordon, 1979). The term “audiation” coined by Gordon, is the process when we hear and comprehend music for which the sound is no longer there. Gordon (1979) contends that the cognitive process is the “musical e quivalent to thinking in language” (pp. 5-6).

PAGE 92

79 When we listen to someone speak we must retain in memory their vocal sounds long enough to recognize and give meaning to th e words the sounds represent. Music is similar, when listing to music we are audi ating sounds that were recently heard. In addition, based on our schema of the tonal a nd rhythmic conventions a person can predict what comes next (Gordon. 1979). However, singing in the mu sic classroom is usually performed as a whole group with one song regardless of th e variety of instructional read ing levels of the student body. Hall, Boone, Grashel, and Watkins (1997) s uggest students should sing independently, on pitch, and with rhythm. The Tune Into Reading study provided opportunities for students to sing independently, supporte d by background music, rhythm and pitch heard through their individual headsets. Goetze, Cooper, and Brown (1990) conduc ted analysis of classroom singing studies over the last 25 year s and concluded methods that included individual singing opportunities and immediate visual and verbal knowledge of results were warranted to increase accuracy in singing. While most singing in the music classroom is done in groups, minimal time is spent with students singing individually, making it difficult to assist each student to develop these speci fic faculties. Levinowitz (1989) found that students sang songs more accurately with text than without. In the Tune Into Reading study students have individual te xts on their computer screen s and scoring mechanism is displayed to record real-time pitch accuracy. In the meta-analysis of over 150 articles, Computer-Based Technology and Music Teaching and Learning, Webster (2002) investigated vari ous studies with computers in music education, including the categories of listening, performing, and composition.

PAGE 93

80 Classrooms are more dominated by technology than ever and students’ skill and understanding of computers often extend be yond those of their teachers. Webster reported that use of computers in the cl assroom, in partnership with teachers’ orchestrating the learning environment, doe s seem to assist in actively engaging the student, increasing motivation and intellectua l stimulation. Individualized instruction facilitates aural instruction can augment th e efforts of music classroom teachers and increase learning in children in a number of different areas. In a pilot study involving 48 strugglin g readers in the sevent h and eighth grades in a rural central west Florida middle school, Biggs, Homan, Dedrick, Minick, & Rasinski (in press) used an interactive singing software program with real time pitch tracking that teaches users to sing in tune and in rhyt hm was used with middle school struggling readers. The computer program, Carry-a Tune was originally developed to improve singing however, it was used in this study to determine its effect on comprehension and instructional reading levels with middle school struggling readers. The 9-week intervention was conducted with 24 struggling mi ddle school readers. A ll participants had failed the state reading test, Florida Comp rehensive Assessment Test (FCAT). Students utilized the software program for 30 minutes 3 times a week. Treatment students were matched with a control group of students by FC AT level, gender, grade level, reading/ language arts teachers and free and reduced lunch. Leveled texts from the Qualitative Reading Inventory (Qualitativ e Reading Inventory, 2004) developed as Cloze passages were administered to all 48 participants a nd served as pretest, posttest, and follow up measures of assessing co mprehension and instruc tional reading levels.

PAGE 94

81 A two-tailed t -test comparing pretest and posttest scores was used to determine the statistical significance at the end of nine weeks. A 2 (Group) x 3 (Time) repeated measures ANOVA of the group grade level averages was utilized for the follow-up testing at the end of the sc hool year. No significant di fferences were found between treatment and control groups’ pr e-test scores however, the posttest results were highly significant for the treatment gr oup. Mean scores of the Treat ment students approached a 2-year gain in their inst ructional reading levels. Current Study and Effective Practice Tune into Reading (TIR) (Electronic Learning Pr oducts, 2006) is an interactive sing-to -read software program that can be used in the music classroom. This technological format provides diverse and inte resting texts. Over two hundred songs are included on the TIR program. All songs were analyzed for readabil ity level. The songs range from first to tenth grade level from traditional folk songs (e.g., Amazing Grace ) to more recent pop songs (e.g., Ain’t No Mountain High Enough ). Direct explicit comprehension instructi on through repeated reading is modeled through singing. The music teacher modeled steps of effective singing by initially showing students how to get their individual vocal range (e.g., alto, soprano). Then the students proceeded with recording their individual vocal range. Once this is accomplished, all of the songs that the student sang matched their individual vocal range. Each student has an individual soundproof microphone headset for listening, singing, and recording while at their computer. The comput er program has two different text formats. The first format, linear sheet music, allows th e students to read the lyrics silently three times, while listening to the background music and tempo. In this way, repeated reading

PAGE 95

82 is embedded into the singing program. This aligns with the recommended number of repetitions suggested by Samuels (1979). Th is is followed by a graphic textual view. This alternative text format provides a visu al display of words broken into syllables without the accompanying musical staff a nd places each syllable accented at the appropriate pitch within each students’ personal vocal range. Along with the visual tracking of the word s, a guideline is provided for accurate pitch and tone that provides a real time track line of the student’s voice while they are singing and recording a song. After singing each time, a score is provided to the student. These scores, ranging from 0-100 represent accur acy of pitch and tone. The teacher uses these scores to determine when to change the level of songs. The students in this study sang and recorded the songs using the visual graphic format three times aloud, and saved the recorded version of their highest score. Strategy instruct ion with diverse texts through a technological format embedded in the cont ent area of music led to engagement and motivation for the learners. Summary This review of the research clarified w hy gaps exist in the literature pertaining to the early adolescent and their literacy learning needs. Historically as noted by researchers (Beane, 2001; Brough, 1995; Cuban, 1992; Spri ng, 1986; Van Til, Vars, & Lounsbury, 1961) these learners have been caught in the tensions of whether the middle school should be more like the elementary school or like the high school. These tensions have also carried over to understand ing the uniqueness of this population of learners, the ambiguity of the role of the middle school teacher, and the delivery of instruction, specifically reading in the content area.

PAGE 96

83 Not unlike their historical predecessors there are current complexities for adolescents that are politically, sociall y, and academically influenced. The current dilemma of accountability and evaluation thr ough high-stakes testing has compromised what has been learned to date about the co mplexity of the early adolescent literacy learner. It has also detoured effective pr actice of literacy em bedded in the content classroom, by not addressing the unique need s for this population, especially when more literacy needs are needed to meet the challenges for the new millennium. How do we prepare students to be flue nt active independent readers and comprehenders? Biancossa and Snow (2006) suggest the early adolescen t literacy learner n eeds explicit direct literacy instruction, which is embedded in th e content classroom to build comprehension. This can be achieved through the use of diverse and interesting texts that are accessible at the reading level of the student. Delivery of these texts could be through a technological format, which can be motivating and engaging for the adolescent. This study took place in the music content classroom, where si nging instruction is taught using explicit instructions in rereading text to build co mprehension. The current study will add to the body of knowledge on the early adolescent stra tegic processes and the need to provide literacy instruction in the cont ent areas to these students of varying reading abilities.

PAGE 97

84 CHAPTER THREE: METHOD Chapter Three presented the methods used to conduct this study. This chapter contained five sections. The first section revealed the purpose of the study and outlines the research questions. The s econd section described the desi gn of the study, the research context, and the participants. The thir d section presented the study’s ethical considerations, instruments, meas ures taken to ensure reliabi lity of the data, researchers’ pre-study involvement, and the procedures. Th e fourth section provided specific details concerning data collection. The final section explained the manner in which data were analyzed and interpreted. The Purpose of the Study and Research Questions The purpose of this concurrent mixed met hods study was to inve stigate the use of an interactive si ng-to-read program Tune Into Reading (Electronic Learning Products, 2006) as an alternative text, embedded w ithin a heterogeneous music classroom. Measured by the Qualitative Reading Inventory-4 (QRI-4) (Leslie & Caldwell, 2006), the fluency, word recognition, comprehension, a nd instructional read ing level of the treatment students were compared to their c ounterparts who sang as part of the regular music program. This investigation also provide d a description of the peers’ interactions during the literacy task assigned by the musi c teacher. The intent of this study was to address the following research questions:

PAGE 98

85 Quantitative Research Questions 1. To what extent is the reading perf ormance of word recognition, fluency, comprehension, and instructional reading level, as measured by the QRI-4, of students using the Tune Into Reading program, different from their regular music curriculum counterparts? 2. To what extent does the Tune Into Reading program differently impact the reading scores of students who are “below, at, or above” grade level as determined by the Florida Comprehensive Assessment Test (FCAT) reading scores? Qualitative Reading Question 1. How do middle school readers interact with their p eers, within the context of their music classroom? The first quantitative research questi on addressed the readers’ use of the interactive sing-to-read program Tune Into Reading as an alternative text, and then was compared to their counterparts who are sing ing as part of the regular music program.. Prior to the treatment, I admini stered a pretest using the QR I-4. Scores from the pretest ensured that the students in the re gular music class and the class using Tune Into Reading were not different in their performance in fluency (measured by words per minute), word recognition (measured by oral reading accura cy), comprehension (measured by implicit and explicit questions after th e reading), and instructiona l reading level (measured by combining scores from word recogniti on and comprehension questions) before implementation. After the implementation of the interactive singtoread program, Tune Into Reading I administered a posttest using the QRI-4 and comp ared the posttest scores with the pretest scores to determine if students in the experimental group gained

PAGE 99

86 significantly over their counter parts in the control group. Th e students were initially assessed at posttest with a r eading passage on the same instru ctional level attained during the pretest. The students were ne xt assessed at posttest at th e highest instructional reading level they attained. The second quantitative research question investigated whether an interaction effect of the repeated reading methods o ccurred on the readin g performance of the students “below, at, or above” grade level as determined by the Florida Comprehensive Assessment Test (FCAT) 2006 in reading, while using the sing-to-read program, Tune Into Reading, as an alternative text. The results in reading achievement level scores (achievement levels 1-5), according to the st ate of Florida Department of Education, are reported as: (a) students who sc ored a Level 1 or 2 are cons idered below proficiency in meeting grade level benchmarks, (b) students who scored a Level 3 are considered at grade level, and (c) students who scored at a Level 4 or 5 are considered above grade level ( FCAT Briefing Book, 2005). Concurrently, the qualitative observations were used to probe for significant themes by describing aspects of peer inter actions (peer talk, peer modeling, and peer social reinforcement) among students w ho sang using the interactive program Tune Into Reading, versus the peer interactions among stude nts who were singing in the traditional music class. Design of the Study In order to address the re search questions, I used a mixed methods approach. The purpose of this approach was to collect, an alyze, and mix or integrate both quantitative and qualitative data during the research pro cess within a single study (Creswell, 2003;

PAGE 100

87 Tashakkori & Teddlie, 2003). Both types of da ta were used because neither quantitative nor qualitative methods in isolation suffici ently capture the trends and details of situations, such as the comple x issues of how the use of an alternative text supports literacy learning of the early adolescent and how these adolescents interact with their peers during the literacy task. When used in combination, quantita tive and qualitative methods complement each other and provide a more complete picture of the research problem (Green, Caracelli, & Graham, 1989; Johnson & Turner, 2003; Tashakkori & Teddlie, 2003). This study used a concurrent mixed methods design consisting of two distinct phases (Creswell, Plano Clark, Guttman, & Hanson, 2003; Tashakkori & Teddlie, 2003). The quantitative numeric data and qualitativ e text data were collected and analyzed concurrently. Integration of the data occurr ed during the interpre tation of the study’s findings. This interpretation can either note the convergence of the findings as a way to strengthen the knowledge claims of the study or explain any lack of convergence that may result (Creswell, 2003). Quantitative Phase The first two questions were answered utilizing a quasi-experimental design. The statistical technique that was used to answer the first question was analysis of variance (ANOVA) with repeated measures to assess differences in mean trend lines over time between the experimental and control gr oup. Multivariate repeated measures ANOVA was conducted to assess the collective differe nces on the dependent variables overtime and by group (Stevens, 2002). The multivaria te repeated measur es ANOVA assessed if the combination of noncommensurate depe ndent variables differed over time and by

PAGE 101

88 group. Simultaneous differences from pretests to posttest by group were further analyzed by conducting t -tests and determining effect sizes. The independent variable for the firs t question was the literacy approach consisting of two levels: the early ad olescents who use the alterative text Tune Into Reading during the literacy tasks and those who are part of the regular music program (treatment and control). The de pendent variables were the sc ores from the QRI-4 on: (a) fluencytimed and measured by words per mi nute, (b) word recognitionmeasured the percentage of accuracy during the oral reading of the passages, (c) comprehensionmeasured by the percentage of correct respons e to questions asked, and (d) instructional reading level assigned a grade level (e.g., 6th) measured by the combination of scores on word recognition and comprehensio n at two points in time (prete st and posttest). Initially, the students were assessed at posttest with a reading passage on the same instructional level attained during the pretes t. The students were next as sessed at posttest on highest instructional reading level they attained. The second question also addressed the stud ents using the interactive sing-to-read Tune Into Reading program and those singing in their regular music class. The purpose was to investigate whether the repeated read ing method with the sing -to-read alternative text program had a different effect on the pe rformance of students who scored below, at or above in their reading level, as de termined by their FCAT level scores. Repeated measures ANOVAs were used to answer this question. It assessed differences in mean trend lines over tim e for the experimental and control groups classified according to belo w, at, or above grade level in FCAT reading scores. The dependent variables remained the same (pretest and posttes t scores from the QRI-4).

PAGE 102

89 However, the independent variables were th e students in the two literacy approaches grouped by their 2006 FCAT level scores in reading. Qualitative Phase The qualitative phase in this study used an interpretive case study approach, with the data collection occurring through participant observation. Inductive analyses were conducted to identify conceptual themes or patterns in the data, and create categories needed (Bogdan & Biklen, 2003; Cohen, Manion, & Morrison, 2000; Merriam, 2001). These themes and categories were analyzed to identify subcategories, which helped to describe peer interactions (e.g., talk, peer modeling, peer reinforcement) during the literacy task (rereading through singing) assigned by their music teacher. This was also considered a bounded case study because it had a defined time, a distinct social inte raction focus, and a physical b oundary (Stake, 1998). The case study was bounded in the context of one literacy task rereading through singi ng, for participants who used the Tune Into Reading program and those who were in the regular singing class, during the fourth quarter of the sc hool year at the west central Florida middle school (March 26, 2007May 25, 2007). In a ddition, the physical boundaries included two cases (one treatment group using the altern ative text and one cont rol group as part of the regular singing program) who were sing ing during their regularly schedule music class period. This interpretive case study approach was used to describe peer interactions during the assigned literacy task. Thus, the quantitative data and results were used to provide a general picture of the research pr oblem: whether the use of an alternative text Tune Into Reading supported literacy learning of ear ly adolescents and improved their

PAGE 103

90 word recognition, fluency, comprehension and in structional reading le vel. The qualitative data and analysis were used to describe th e peer interactions during the literacy task assigned by their teacher. Mixing the Methods Priority was given to the quantitative appro ach because it looked at the statistical relationship between rereading through singing of the particip ants who used the sing-toread program Tune Into Reading and their counterparts in the regular music class. However, concurrently qualitative case st udy methods were used to better understand and describe the peer interactions occurring during the literacy task assigned by their teacher. The integration of the two types of data might occur at several stages in the research process: the data collection, th e data analysis, or the interp retation (Creswell, 2003). In this concurrent mixed method study, the mixing of the data occurred during the qualitative findings section of the resear ch project. The quantitative results and qualitative descriptions were mixed applying a triangulation st rategy in order to provide a clearer picture and answer the research que stions. Figure 1 presents a diagram of the mixed methods concurrent desi gn procedures in this study.

PAGE 104

91 Figure 1 Diagram of the Mixed Methods Concurrent Design and Procedures Research Context The School Site This study was conducted in a rural west central Florida middle school. This public middle school had 1079 students enrolled, and served grades sixth through eighth, ( School Improvement Plan 2006). The school term starts in August and extends through the end of May. The terms are divided into four quarterly reporting periods. I chose the school site because I had established a rapport with the principal and teachers prior to the study. In addition, the music teacher and I worked on previous research projects that have investigated literacy that is em bedded in her music classroom. The staff included 89 full-time teachers a nd 3 administrators. Ethnically, 95% of the staff were Caucasian, 1% were African American, and 4% were Hispanic. In addition, 72% were female and 28% were male. The school had one reading coach and 15 reading and/language arts teachers and support staff. All reading and language arts teachers were reading and ESOL (English as a Second Language) endorsed, through University of Quantitative Data Collection Quantitative Data Analysis Integration o f Quantitative & Qualitative Results Procedures QRI – 4 Pre /Post test (n=64) Data Screening Repeated Measures Factorial Interpretation and explanation of the quantitative and qualitative results Interpretive Case Study Cases (2) Field Notes Qualitative Data Collection Qualitative Data Analysis Coding and thematic analysis Within case across case theme Cross thematic analysis

PAGE 105

92 South Florida (USF) partnership professional de velopment courses, district professional development courses, or department of edu cation state programs. In addition, all content teachers had professional development rela ted to reading in the content areas. The ethnicity of the students is reported in Table 2. Table 2 Percentage Enrollment By Et hnicity At The School Site Total White African American Hispanic Asian Multiracial Enrollment 1079 77% 7% 11% 1% 4% This middle school is the cluster site for the district’s Exceptional Student Education program (ESE). It serves 240 ES E students (22%) with significant cognitive, behavioral and /or physical disabilities fr om around the district. Specifically, this population of students all have an active Indi vidualized Educational Plan (IEP), and the students require direct and extensive instru ction to acquire, maintain, generalize and transfer skills. In addition, students with significant cogni tive disabilities are students whose cognitive abilities are 2.0 standard deviations or more below the mean of their grade level peers (Florida Department of Education, 2006). Additionally, less than 1% of the population is designated as qualifying fo r the ESOL program. The free and reduced lunch program benefits 51% of the student popul ation at this school site. This qualifies the school as a Title 1 school, which receives funding from the state and national level to assist in providing remediation for struggli ng students based on the percentage of free and reduced lunch.

PAGE 106

93 The achievement levels from the 2006 FC AT results in reading for student’s in grades 6 through 8 at this middle school were reported as: (a) 47% were below grade level, (b) 35% were at grade level, and (c) 18% were above grade level The middle school has not made Annual Yearly Progress (AYP) in reading for four consecutive years, however, they made a grade of A in Florida’s A++ program. A needs assessment was completed to address the issues and re view reading for the students. Based on the results of the needs assessment the school ha s taken several steps to improve reading. They have: (a) increased the reading re mediation staff, (b) provided after school tutoring, (c) continued with the Accelerated Reader Program through the purchase of more books at more levels, (d) increased student access to FCAT Explorer and supplemental technology tools, and (e) worked with the reading coach and professional development partnered from USF to support teachers in read ing and reading across the content areas. School instruction is provided through interdisciplinary team teaching by grade level. There are eight teams (two at each gr ade level) made-up of teachers in the core content areas (math, social stud ies, language arts, and scienc e) plus one remedial math and one remedial reading teacher. The teams of students stay together as cohort groups for three years. Elective classes (Art, Music, Computers, and Consumer Education) are assigned to the students at th e beginning of the school year mixing grade levels across teams. Students are assigned an elective class per quarter (Wheel Clas s), so they have an opportunity for each of the four elective classes per year. All students have a heterogeneously gr ouped language arts period daily for 90 minutes. However, students who require remedi ation (globally define d as those students who failed the reading portion of the FCAT 1 and 2) receive 45 minutes of reading

PAGE 107

94 support during this 90 minute period. The first type of reading class includes students who performed at the lowest level on the FC AT: Level 1 (“Intensive” reading course). The second type of reading class includes students who performed at the second-tolowest level on the FCAT: Level 2 (“Corrective” reading course). Students at Level 3, 4, and 5 on the FCAT have classes in reading, a nd they use the FCAT Explorer in reading and Accelerated Reader with leveled texts. The FCAT Explorer is a free online e ducational program for Florida students, which provides FCAT sample questions in r eading and math, related to the Sunshine State Standards ( FCAT Briefing Book 2005). The Accelerated Reader program (Renaissance Learning, Inc., 2006) is an indi vidual computer program using multiple grade level texts of different readability levels. The students read the books at their instructional reading level and take a comput er test, and then the teacher receives the print-out of the results. In addition to the la nguage arts teachers’ li teracy instruction, all content area teachers at this school site are required to incorp orate literacy strategies in their lessons daily and must provide lite racy objectives in their lesson plans (Improvement Plan, 2006). The Music Classroom In order to facilitate the visualizatio n of the enactment of the literacy task (rereading through singing) a desc ription of the physical configuration of the music class is provided. However, one must first enter th e base of a single stor y rectangular shaped middle school to find the music classroom. The base of the rectangle houses the decision making and policy enforcement center of the school, containing the front desk and the different layers of administrative offices. Once allowed in the school, in an attempt to

PAGE 108

95 reach the music classroom and see the inne r working of the school and its physical layout, you must exit the admini stration building through a door parallel to the front desk. Outside the administration building the sound of music fills the air as different musical genres echo throughout the outdoor gardens. Picnic tables and benches, as well as a bird aviary, an alligator pen, and the school’s Holocaust Memorial, are scattered around the center of the rectangle. All buildings at this school have outside access and to enter any of the buildings in the school you must fo llow a covered pathwa y that outlines the parameter of this rectangle. The right side of the rectangle contains the interdisciplinary grade level classrooms sequentially organized from 6th through 8th grade. The gym is located at the top of the rectangle and the service buildings, housing, the guidance center, the media center, the cafeteria, and the music wi ng are at the left side of the rectangle. The music wing is located in the cafeteria building and runs para llel lengthwise to the cafeteria. The long hallway wall in this musical wing displays hand painted music notes, messages to the students, and different characters singing and playing instruments. The music wing is comprised of two large s ound-proof classrooms: (a) the band room is first, and then (b) the chorus room, where this study took place, is second. When you open the door to enter the chorus classroom the rhythms and sounds escape temporally through the soundproof door. The walls of this large room are print rich, covered with songs, musical notes, and schedules, extending all the way up the walls to the 100 foot ceiling. The left wall of the classroom is a hand painted musical scal e displaying symbolic music notation and corresponding words. The right wall has an overhead screen projec ting song lyrics from the projector. The white board at the front of the classroom has notes to the students and

PAGE 109

96 outlines the daily agenda, or assignments to be completed. The top of the back wall is the daily schedule and times for each class. A painted bookshelf landscapes the back wall of the classroom and is autographed with handprints and names of the artists who created it. The classroom seats students in 3 semi-circle stadium steps that descend to the central stage of the classroom. To enter the stage, one could use the stai rs or the ramp, wide enough fo r a wheelchair and hand rails to support balance. The performance areas’ focal point is the piano surrounded by a garden of musical instruments: drums of all sizes, (both handmade and store bought), guitar, auto-harp, and a variety of different rhythm sticks. A music stand and a large African drum, which begs to be hit, are the standing position for the actors that enter the stage. Before the walk to the stage, a soun dproof computer lab housing 15 computers can be seen as you peer through the twoway glass window. Audio visual equipment hides in different corners of the classroom. A large television set, with a VCR and CD disk player, rests near the wh ite board at the front of the classroom. A table outside the teacher’s office holds the tape recorder and CD player, while the overhead projector gets pulled out daily and then neatly tucked next to a bookcase with song books and clipboards. The bell rings and the students rush into the classroom, select their instrument of choice, and sit down. The teacher enters the center stage and brings the group to attention with the beat of the drum: the child ren echo this beat and class begins. Figure 2 provides a floor plan of the music classroom.

PAGE 110

97 Figure 2 Floor Plan of the Music Classroom Participants Sample Design A total of 64 students from one rural west central Florida middle school music classroom participated in this study. A mu sic classroom was chosen because it was appropriate to investigate si nging as a method of rereadi ng to build fluency embedded within the natural context of a chorus cl assroom. Biancarosa and Snow (2006) suggested when “instructional principles of literacy are embedded in content subject-areas, teachers Computer Lab Piano Desk White Board Office Fire Door Bookcase Stadium Seating Music Stand Storage Closet Audio Visual E q ui p ment Closet for Instrument Entrance Door

PAGE 111

98 provide or reinforce instruction in the skills and strategies that are particularly effective in their subject area” (p. 24). The sampling choice for this study was that of convenience. All of the study participants volunteered, were from the same school site, attended the same music class, and had the same music teacher. Although convenience sampling choice limits the generalizablity of the findings to a larger population, this decision: (a) was consistent with the purpose of this study, and (b) is supported through the li terature on technology and reading for middl e school students. As previously noted, the purpose of this study was to provide a description of the phenomena, rereading through singing using Tune Into Reading (Electronic Learning Products, 2006) program as an alternative text. This concurrent mixed methods study investigated the use of an interactive sing-to-read program embedded within a heterogeneous music classroom. Quantitatively, as measured by the Qualitative Reading Inventory-4 (QRI-4) (Leslie & Caldwell, 2006) the fluency, word recognition, comprehension, and instructional reading leve l scores of the treatment students were compared to those of their counterparts who sang as part of the regular music program. Individual assessment of this sample provide d opportunities to assess each participant, completing the full battery of the instrument while also noting and describing individual reading behaviors. Concurrent ly, qualitative observations were used to describe aspects of peer interactions (peer talk, peer modeling, and peer social reinforcement) among students who sang using the interactive program Tune Into Reading, versus the peer interactions among students who sang in the traditional music class.

PAGE 112

99 In addition, for the purposes of the st udy, the literature supported the sampling choice and the current sample size. A meta-a nalysis of the effects of technology and reading for middle school learners was c onducted by Pearson, Ferdig, Blomeyer, and Moran (2005). The researchers were comm issioned by the North Central Regional Educational Laboratory (NCREL) Center for Technology to i nvestigate experimental and quasi-experimental studies over the last decade in literacy and technology. Pearson et al. found little experimental research for read ing and technology use in the middle grades. The researchers made the following recommenda tions for future studies to investigating the use of literacy learning through t echnology for middle school students: 1. More experimental and quasi-experimenta l studies using some sort of correlated design (pretests used as covariat es for posttest or repeated measures). 2. Balance issues of focus on control and prec ision for about five weeks, longer studies might have maturation eff ects or other confounding variables. 3. Smaller sample sizes are more manageable then larger samples. There might be a trade-off between statistical power and experimental precision, however, it may be easier for researchers to maintain a high degree of fidelity to treatment in smaller studies because of the greater manageability prospects. 4. Follow the Complementarity Principle: (a ) start with a small descriptive study, then (b) conduct a formative experiment that narr ow the range of relevant variable, followed by (c) carefully controlled randomi zed experiments, and finally (d) conduct a full scale experimental study. 5. More studies that explore the relationship between commercial produ cts developed to address the literacy ne eds for middle school. Little research has investigated

PAGE 113

100 commercial technology products used for improving literacy acquisition at the middle school level. (pp. 19-23) This mixed methods study used a quasiexperimental design for the quantitative phase, and an interpretive case study design fo r the qualitative phase to investigate a commercial interactive sing-to-read program, Tune Into Reading, with middle school students in a music classroom. This seven-week descriptive study used a smaller sample size to maintain a high degree of fidelity to treatment, and to include measurements (pretest and posttest) for both gr oups (treatment and control). To qualify for inclusion in the study, st udents were in grades six through eight and were a part of the elective Wheel Musi c Class during the fourth quarter of the 20062007 school year (March 12, 2007May 31, 2007). The Wheel Music Class is an assigned elective class of new cohorts (a mix of sixth through eighth grade students) each quarter of the school year. The school year is divided into four quart ers starting at the end of August and running until the end of May. There were four intact Wheel Music Cl asses during the fourth quarter of this school year for this music classroom. Ra ndomly assigning each individual student in intact curriculum classes to a treatment and control was not an option in this study (e.g., teacher lesson formats, scheduling, various gr ade levels). Therefore, participants were randomly assigned to treatment or control conditions by classes. Prior to assigning each of the classes to treatment or control conditions, the numbers one through four were written on a piece of paper and placed in a bowl. A nonparticipant teacher from the school made four quick picks, alternating assignment for treatment then control. This way each cla ss had an equal chance of being assigned to

PAGE 114

101 either treatment or control. Classes one and three were assigned to receive the experimental treatment and classes two and four were assigned to the control. The classes were then combined. Classes one and three became the treatment group, and classes two and four became the control group. Although the treatment and control groups were randomly assigned as classes, this told us little about the characteristics of th e individuals within each group. In order to answer the research questions and compare the two groups, it was necessary to match as many of the sample characteris tics of the subsets as possibl e prior to the experimental treatment. Sample Characteristics Many variables contribute to unde rstanding how and why students perform during the complex process of reading. The cont rol of all variables that contribute to understanding the outcomes in r eading performance for these two groups is not within the scope of this study: therefore, it was nece ssary to provide information that matches characteristics of the two groups so that they could be compared prio r to the experimental treatment. A total of 64 students ages 12 to14 participated in this study. The treatment and control groups had 32 students each. Initia lly, the treatment group had 33 students, whereas, the control group had 35 students. Tw o students, one treatment and one control moved. In addition, one student in the control group chose not to participate in the study. The changes in the total number of participants occurred prior to th e pretest or any data collection. Therefore, these three cases were dropped from the study. In addition, attendance was taken for each session (14 sessions, 2 times a week, for 7 weeks) in both

PAGE 115

102 groups. The music teacher provided a make-up session time available for students in both the treatment and control group each week. A total of six students (four treatment participants and two control participants ) missed one session during this seven-week period, and all six students voluntarily ma de up the time during a make-up session. It was originally assumed that each of the Wheel Music Classes would have a cohort of 6th through 8th grade students in each class, because of the inter-grade level structure of the elective cla sses at this school. However, after randomly assigning the students to a treatment or c ontrol groups, there were no 6th graders in either subset. In the treatment group 34% of the students were 7th graders and 66% were 8th graders, whereas, the control group had 33% 7th graders and 67% 8th graders. Gender is a crucial variable for early a dolescent literacy learners. Males and females bring different discourse styles and ways of understa nding literacy to the middle school classroom (e.g., Moje, 2000). In this study, the treatment group had 37% females and 63% males, whereas, the control group had 41% female students and 59% males. Along with gender, the other classification variables (ethnicity, language, exceptional learning needs, and social economic status) in fluence adolescents’ literacy development and their understanding of what they read and how they approach reading in school (e.g., Phelps, 2005). The ethnic background of the students wa s predominately White (81% in the treatment group, and 78% in the control group) African American students accounted for 6% of the treatment population and 9% for the control, and the percentage of Hispanic students was 13% for both groups. Students iden tified as receiving services to support their learning, Exceptional Student Education (ESE) or language needs English Language

PAGE 116

103 Learners (ELL) were: (a) 6% for ESE st udents in both groups, and (b) 3% for ELL students in both groups. Stude nts’ of poverty, low socioec onomic status (SES), is a critical issue for reading achievement. Some researchers contend that the academic achievement gap in reading is influenced by social, familiar, and economic factors. Allington (2002) asserts we hear more about the Black/White achievement gap or the urban issues in America schools and yet the rich/poor gaps in achievement are larger. In this study 72% of the treatment group student s were considered to be of low SES (determined by free or reduced lunch programs), and 28% were considered to be of high SES. Whereas, 75% of the control group were low SES, and 25 % were considered high SES. Table 3 presents the percentages of cl assification variables for the students in the treatment and control groups. Table 3 Students’ Classification Variables Percentages by Treatment and Control Group Gender Grade Level Ethnicity ESE ELL SES Male Female 7 8 WhiteBlack-Hispanic Low-H igh Treatment 63% 37% 34% 66% 81% 6% 13% 6% 3% 72% 28% (n=32) Control (n=32) 59% 41% 32% 68% 78% 9% 13% 6% 3% 75% 25% A cursory examination of Table 3 of the percentages comparing the classification characteristics of both treatment and control grou ps, appear to suggest that that the groups are predominantly White low SES students. Male 8th graders represent a larger proportion for treatment and control groups then their female counterparts, or 7th grader peers. In addition only a small percent of the adolescent s receive support serv ices for learning or language needs. However, it is important to assess if there are any significant differences

PAGE 117

104 between the sample characteristics of the two groups. Therefore, Chi-square tests at an alpha level of .05 were used to analyzed diffe rences in gender (male and female), grade level (7th and 8th grades), ethnicity (White and Black ), and SES (low and high) for the treatment and control groups. Th e results indicated that the proportions of classification characteristics do not differ significantly across groups, reported as: (a) gender, x2(1) = 0.0656, p = .7978 (b) grade level, x2(1) = 0.0709, p = .7901, (c) ethnicity, x2(1) = 0.2196, p = .6393, and (d) SES, x2(1) = 0.0801, p = .7772. These results verify that the treatment and control groups displa yed homogeneity in the proportions of the classification variables, of ge nder, grade level, ethnicity, and socioeconomic status. The matched characteristic s of the sample groups provide useful information however; it did not address th e research questions or provi de needed information, about comparing the characteristics of reading pe rformance for each of the groups. Prior to conducting pretests for both the treatment a nd control groups, each group was stratified by FCAT level reading scores. The primary purpose of the FCAT in readi ng is to assess student achievement of higher order thinking skills (Florida De partment of Education, 2005). FCAT level reading scores range from highest score (level 5) to lowest score (level 1). The scores for the treatment and control groups were stratif ied according to thei r FCAT level as: (a) Level 4 and 5 above grade level, (b) Level 3 at grade level, and (c ) Level 1 and 2 below grade level. When this was accomplished a percentage was noted for each level by treatment and control groups. Table 4 displa ys the percentages by groups stratified by FCAT level reading scores as, above level, at level, and below level.

PAGE 118

105 Table 4 Group Percentages of Students FCAT Level Scores Group Above Grade Level At Grade Level Below Grade Level Treatment (n=28) 29% 42% 29% Control (n=28) 29% 42% 29% ________________________________________________________________________ *Note each group was missing FCAT scores for some me mbers: Treatment Group (-4) and Control Group (-4) The percentages showed an equal distri bution of FCAT level reading scores between the two groups, however, a concern wa s the missing reading scores for some of the participants. In the treatment group four students did not have FCAT level reading scores, whereas, in the control group four st udents did not have r eading scores, also. A goal of this study was to understand and comp are students of varying reading ability during the literacy task of rereading through singing. Conseque ntly, it can not be assumed that they are compatible groups based on missi ng data, which could hi ghly influence their .scores in reading. In addition, reading is a very complex process. Using FCAT reading level scores alone does not pr ovide sufficient information a bout the reader. As noted in the literature review Amrein and Berliner (2002) overa ll contend that “ there is no compelling evidence from a set of states wi th high-stakes testing polices that those policies result in transf er to the broader domains of knowledge and skill for which highstakes test scores must be i ndicators” (p.54). Therefore, the us e of a high-stake test scores alone can not account for the many variables associated with unders tanding the reading process and relating that to th e characteristics of this group of early adoles cent literacy learners and their fluent reading behaviors. Accordingly, it was necessary to conduct an analysis using reading pretest scores fo r both the treatment and the control groups.

PAGE 119

106 Four Analyses of Variance (ANOVAs) at a .05 alpha level, were conducted to compare scores from the QRI-4 pretest for the treatment and control groups in fluency (wpm), word recognition, reading comprehens ion, and instructional reading level. The results of the analysis revealed no statistical significance difference in pretest reading scores for the treatment or the contro l groups in fluency (WPM) p= .196, word recognition (WR) p=.180 read ing comprehension (COMP) p=1.00, or instructional reading level (RL) p=.720. Tabl e 5 provides a summary the de scriptive statistics for the treatment and the control groups’ QRI-4 pretest. Table 5 Summary of the Descriptive Statistics of the Qualitative Reading Inventory Pretest Scores Treatment Control (n=32) (n=32) Mean SD Skewness Kurtosis Mean SD Skewness Kurtosis Fluency 125 32.9 0.068 -.706 136 36.09 -0.604 -0. 152 Word Recognition 0.98 0.01 -.384 -1.55 0.98 0.02 -1.84 1.83 Comprehension 0 .77 0.04 2.24 3.36 0.76 0.03 1.78 6.07 Reading Level 5.45 1.17 -0.10 1.65 5.58 1.22 0.35 -0.47 ________________________________________________________________________ In conclusion, prior to experimental tr eatment, the treatment and control groups displayed homogeneity in proportions of the classification variable s, of gender, grade level, ethnicity, and SES. In addition, the groups were no st atistically different on FCAT reading level scores. Furthermore, there were not significant differe nces in the pretest scores of the QRI-4 in fluency (wpm ), word recognition, comprehension, and instructional reading level prior to experimental treatment.

PAGE 120

107 Ethical Considerations I considered several ethical considerati ons before collecting the data, during the data collection, after the data were collected, and on completi on of the research project. Prior to Data Collection: Permission from the school and teacher where the study occurred was obtained. The study was reviewed and authorized by the Institutional Review Board from the University. Informed consent forms were used to obtai n assent from the child and consent from the parents. Parents of the participants were sent a letter explaini ng the study and the role their child would play as a participant. No name s were used that identify the children or their school. I provided my telephone numbe r if any particip ant had questions. Along with the letter and the informed consent form, all participants were informed during a meeting that the st udy was voluntary. The participan ts would let me know if they did not wish to continue or in the case of the child, the parent or teacher would advise me if the child no longer wi shed to be a part of the study. During data collection Data were backed up regularly using coded disks. Security codes were in place to control access to the data. All data were stored in a locked file cabinet at the university. Completion of the Project All the field notes and data were kept secure. No identifiers (names, schools) of the participants were used in any written report.

PAGE 121

108 All research material will be kept for three years. When the data are no longer needed it will be shredded, electronic data will be destroyed. Instruments Qualitative Reading Inventory-4 The Qualitative Reading Inventory-4 (Q RI-4) was used at two points in time to investigate the impact of usi ng the alternative text program Tune Into Reading compared to the regular music curriculum. The following is a summary of the reliability and validity of the QRI-4 scores taken from the technical development report (Leslie & Caldwell, 2006). In addition the scoring procedures used are described. Validity and Reliability of the QRI-4 Scores The QRI-4 is intended to determine inst ructional reading levels for students and for diagnostic purposes (strengt hs and weaknesses in their re ading) to determine fluency, word recognition, and comprehens ion. Therefore, the crucial te st properties to determine reliability and validity are consistency, constr uct representation, and penetration (Cross & Paris, 1987). Consistency relates to the reliability of the QRI-4, and construct representation and penetration relate to the validity of the test. The QRI-4 measures consistency of scores in three ways: inte r-scorer reliability, internal consistency reliability, and alternate-form reliability. Leslie and Caldwell (2006) wanted to investigate whether the QRI-4 was consistent across examiners, to ensure that differences in judgment did not affect the consistency of the examiners’ ratings. They used three expert scorers with master’s degrees in reading and scorers who did not have extensive traini ng in the subject. The

PAGE 122

109 judges scored 304 student passages for reading levels and agreed on 299 of them, for an inter-scorer reliability of .98, indi cating a high degree of consistency. The internal consistency reliability or how well the score is representative of a student’s true reading comprehension was also assessed. Cronbach’s (1951) alpha reliability indicated a high degree of consis tency (.98) for comprehension. The standard error of measurement (SEM) should be betw een .00 and .25/n(i)-1, with lower numbers being more desirable (Crocker & Algina, 1986). The SEM was found to be between .10 and .23 for each passage and grade level. The reliability increases and the SEM decreases when students complete two passages of the same type (e.g., two narrative passages),as the number of similar questions th e student must answer rises. Alternative-form reliability methods were used to determine the consistency of test results over time or cond itions, in order to ensure student s were placed in appropriate instructional levels. This was accomplished by having students read two similar passages (e.g., two narrative passages). The reliabilities of the instructional-leve l decisions were all above .80, and 75% were above or equal to .90. These methods also found that 71% to 84% of the time the same instructional le vel would be found on both passages, according to the comprehension scores for each passage. Leslie and Caldwell wanted consistency in the QRI-4 ability to successfully illustrate the student’s strengths and weaknesse s. The QRI-4 would be considered reliable in this regard if a student performed sim ilarly when orally reading a passage and on a word list of a comparable le vel of readability. Two examin ers independently scored 108 students to determine their level of word recognition and comprehension. The scorers agreed on the diagnostic category for the abili ties of the students 87% of the time. When

PAGE 123

110 the judges did not agree, it was generally wh en the student’s patterns of strengths and weaknesses were unclear. Another concern was that the QRI-4 shoul d be sensitive, or responsive, to both immediate and long-term changes in students’ abilities. Leslie and Caldwell examined the changes in students’ reading abilities by a ssessing them over a four-month period. They found the QRI-4 could successfully meas ure change in word recognition and comprehension over this short time period. Longitudinal studies were also completed, over both the course of one sc hool year, and over several school years. Researchers found that the QRI-4 was also sensitive to change s in abilities over a l onger period of time. Content-validity evidence speaks to the ex tent to which the sample of items on a test is representative of some defined domain or cont ent (Ary, Jacobs, & Razavieh, 1996). Researchers evaluating the QRI-4 wanted to represent the field of reading in a systematic manner that reflected research findings as well as classroom practice. To accomplish this approach, researchers include d both narrative and expository passages for a wide range of levels, from pre-primer to high school. The passages at the beginning levels include pictures so they represent age-appropriate materials children generally encounter. However, reading research shows th e importance of prior knowledge when reading and the significance of miscues in oral reading that alter the meaning of the passage as compared with miscues that do not (Snow, 2004). To provide for these findings, researchers included a measure of prior knowledge in the QRI-4 and two ways to score oral reading accuracy.

PAGE 124

111 Researchers also provided three ways of measuring comprehension, which include the use of explicit que stions, the use of implicit qu estions, and retelling. Word lists contained words that could be figured out using the rules of the English language and words that could not be figured out becau se the spellings were irregular. The QRI-4 also provides a way to evaluate a studen t’s oral reading fluency by measuring the student’s correct words per minute when read ing aloud. Researchers in cluded all of these factors in order to create a valid test th at fully covers the domain of reading. Criterion-related validity was measured by comparing students’ instructional level based on the QRI-4 with students’ equivalent scores on standardized reading tests, including the California Achievement Test, the Iowa Test of Basic Skills, and Terra Nova tests. The researchers examined the correlati on (within grade) between the instructional level obtained from the QRI and the stude nt’s national curve equivalent (NCE) or standard score on a group administer ed standardized reading test. The standardized test data for grad es 1 through 3 were obtained from the California Achievement Test or the Iowa Test of Basic Skills. The standardized test data from grades 4 through 9 were the Terra Nova Test. Statistically significant correlations were found between the instructi onal level in narrative texts a nd standardized tests scores for all grade levels. Table 6 displays the correlation between the instructional level obtained from the QRI and the students’ scor es on the various standardized tests.

PAGE 125

112 Table 6 Correlation of Instructional Level Qualitative Reading Inventory Scores and Standardized Tests Scores by Grade Level Sample Total Grade Level Correlations Standardized Test n=50 206 1 .85 California Achievement/ n=32 2 .65 Iowa Test of Basic Skills n=39 3 .55 Grades 1-3 n=31 4 .66 n=35 5 .44 n=21 6 .27 Terra Nova n=17 7 .43 Grades 4-9 n=22 8 .47 n=19 9 .52 Leslie and Caldwell assessed construct vali dity by determining whether the QRI-4 successfully measured word-recognition ability and comprehension. Expectations were that word identification, oral reading accuracy and reading ra te would be strongly related to comprehension when dealing with be ginning readers, while prior knowledge of concepts in the passage would be connected with comprehension with more advanced readers with a higher level of word rec ognition. Researchers found word identification from word lists; oral reading accuracy, sema ntically acceptable accuracy rate, rate of reading, and corrected rate we re positively correlated and st atistically significant through the 3rd grade from .34 to .59. Statistically si gnificant correlations between prior knowledge and comprehension existed from the primer level and above, but correlations were much stronger above the 3rd grade. The correlations from the primer level to the 2nd grade level ranged from .18 to .30, while the correlations from 3rd grade to middle school ranged from .35 to .86.

PAGE 126

113 The researchers also found students’ comp rehension at beginning reading levels was best predicted by the per centage of miscues that do not change the meaning of the passage and whether they read narrative or expository passages. At more advanced reading levels, researchers f ound comprehension was most successfully predicted by the reader’s background knowledge of the concept being presented. Qualitative Reading Inventory Administ rating and Scoring Procedures The QRI-4 is an informal reading invent ory that provides grade level word lists, and narrative (litera ture) and expository (science, social studies, historical) passages for pre-primary through high-school reading leve ls. The choice of using narrative passages for the participants at pretest and posttest in this study came as a result of reviewing the technical report which provided support for the validity and reliability of only the narrative genre. All passages in the QRI-4 are assign ed ordinal numbers corresponding to readability levels (e.g., 1st grade reading level). However, that was not the case for upper middle school (7th and 8th grades) and high school (9th and 10th grades). They are labeled as upper middle school and high school w ith no corresponding readability levels. Instructional reading level is a dependent va riable in this study and therefore it was important to determine the readability leve ls for the middle and high school narrative passages. A Fry (1979) readability analysis co-sco red with another literacy expert and approved by a university literacy professor was calculated re sulting in a readability level of 7.5 (7th grade 5th month) for the upper middle school passages, and 9.5 (9th grade 5th month) for the high school passages. Theref ore, when calculating the instructional

PAGE 127

114 reading levels for all participants in the study, all readi ng levels were ex trapolated (e.g., 6.0 sixth grade zero months) so that all the scores could be commensurable. The purpose of this instrument accordi ng to Leslie and Caldwell (2006) is to determine: (a) timed automaticity of words in context (fluency), (b) accuracy of oral reading (word recognition), (c) the level of understanding in reading by answering explicit and implicit question (comprehensi on), and then (d) a reading level by combining word recognition and comprehens ion level scores (i nstructional reading levels). Leslie and Caldwell c ontend that unlike other reading inventories this instrument has extensive piloting with approximatel y 1,000 students at multiple grade levels. The administration of this assessment begi ns by determining the appropriate grade level passage for the individual students. Th e authors recommend that using either the graded word list provided in this assessment or any extant data, which approximates their reading level. In this study the FCAT reading level scores were used to approximate the appropriate the beginning read ing levels for assessment for two reasons: (a) it addresses the second research question of this study c oncerning the comparison of the relationship with reading performance and FCAT levels, and (b) the primary purpose of the FCAT in reading is to assess student achievement of the higherorder thinking skills (Florida Department of Education, 2005). Therefore it was assumed that a student who attained a higher FCAT level score in reading (e.g., leve l 4) would be above grade-level peers in reading. FCAT level reading scores (level 15) ranging from highest score (level 5) to lowest score (level 1) were used to determ ine the grade-level passa ge to start with the students. For that reason a student in grade seven who scored at a level 2 in his or her FCAT level reading score first passage would start with a 6th grade reading level.

PAGE 128

115 Students however, were given as many passages as necessary until they reached frustration in order to determine their highest instructional reading level (described below). Scores from the reading instrument are calculated for the separate components of reading as (a) fluency measured by the rate the student reads the words per minute, (b) word recognition measured by oral readi ng accuracy, (c) comprehension measured by implicit and explicit questions after the reading, and (d) instructi onal reading level determined by combining level scores fr om word recognition and comprehension questions. The following describes the scori ng procedure for each of the components. The administrator goes over the procedur es for the assessment with the student. The student and administrator both have a copy of the passage however, only the administrator has a copy of the comprehensi on questions. Reading ra te is calculated to determine automaticy in fluency. The administrator uses a timer with a second hand noting the student’s start and end times on the assessment. To obtain the reading rate in words per minute the following formula is us ed: WPM= (number of words in the passage x 60) / divided by the number of seconds it took the students to read the passage. Word recognition is measured by the num ber of miscues in the student’s oral reading. Miscues are mistakes the student makes by substituting, omitting, or inserting words, or if the administrator tells the student a word because he or she does not know it. The administrator circles mistakes on his or her copy of the passage while the student orally reads their passage. When a student self-correct s or repeats words or phrases this is not considered an error. However, the ad ministrator of the as sessment should note the self-correction because it provides evidence of comprehending or in some cases offers

PAGE 129

116 evidence of struggling with th e passage. In addition, an omission of an entire line by a student is counted as one miscue because it is considered as a loss of place. At the end of each passage the administrator counts the number of miscues, and the results determine whether the performance reflect s an independent, instructiona l, and /or a frustration level in a student’s word accuracy in reading. The total accuracy for reading level performance in word accuracy is (a) indepe ndent levelreads text with 98% accuracy, (b) instructional level-reads text with 90% to 97% accuracy, a nd (c) frustration leve l – reads text below 90% accurately. A chart after each passage pr ovides the number of miscues designated for each reading level. To determine percentages for word r ecognition in reading, the administrator subtracts the number of miscues from the num ber of words in the passage (total words are listed at the bottom of each passage). This yields the number of words read correctly. Then the administrator divides the number of words read correctly by the number of total passage words, rounding up to find the percentage of total accuracy. Comprehension is measured by the student s’ responses to e ither eight or ten implicit and explicit questions asked after the reading. Only the administrator has a copy of the questions. The questions are scored as either right or wrong, and under each question the correct responses are provided to the administrator. At the end of each passage the administrator counts the number of correct responses and the results determine whether the performance reflects an independent, instructional, and /or a frustration level in the student’s comprehens ion in reading. The total correct responses for reading level performance in comprehe nsion are (a) independent levelanswers questions correctly 90% or a bove, (b) instructional levelanswers questions correctly

PAGE 130

117 67% 89% and (c) frustration level – answers questions be low 67%. A chart after each passage provides the number of correctly answered questions needed for each reading level. To determine percentages for comprehe nsion in reading, the administrator divides the correct responses by the total number of questions. Instructional reading level is determined by the combination of word recognition level plus comprehension reading level, on a particular grade level passage. Therefore, a student who reads a 6th grade passage and scores at the independent level for word recognition, and the instructional level in comprehension, would represent a 6th grade instructional reading level. Ta ble 7 displays how the combin ations of levels determine the students reading level. Table 7 Determining Instructional Reading Levels from the Qualitative Reading Inventory Word Recognition + Comprehension = Total Passage Level Independent + Independent = Independent Independent + Instruc tional = Instructional Independent + Frustrat ion = Frustration Instructional + Indepe ndent = Instructional Instructional + Instru ctional = Instructional Instructional + Frustr ation = Frustration Frustration + Inde pendent = Instructional Frustration + Instructional = Frustration Leslie and Caldwell (2006) recommend that if the assessment is being used as a pretest/ posttest measure, that the posttest passage should be at the same instructional level attained during the pretest. Then the administrator continues to test the students

PAGE 131

118 until they reach frustration. One level above fr ustration is their new instructional reading level. Florida Comprehensive Assessment Test The primary purpose of the FCAT is to assess student achievement of higherorder thinking skills for reading, writing, mat h, and science. Students take the FCAT in grades 3 through 11. In grades 4, 8, and 10 student s take the writing portion of the test. In grades 5, 8, and 11 students take the science por tion of the FCAT, and students in grades 3 though 10 take the reading and mathematics por tions. Florida Comprehensive Assessment Test (FCAT) level 2006 scores in reading were used in this st udy to divide students before treatment in three groups. The students in this study were divided into groups for the purpose of data analysis based on their levels as “below, at, or above” in reading. The following is a summary of the reliability and va lidity of the FCAT level scores as reported by the Florida Department of Education (FCAT Briefing Book 2005). Reliability and Validity of the FCAT Scores Criterion-referenced tests ar e designed to identify an individual’s status with respect to an established st andard of performance. For the FCAT, these established standards are the Sunshine State Standard s. The FCAT’s secondary purpose is to compare the performance of Florida student s with students across the nation, which is accomplished by using a norm-referenced test (NRT) for reading and math. The current NRT is the Stanford Achievement Test 10 ( SAT 10 ), published by Harcourt Assessment, Incorporated, 2005 A research based norm-reference achievement test provides information on student performance based on its nationwide standardization program conducted in the spring and fa ll of 2002 on the K-12 population.

PAGE 132

119 The 2003 Florida legislatur e enacted HB 915 that required the Department of Education to determine the score relationships of the SAT, ACT, PSAT, and PLAN to the FCAT. They conducted concorda nce studies, a technical proc edure for converting scores from one standardized test to another. Th e study was based on students who had taken the FCAT in spring of 2000 or 2001 and had taken one of the other four tests. They found positive correlations between FCAT scores and the scores on the other four tests, all within the range of correlations between thos e of the SAT and ACT. The state of Florida had by far the strongest correlations, with a .96 correlation between high and low stakes test score levels and a .71 correlation between the year-to-year gains on high and low stakes tests (Florida Depart ment of Education, 2006). The degree of difficulty of FCAT items is categorized in two ways – by item difficulty and cognitive complexity. Item di fficulty consists of two meanings. Before testing, it is the prediction of the percen tage of students who will choose the correct answer. After testing, it is the percentage of students who actually chose the correct answer. When 70% of the students chose the correct answer items are categorized as easy. When 40-70% of the students answered co rrectly items are considered average, and challenging questions are answered correc tly by fewer than 40% of the students. The cognitive complexity refers to the cognitive demand associated with each item. This is currently determined using a system based on Webb’s (2002) work related to the Depth of Knowledge Levels. Webb developed four levels of cognitive complexity as an alignment method to examine the consis tency between the cognitive demands of the standards and the cognitive demands of the assessment. Bloom’s taxonomy was previously used to determine the cognitive complexity, but it was found to depend too

PAGE 133

120 much on the abilities and prior knowledge of th e students as opposed to the expectations of the items. Therefore, the cognitive comp lexity classification no longer relies on the student’s approach to the question bu t on the actual test item itself. After a student takes the FCAT in reading and mathematics, the student receives a developmental scale score that ranges from 0 to 3000. These scores provide additional information to help interpret scores from th e FCAT Sunshine State Standards (SSS) test. Developmental scores are used because simply looking at the scale sc ores that the FCAT reports, which range from 100 to 500, do not re flect students’ progre ss within a level. Students should receive higher developmental sc ores as they move from grade to grade according to increased achievement. Since reading and mathematics are tested every year, this score is used to help parents and schools understand st udents’ year-to-year progress. Based on the developmental scale score, the student is then assigned one of five Achievement Level Classificat ions ranging from 1 to 5. A level 5 score indicates the student ha s had success with the most challenging content of the SSS and has answered most of the test questions correctly, including the most challenging questions. Students who ear n a level 4 score have had success with challenging content of the SSS, and have answer ed most of the test questions correctly, but may have had only some success with questions concerning the most challenging content. Level 5 and 4 are considered above grade level in reading. A level 3 score means that the student had partial success with th e challenging content of the SSS, but their performance is inconsistent. They may have answered many of the test questions correctly, but they are gene rally less successful with th e most challenging questions. Level 3 denotes meeting the basics for the grad e level or at grade level. Students at this

PAGE 134

121 level are considered on grade level in read ing and mathematics. A student who earns a level 2 score has had limited success with the challenging content of the SSS. A level 1 score indicates little success with the chal lenging content of the SSS. Both Level 2 and level 1 are considered below grade level a nd not meeting grade level expectations. Reliability of the Data The following section reports how I ensu red the information in the concurrent mixed method study was reliable. The quantita tive phase addressed the measure taken to address reliability through m easures of internal consistency and interrater reliability. Internal Consistency Reliability The reliability of a measuring instrument is the degree of consistency with which it measures whatever it is supposed to m easure (Nunnally, 1978). One way to measure reliability involves assessing a test’s internal consistency, the extent to which all test items are measuring the same thing. Cronbach’s alpha is the most common estimate of internal consistency of items in a scale. Alpha measures the extent to which items responses obtained at the same time correlate highly with each other. However, when items are dichotomously scored, as in this study, as right or wrong (0 and 1) Kuder – Richardson 20 (KR20) is used to assess a test’s internal consistency. Kuder and Richardson devised a procedure for estimating the reliability of a test in 1937. It has become the standard for estimating reliability for single administration of a single form. Kuder-Richardson measures inter-item consiste ncy. It is tantamount to doing a split-half reliability on all combinations of items resultin g from different splitt ing of the test (Sapp, 2006).

PAGE 135

122 In this study, comprehension reading scor es for 10 students (5 treatment and 5 control) consisting of 10 questions were labe led as right (1) or wr ong (0). The alpha was computed for internal consistency on the 10 students followed by internal consistency measures for the 5 treatment and the 5 contro l groups separately. The raw coefficients for each of these variables were .75, .72, and .70 respectively. Nunnally (1978) suggests .70 as an acceptable reliability coefficient; sma ller reliability coefficients are seen as inadequate. These numbers are considered satisfactory following Nunnallys’ guidelines and indicate that for these variables, the te st scores in reading comprehension had an acceptable level of internal consistency. Interrater Reliability Training and Scoring In addition to internal consistency, anothe r reliability issue is the consistency of scoring of test items. To measure the extent to which I accurately and reliably applied the scoring criteria from the Qualitative Reading Inventory-4 (QRI-4) for fluency, word recognition, comprehension, and reading level, a stratified random sample of 20 students (10 treatment and 10 control), at pretest a nd posttest were double-scored. Prior to any work completed by the second scorer, I c onducted two training sessions. The second scorer was a literacy education professional wi th extensive experience in reading content and pedagogy. In addition, she has for the last three years used the QRI-4 in the field with me on various research projects. The first session explained the procedur e for co-scoring with a student. Since the co-scorer was familiar with the instrument, th e first session developed the procedures we followed in the field. The second session wa s a practice session with the procedures for co-scoring with a student which included look ing at rate, measuring words per minute for

PAGE 136

123 fluency, word recognition, miscues (number of mistakes made by the students), comprehension questions answered correctl y, and assessment of instructional reading level based on the scores from word rec ognition and comprehension. After the scoring session was complete, both the scorer and I ca lculated the assessment independently and then discussed any differences in scores. The scorer then went out into the field on two occasions during pretest and posttest. Ten students were selected using a stratified random sample from the treatment and control groups. The same students, select ed at pretest were co-scored during the posttests. Two Pearson correlation coefficients were calculated on tw o of the dependent variables of this study, fluency and word recogn ition to investigate the relationship of the scores between the co-scorer and researcher. The fluency scores a nd the word recognition scores were both highly corre lated r=.999. The correlation results for word recognition also showed a strong relationship that was signif icant r = .943. Procedures The following section describes the proce dures used for the tr eatment and control groups during the literacy task of rer eading through singing. However, before a description of the procedures for both of the groups, a discussion of my pre-study involvement with the intera ctive sing-to-read program Tune Into Reading is necessary because the protocol, developed from previous research, was used in the current study. Pre-Study Involvement Developing th e Protocol for the Current Study Over the past three years I have b een involved in several quasi-experimental studies investigating the impact of the interactive sing-to-read program, Tune Into Reading (previously referred to and adapted from the Carry-A-Tune program).

PAGE 137

124 Interestingly, the program was designed to improve singing; however the developer of the interactive sing-to-read program, Tune Into Reading received a call from a parent of a middle school student who struggled in her r eading suggesting to the developer that the use of the program improved the student’s r eading. The developer brought the program to The University of South Florida and asked a literacy professor if a study could be conducted on this assumption. I was assigned as a research assistan t to conduct a pilot study. The purpose of the initia l study and the following re plication studies were to investigate the impact on reading performance measured by the Qualitative Reading Inventory of the students who used this singing program comp ared to their counterparts who did not. A total of four hundred west central Florida struggling readers ( the struggle was determined by FCAT scores levels 1 and 2) from three school distri cts in grades four through twelve were participants over the last three years. The initial study ( n =48) was conducted in a middle school music classroom for 9 weeks, 3 times a week, for 30 minutes each session. During this study I developed a protocol for use with the sing-toread program that was used in the curre nt study. This protocol was adapted from Samuels’s (1979) theoretical recomme ndations for building reading fluency. The program Tune Into Reading uses a vocal-range anal yzer that tracks the singer’s pitch and rhythm, compar ing it to the correct pitch of the song. Each student uses a microphoned headset linked to the computer to sing along repeatedly and to record his or her singing. Following Samuels’s (1979) theoretical recommendation for building reading fluency with struggling readers, I developed a protoc ol for treatment using this interactive sing-to -read program.

PAGE 138

125 Samuels (1979) recommends: (a) students be provided a model of fluent reading, (b) reading material should be at their in structional reading leve l (reading with 90-94% accuracy), (c) practice rereading the material at least three times independently, and then (d) orally read the passage for asse ssment and feedback. Following these recommendations I adapted the r eading fluency protocol to m eet the needs for this study. The first recommendation was to provide a model of fluent reading. In this study the students had background music with word s (broken into syllabl es) emphasizing pace, pitch, volume, rhythm, and tone. This provide d a model of reading fluency specifically, relating to prosody of text. Then, as r ecommended by Samuels (1979), the reading material used to build fluency should be at the students’ instru ctional reading level (students can read passage with 90-95% accura cy). There were 24 songs on this program, and to determine readability for the songs a literacy professor a nd I co-scored all the songs. We both had individual copies of the song lyrics and independently scored the songs for readability levels using the Fry (1979) readability formula. When this was accomplished we compared each song and if there was any disagreement we discussed it and made the appropriate adjustments. In the end each song on the program had a readability grade level so that the students could sing songs at thei r instructi onal reading level. The Fry readability formula is calc ulated by the averag ing the number of sentences and syllables per hundred words. Th ese averages are plo tted onto a specific graph, and the intersection of the average number of sentence and average number of syllables determines the reading level. Figur e 3 is a copy of the graph used for the Fry (1979) readability formula.

PAGE 139

126 Figure 3 Fry Readability Formula Graph (1979)Copyright Free Instructional reading levels were determined from leveled passages using the Qualitative Reading Inventory administered to all 48 particip ants and served as pretest, posttest, and follow up measures of assess ing comprehension and reading levels. Therefore, when the students sang their songs they used material on their instructional reading level. Once the students’ instructi onal reading level was determined, and the songs (reading material) were at the stude nts’ correct level, Samuels recommended rereading the passage at least three times. Prior to using the software each studen t enters the signs-in component of the computer. After typing in their name all data collected for the student became permanently stored into his or her personal po rtfolio on the computer specifically, all the students singing scores, record ings, and their individual voc al range. In order for the students’ to get his or her vo cal range they record themselves singing at their highest vocal level followed by their lowest vocal level, holding a singl e note or vowel sound (e.g., do or ah). Then the program calculates the vocal range by combining the high and

PAGE 140

127 low levels. As a result all of the songs th at the students sang were at the individuals’ appropriate reading level. The sing-to-read program has two different textual formats for rereading. The first text format, linear sheet music, allows the stud ent to read the lyrics silently three times, while listening to the background music and tempo. This aligns with the recommended number of repetitions suggested by Samuel s (1979). The linear sheet-music view is followed by a graphic textual view, where stude nts record their singi ng. This alternative text format provides a visual display of words broken into syllables without the accompanying musical staff and places each sy llable accented at the appropriate pitch within each student’s personal vocal range. The graphic view of the song that is used to guide students’ pitch matching wh ile they sing selected songs. Along with the visual tracking of the wo rds, a guideline for accurate pitch and tone provides a real time track line of the student’s voice while he or she is singing and recording a song. After singing each time, a score is provided to the student. These scores range from 0-100 on their representation of pitch accuracy and tone for the song. The students in this study sang and recorded the s ongs using the visual graphic format three times aloud. Then the program saved all their re corded versions of their highest score for each song. Therefore, I could review th eir singing and assess their progress. How the Students Used the Tune Into Reading Program The teacher was a veteran music teacher of 20 years. She had used the program for three years and was the same music teach er with whom the protocol was developed. This was important so that teaching can be undisturbed by trying to learn the program. Prior to starting the experimental treatment with the students the teacher reported to me

PAGE 141

128 there were no participants in this study who has used the program. In addition, the students were aware of the selection process and they appeared to be comfortable with how it was handled. Following the protocol fr om previous studies the music teacher introduced the students to the in teractive sing-to-read program. Tune Into Reading However, unlike previous studies the curr ent study was a seven-week treatment and sessions were twice a week for forty-five minutes per session. Using an overhead projector the music teacher presented the Tune Into Reading program to the whole group of students. She went over all the components of the program, showing the students: (a) how to sign-in, (b) how to determine their vocal range, (c) how to use the two different textua l formats, (d) how many times to listen to the song and reread silently, (e) how many ti mes to record their singing, (f) how to interpret their scores and how this represen ts the accuracy of matching the pitch of the song while singing and recording, and finally (g) how to access their individual folder that contained the songs they would work w ith for each week. Then the student went to their individual computers and the teacher had them sign-in and record their vocal range. The teacher walked around and made sure th at the students had this in place. All of the students practiced the flue ncy protocol using the same song Hot Cross Buns. This particular song was used because it has a 2nd grade readability level. Therefore, all of the students were able to read the words of the song while they were learning how to use the program. When the students returned for the next session, they had individual songs in a folder under their name at their individual instructional r eading levels. Instructional reading levels for the students were determ ined through their pretest scores from the QRI-4.

PAGE 142

129 Control Group The same music teacher worked with th e control students for seven weeks, two sessions a week, for 45 minutes per session, no t unlike their counterparts using the singto-read software, Tune Into Reading The students during this seven-week study learned three songs with multiple stanzas, while lear ning and individually playing simple drum rhythms to accompany their singing. The music teacher suggested, “drum circles are a way to build a sense of community in the classroom. They keep the students motivated and engaged in the singing process…. A nd drumming provides a rhythmic background that supports the student while le arning a song” (March 26, 2007). Initially, the music teacher presented the si mple drum patterns to the students. All of the students had individual drums, as did the teacher. She taught the rhythmic pattern and the students echoed the same pattern dur ing the first two sessions. This was followed by teaching a song. The procedure for teaching a song went as follows: 1. The song was presented to the en tire group using an overhead projector. 2. The meaning of the song was discussed along with some pertinent vocabulary words within the song. 3. The music teacher sang the song firs t, and then the students followed along reading the text on the projector screen. 4. The song was broken down by stanzas th e teacher sang first, and then the students echoed her singing for each stanza. 5. Each stanza was sung repeatedly three times.

PAGE 143

130 6. When the song in its entirety was repeat edly sung, the students played the simple drum rhythm while they sang. 7. The music teacher spent two weeks on each of the three songs. 8. The final week was a performance of the students’ singing and playing the drums for a school assembly. Data Collection Quasi-Experimental Design Data Collection Quantitative data collection consisted of administering the QRI-4 assessment to participants in both the alternative text Tune Into Reading program and the regular music curriculum program at two points in time (prete st and posttest). Prior to the experimental treatment and upon approval of the informed c onsent forms, groups by class were assigned randomly to the control and experimental c onditions. One treatment group of 32 students used the alternative text program Tune Into Reading and one control group of 32 students sang as part of their regular music program. Sc ores from the pretest were used to ensure that the students in the expe rimental treatment and control groups were not different in their performance in word recognition, flue ncy, comprehension, and instructional reading level before the experimental treatment. The students were individually tested during their Wheel Music Class periods. Each Wheel Music Class period ran for 50 minutes each day, and each student took approximately 25 minutes to test during these periods. As previously discussed this study included four Wheel Music Classes that were randomly assigned to a treatment or a control condition. The four classes had differe nt class periods each day and there were different numbers of participants in each class. A total of eight students, four from the

PAGE 144

131 treatment and four from the control condition, were tested daily. The total testing time was accomplished in 10 days for all participants All pretests for the 64 participants were completed within two weeks (March 19thMarch 30th, 2007) prior to the 7week experimental treatment (April 2nd –May 15th, 2007). Table 8 presents the schedule of pretests by class period for the tr eatment and control participants. Table 8 Schedule of Pretests for Treatment and Control Participants Class Period Time Treatment/ Control Number of Participants Duration Period 1 7:30-8:20 Treatment Cond ition 12 Participants 6 days Period 2 8:30-9:20 Control Condition 18 Participants 9 days Period 3 9:30-10:20 Treatment Conditi on 20 Participants 10 Days Period 6 1:45-2:30 Control Condition 14 Participants 7 Days After the implementation of the interactive sing-to-read program, I administered a posttest using the QRI-4 and compared the pos ttest scores with the pretest scores to determine if students in the experimental group had gained significantly over their counterparts in the control group. All postt ests for the treatm ent and the control conditions were completed after the 7-week experimental treatment following the same procedures as the pretests (May 17thMay 31st, 2007). Interpretive Case Study Data Collection I developed a schedule of observations fo r the two cases in this study based on the middle school calendar. A total of 14 classroom visits were made in the music classroom during the fourth quarter of the 2006-2007 scho ol year (April 2 May 15, 2007) over the seven week experimental treatment period. As previously noted, four classes were

PAGE 145

132 randomly assigned by class to the treatment and control conditions. Two classes were combined and became the treatment group and two classes were combined and became the control group. Observations occurred twice a week for both the treatment and control groups in all four classes on the same day. Figure 4 depicts a schedule of qualitative observations for both the treatment and control groups. Figure 4 Qualitative Observations Schedule Monday Tuesday Wednesday Thursday April 2 Observations All Classes Observations All Classes April 9 Observations All Classes Observations All Classes April 16 Observations All Classes Observations All Classes April 23 Observations All Classes Observations All Classes April 30 Observations All Classes Observation All Classes May 7 Observations All Classes Observations All Classes May 14 Observations All Classes Observations All Classes Observational field notes were taken during each class session twice a week during the 50-minute class periods for each of the four classes assigned to the treatment or control condition. Field notes were take n on a pad of paper during the Wheel Music Class periods noting time, place, attendance, and all the peer interactions during the observation. These observations focused on de scribing the relations hip, if any, between the literacy task the music teacher assigne d (rereading through singing). Focusing on the interactions (peer talk, peer modeling, and peer social reinforcement) among students

PAGE 146

133 who were singing using the interactive program Tune Into Reading, versus the peer interactions among students who sang in the traditional music class. Strauss (1993) recommended that to a ssist with this difficult process for beginners, researchers should develop a codi ng paradigm. The paradigm, which applies to this study, consisted of: (1) the literacy task (reread ing through singing) assigned by the music teacher and (2) interactions among the peer groups during the literacy task assigned by the music teacher for the two cases (students using the interactive sing-toread program and students in the regular music class. Following a theory suggested by Ryan (2000) there are generally three ways that early adolescents experience peer interacti ons within the contex t of middle school: (a) through information exchange (discussi on), (b) modeling (p eer observation and imitations), and (c) peer pressure (social rein forcement). I used thes e three categories as preliminary coding categories and as a framework to focus my observations. Information exchange refers to discussi ons and talk amongst the peers, capturing direct quotes from the various conversations that the peers excha nged during the literacy task: Peer 1“ How did you get the song to sl ow down” Peer 2 “ Click on this button” (Observational notes April, 7, 2007). Peer m odeling on the other hand refers to the act of peers observing one another that result in ch anges in behaviors or understanding within the student(s). This is achieved by describing the interactions during the literacy task that documents these changes: [He looked around the classroom for two minutes then he smiled and went back to playing the drum s] (Observational notes, April 7, 2007). Finally, peer pressure occurs through social reinforcement, both negative and positive. Descriptions of peers’ accepting or reject ing behaviors exhibited by their counterparts

PAGE 147

134 through body language, facial expressions, smilin g, or laughing during the literacy task: [T hit the drum wrong… M laughed…and then the class laughed…T turned red and put his head down] (Observational notes, April 7, 2007). Ryans’ ( 2000) three categories became preliminary coding categories. They were then put into a matrix that was used for data analysis. Field notes were reviewed daily afte r all the observations were completed. Initially, I would read through the notes three times to get a holistic sense of the data collected. Then the notes were bracketed and coded as one of the three peer interaction categories. Units of data were, conversati ons amongst the peers, or paragraphs that described peers observing or a pplying pressure to other peer s, were bracketed and labeled as one of the three peer interaction cate gories. This was followed by transferring the bracketed notes to a matrix (Appendix B) with the three categories. The matrix was used to ensure that the observati ons did not stray from the focus of the study. Once the data were transferred the difficult j ob of data analysis began. Fi gure 5 provides an example of the matrix used in this study

PAGE 148

135 Figure 5 An Example of the Observational Notes Trans ferred to the Categorical Matrix for the Peer Interactions Information Exchange Peer discussion/talk direct quotes from conversations during the literacy task Peer 1“ How did you get the song to slow down” Peer 2 “ Click on this button” Modeling Peer Observation/ through descriptions of interactions during the literacy task He looked around the classr oom started to smile and went back to playing the drums Peer Pressure Social reinforcement/ descriptions through looks / comments/ laughs during the literacy task T hit the drum wrong, M laughed and then the class laughed T turned red and put his head down, Data Analysis Data analyses were concerned with the research questions an d the integration of the data to meet the study’s design. The quant itative methods used for data analysis are explained first. This explanation is followe d by the qualitative methods used for data analysis. The final section explains how the data were integrated. Quasi-Experimental Design Data Analysis The research question concerned with this phase was: 1. To what extent is the reading perf ormance of word recognition, fluency, comprehension, and instructional reading level, as measured by the QRI-4, of students using the Tune Into Reading program, different from their regular music curriculum counterparts?

PAGE 149

136 2. To what extent does the Tune Into Reading program differently impact the reading scores of students who are “below, at, or above” grade level as determined by the Florida Comprehensive Assessment Test (FCAT) reading scores? Data for the quantitative phase came from the participants’ performance on the Qualitative Reading Inventory-4 (QRI-4) (Leslie & Caldwell, 2006), reading assessments during the pretest and posttest administrations Leslie and Caldwell recommend that if the assessment is being used as a pretest/ posttest measure, that the posttest passage should be at the same instructional level attained during the pretest. Then the administrator continues testing the students until they reach frustration, so that the new instructional reading level can be determin ed. Therefore, analysis for the first two questions was completed utilizing the same instructional read ing level scores attained on the pretests, and then another analysis was completed at the students’ higher instructional reading level if appropriate. All quantitative data analyses were conducted using SPSS soft ware (Statistical Package for Social Sciences) Version 15.0 (Stevens, 2002). The analyses included computation of differences in mean performa nces between the experimental and control group on the QRI-4. Question 1. The first quantitative research question addressed the readers’ literacy performance after using th e interactive sing-to-read program Tune Into Reading as an alternative text and how this compared to the perfor mance of their counterparts who were singing in the regular music class. Prio r to the treatment, I administered a pretest using the QRI-4. Scores from the pretest were examined to ensure that the students in the regular music class and the students using Tune Into Reading were not different in their

PAGE 150

137 performance in fluency (measured by words per minute), word recognition (measured by oral reading accuracy), comprehension (measured by implicit and explicit questions after the reading), and instructiona l reading level (measured by co mbining scores from word recognition and comprehension questions ) before implementation. After the implementation of the interactive singtoread program, Tune Into Reading I administered a posttest using the QRI-4 and co mpared the posttest scores with the pretest scores using their reading level scores from their pretest initially to determine if students in the experimental group gained significantl y over their counterp arts in the control group. This was followed by a comparison of pr etest scores and postte st scores at the higher instructional readi ng level. Then I analyzed the scor es at their higher reading level at posttest if appropriate. Doubly multivariate repeated measures ANOVA at an alpha level of .05 was used to examine the simultaneous differences in the dependent variables fluency (WPM), word recognition (WR), comprehension (Comp) and instructional reading level (RL) on the same instructional reading level attained at the pretest initiall y at two points in time (pretest to posttest). The multivariate repeated measures ANOVA assessed if the combination of noncommensurate (differing measurement scales) dependent variables differ over time and by group. Before analyses were initiated, prelim inary inspections of all variables were completed to check dist ributions (observations outside the normal distribution). Means, standard deviations, skew ness, and kurtosis, were calculated for all continuous variables, and percen tages for all categorical variab les, were derived in order to describe the sample and be able to co mpare results with data from other published studies.

PAGE 151

138 Simultaneous differences reported by the F test statistics from pretests to posttest by group were analyzed first by checking for sign ificant interactions. If the interactions were significant, then comp arisons were conducted using t -tests on each of the dependent variables and determining effect sizes. Initia lly, the scores were analyzed at the same instructional pretest reading level and th en this was followed by a between-groups analysis of variance (ANOVAs) for each of the four dependent variables at the increased reading level posttest scores. Question 2. The second quantitative research que stion investigated whether an interaction effect of the rep eated reading methods occurred on the reading performance of the students “below, at, or above” grade level as determined by the Florida Comprehensive Assessment Test (FCAT) 2006 in reading, while us ing the sing-to-read program, Tune Into Reading, as an alternative text. The results in reading achievement level scores (achievement levels 1 th rough5), according to the state of Florida Department of Education, are reported as (a ) students who scored a Level 1 or 2 are considered below proficiency in meeting grade level benchmarks, (b) students who scored a Level 3 are considered at grade level, and (c) students who sc ored at a Level 4 or 5 are considered above grade level ( FCAT Briefing Book, 2005). The students were grouped by FCAT reading level scores and th en analyses were conducted on the four dependent variables for the three levels. Repeated measures ANOVAs at an alpha level of .05 were used to examine the differences for each of the dependent variab les fluency (WPM), word recognition (WR), comprehension (COMP), and instructional read ing level (RL) at the same instructional pretest level for each of the three FCAT le vels. The repeated measures ANOVA assessed

PAGE 152

139 if each of the dependent variables differe d over time, instructional group, and by FCAT Reading Levels. Before analys es were initiated, means, st andard deviations, skewness, and kurtosis, were calculated for all the continuous variables. Differences reported by the F test statistics from pret ests to posttest by groups were analyzed by first checking for significant group level interaction. If the interactions were significant, then comp arisons were conducted using t -tests and determining effect sizes for each of the dependent variables for the three levels at the same instructional pretest reading level. This was followed by a between-groups analysis of variance (ANOVAs) for each of the four dependent va riables at the increased reading level posttest scores. Interpretive Case Study Data Analysis The analysis of the data required qualitative analysis procedures. Patton’s (2002) guidelines for content analysis recommended r eading through the data at a specific time and making notes in the margins pertaining to specific notions about meanings. Moerman’s (1988) suggestions for conversati on analysis guided the analysis of peer interactions through conversati ons. In addition, Miles and Huberman’s pattern analysis (1994) was used to code data and look for emerging patterns. The data analysis for this case study invol ved a careful review of data gathered from the observations of peer interactions within the treatment and control groups during the literacy task of rereading through si nging. This study consisted of two cases. The experimental treatment group using th e interactive singtoread program Tune Into Reading and the control group singing as part of their regular music cl ass. Therefore, the constant comparative method was used to analyze the data (Glaser & Strauss, 1967;

PAGE 153

140 Patton, 2002). Using constant comparative fo rm of analysis, I be gan the process of analyzing text after e ach observation (Glaser & Strauss, 1967). It involved “continually comparing one unit of data with another in orde r to derive conceptual elements of theory” (Merriam, 2002, p.8). The comparison initially t ook place within each case but eventually moved across cases. My first task involved typing the field not es from the observations of the Wheel Music Classes. The notes were typed-up daily after all the classroom observations so the information could remain fresh. Once this task was accomplished, I began the difficult task of reading and analyzing the data. Firs t, I read the field notes from the classes through three times to gain a hol istic sense of the data. Then I returned to the data and bracketed the categories of peer interactions and labeled them as information exchange, modeling, and peer pressure so th at it could be transferred to the peer interaction matrix (Appendix B). I then read each line of the da ta in the matrix and highlighted units of meaning, patterns where repeated phrase s and or words occurred (Patton, 2002). Construct names emerged from these data. Th e construct names came directly from the data. One example that illustrates how this was done was from a phrase that described peer modeling, “In the four corners of the co mputer lab small groups of females look at one another and start to laugh softly, as they secretly gl anced around the room”. This sentence was highlighted and was bracketed with the construct name, Peer Observation Once in the matrix the data were further analyzed to determine the elements of peer interactions during the li teracy task. After the elements were identified and assigned construct names, they were added to th e Construct Key (Appendix C). I used the Construct Key to be consistent with constr uct names from the emerging data, but also

PAGE 154

141 added any new emerging constructs from th e consecutive observations to the construct key The elements were then grouped according to the construct names. The elements with the construct names assigned were then cut-up and placed in a folder. The frequency of each construct was tallied to determine whether or not an element was emphasized during the peer inter actions. The frequency calculations were followed by organizing the constructs into categories. E ach category of constructs was placed on a bulletin board and further analysis determined the themes that emerged from these data. These themes were presented first as individual cases, then a cross case analysis. I repeated this process fo r 28 observations (14 observat ions for the treatment case and 14 observations for the control case), th en I analyzed these data again with the finalized Construct Key. To ensure that the qu alitative phase of this study is credible, qualitative researchers with b ackground in literacy were uti lized as a second observer and conducted an analysis check of the data. Credibility of the Data The qualitative phase was devoted to addres sing the issues of credibility in this study. Credibility ensures the ac curacy of the data. The researcher is responsible to ensure the truthfulness of the findings and to report the findings with care. Therefore, to address the issue of credibility a second observer was used an d analysis checks were done with two qualified literacy rese archers. In addition, a triangulation strategy for this concurrent mixed methods study is described and also addresses supporting the credibility of this study.

PAGE 155

142 Second Observer A second observer ensured the analysis wa s systematic and verifiable, strategy suggested by experts in qualitative research. This enhances the accuracy of data recording (Miles & Huberman, 1994). In this study a second observer was utilized. The second observer’s was a literacy education professi onal with extensive experience in reading content and pedagogy. In addition, she has a strong qualitative research background. Prior to any observations, I conducted a trai ning session with the second observer. During the training session I discussed the paradigm that applies to this study: (1) Literacy task (rereading through singing) assigned by the music teacher, and (2) Interactions among the peer groups during the literacy tasks assigned by the music teacher. Following a theory suggested by Ryan (2000) that early adolescents experience peer interactions within the context of middle school genera lly in three wa ys: (a) through information exchange (discussion), (b) mode ling (peer observation a nd imitation), and (c) peer pressure (social reinforcement). These ge neral categories were used as a framework to focus our observations in the field. Once in the field we each took observati onal notes with both the treatment and control groups. Immediately afte r the observations a discussion occurred. This helped to ensure I was capturing and accurately recording the peer interactions during the literacy task. Analysis Checks Two qualitative researchers with backgrounds in literacy education and extensive experience in reading content and pedagogy r ead several transcript s. The qualitative researchers checked for credib ility at two points during the qualitative phase of the study.

PAGE 156

143 The researchers were given the Cons truct Key (Appendix C) I developed for coding purposes. The Construct Key included the constructs with descriptions. They were given several transcripts of field notes. One lit eracy expert was given transcripts from the group using the interactive software program and the other literacy expert was given transcripts from the group in the regular si nging class. Their coded transcripts were compared to the same transcripts I coded to determine the clarity of the constructs and definitions. We discussed any areas of disagr eement and reworded descriptions presented in the construct key that were unc lear for a better understanding. Concurrent Triangulation Strategy Triangulation involves both qualitative and quantitative formats to better measure concepts gauged individually (Creswell, 2003). This technique is an attempt to confirm, cross-validate, or corroborat e findings within a single study (Morgan, 1998; Steckler, McLeroy, Goodman, Bird, & McCormick, 1992). In incorporating the two, a researcher can look for or measure data normally associat ed with quantitative methodologies such as outcomes as well as data commonly used in qualitative research such as perceptions (Tashakkori & Teddlie, 1996). In combination, th is strategy can target a larger or more varied series of indicators or data sets usually limited within conventional research formatted studies (Creswell, 2003). In addi tion, “it can result in well-validated and substantiated findings” (Creswell, 2003, p. 217). It also limits the weaknesses inherent in both formats and enhances their strengths as th e diversity establishes a greater reliability and reduces errors or threats. Triangulation of the data occurr ed in Chapter Four of this study.

PAGE 157

144 Integration of the Data Priority was given to the quantitative approach because it looked at the statistical relationship between the treatment group who used the sing-to-read program Tune Into Reading and the control group who were rereadi ng through singing in their regular music class. The analysis for this approach was ex ecuted first to answer the first two questions of this study. However, concurrently quali tative case study methods were used to better understand and describe the peer interactions occurring during the lite racy task assigned by their teacher. The integration of the two types of data occurred during the qualitative findings section of the research project. The quantitative results and qualitative description were mixed in order answer the research questions and to provide a clearer picture Chapter Summary Chapter 3 presents the methods that were used to conduct this study. It outlines the research questions, describes the de sign of the study, and describes the study population and participants. In addition, this chapter delineates ethi cal considerations, instruments, and reliability of the data. Fina lly, it outlines the procedures, data collection, data analysis, and credibility for the study.

PAGE 158

145 CHAPTER FOUR: FINDINGS The purpose of this study was to investigat e the use of an inte ractive sing-to-read program, Tune Into Reading, as an alternative text embedded within a heterogeneous music classroom. As measured by the Qualitative Reading Inventory-4 (QRI-4) (Leslie & Caldwell, 2006), fluency, word recognition, comprehension, and in structional reading level of the treatment students were compared to their counterparts who sang as part of the regular music program. This investigation also provided a description of the peers’ interactions during the literacy tasks assigne d by the music teacher. This chapter presents the results of this concurrent mixed met hods study organized according to the research questions. The first two questions were conc erned with the quantitative phase of the study. The descriptive and infere ntial statistical results, as well as interpretations, are provided. The third question is concerned with the qualitative phase of the study. Peer interactions during the litera cy task of rereading through singing were examined and described. The statistical findings and the qual itative description were integrated within the qualitative findings in this study. Question One: Quantitative Findings for Treatment and Control Groups The findings in this section address th e following research question: to what extent, as measured by the QRI-4, is the reading performance of word recognition, fluency, comprehension, and instructiona l reading level of students using the Tune Into Reading program different from their regul ar music curriculum counterparts?

PAGE 159

146 The first quantitative research questi on addressed the readers’ use of the interactive sing-to-read program, Tune Into Reading, as an alternative text compared to their counterparts in the regular music class. Prior to the treatment, I administered a pretest using the QRI-4. Scores from the pret est ensured the students in the regular music class and the students in the class using Tune Into Reading were not different in their reading outcomes, specifically in Fluency (WPM) measured by words per minute, Word Recognition (WR) measured by oral r eading accuracy, Comprehension (COMP) measured by implicit and explicit questions after reading, and In structional Reading Level (RL) measured by combining scores from word recogniti on and comprehension questions before implementation. After the im plementation of the interactive singtoread program, Tune Into Reading I administered a posttest using the QRI-4 and compared the posttest scores with the pretes t scores to determine if students in the experimental group gained significantly over their counterparts in the control group. Initially, the students were assessed at pos ttest with a reading passage on the same instructional level attained dur ing the pretest. This was followed with statistical analysis of the posttest on the highest instructional reading level attained by the students. Results A doubly multivariate repeated measure ANOVA at an alpha level of .05 was conducted on Fluency (WPM), Word Rec ognition (WR), Comprehension (COMP), and Instructional Reading Level (RL) from pretes t to posttest by treatme nt group (Control vs. Treatment). Students were in itially assessed using the same instructional reading level scores attained during their pr etest. Means, standard deviat ions, and values for skewness

PAGE 160

147 and kurtosis (Table 9) for WPM, WR, CO MP, and RL from pretest to posttest by treatment group (n=32) and cont rol group (n=32) are presented. Table 9 Descriptive Statistics for Flue ncy (WPM), Word Recognition (W R), and Comprehension (COMP) Pretest Posttest Group M SD Skewness Kurtosis M SD Skewness Kurtosis WPM Control 136.56 36.09 -0.06 -.015 146.81 53.18 1.02 1.02 Treatment 125.28 32.95 0.07 -.071 160.34 47.52 -0.06 0.40 WR Control 0.98 0.02 -1.84 1.83 0.98 0.01 -1.38 0.42 Treatment 0.98 0.01 -0.38 -1.56 0.99 0.01 -1.15 1.44 COMP Control 0.76 0.03 1.78 6.07 0.75 0.03 2.50 11.79 Treatment 0.77 0.04 2.24 3.36 0.85 0.07 -0.27 -0.15 RL Control 5.58 1.22 0.35 -.047 5.58 1.22 0.35 -0.47 Treatment 5.45 1.17 -0.10 1.65 5.45 1.17 -.010 1.65 *Note Instructional reading level is the same at pretest and posttest An examination of Table 9 suggested hi gher reading achievement scores were attained for students classified as treatmen t than by students classified as control at posttest in WPM, WR, and Comp on the same instructional reading level attained at pretest. The treatment group e xhibited a means change in WPM from 125 at pretest to 160 at posttest, showing an increase of 35 in WPM scores; whereas the control group went from 137 at pretest to 147 at posttest a difference of 10 in the WPM scores. In addition to the WPM changes, the treatment group exhibited a means change in WR from .98 at pretest to .99 at posttest, illustrati ng an increase in word recognition scores. Whereas, the control group showed no increase in word recognition scores across the two points in time, .98 at pretest and posttest respectiv ely on the same instructional reading level attained at pretest. Furthermore, th e treatment group exhibited a means change in

PAGE 161

148 COMP from .77 at pretest to .85 at posttest, demonstrating an increase in comprehension scores of .08; while, the cont rol group’s s decreased acro ss the two points in time in COMP, .76 at pretest and .75 at posttest on th e same instructional level attained at pretest.. However, chance must be eliminated as a plausible explanation for the observed sample differences found in the population. A doubly multivariate repeated measures ANOVA was conducted with an alpha level of .05. Due to the kurtosis and skewness numbers found in the descriptive statistics, normality was assumed for two of the group distributions, WPM and WR for treatment and control at two points in time (pretest and posttest). For the third group distribution, CO MP normality appeared questionable for the control group at two points in time due to le ptokurtic kurtosis. Speci fically, a distribution with positive kurtosis (6.07 at pretest and 11.79 at posttest) exhibits a superior acute "peak" around the mean (a higher probability than a normally distri buted variable of values near the mean) and "f at tails" (a higher probability than a normally distributed variable of extreme values). Consequentl y, Stevens (1996) conte nded that “deviation from multivariate normality has only a small effect on Type 1 error” (p. 243). In addition, Tabachnick and Fidell (1989) re port that “for grouped data if there is at least 20 degrees of freedom for error in the ANOVA, the reported F test is said to be robust to violations of normality” (p. 71). Homogeneity of variances might be assume d, as the largest variance ratio was less than 2, which was not large enough to be considered problematic. Because the sample sizes were equal in each group, the analysis was expected to be relatively robust to

PAGE 162

149 violations. Based on the analysis assumpti ons, it appeared reasonable to conduct the doubly multivariate repeated measures ANOVA. There was a simultaneous difference on WPM, WR, and COMP at the same instructional reading level from pr etest to posttest by treatment group, F (4, 59) = 10.539, p <.001, 2 = .417. The Wilks’ Lambda for within subjects (time) was F( 3, 60) = 14.623, p <.001, 2 = .422. The Wilks’ Lambda for within s ubjects time by treatment interaction was F (3, 60) = 12.039, p <.001, 2 = .376. Table 10 presents an ANOVA of WPM, WR, COMP, and RL. Table 10 ANOVA Table on Fluency (WPM), Word Recognition (WR), Comprehension (COMP), And Instructional Reading Level (RL) DV F Sig. 2 WPM Group (gp) .014 .906 .000 (2870.637) Time (T) 18.957 .000 .234 gp T 5.684 .020 .084 (866.47) WR Group (gp) .278 .600 .004 (.000) Time (T) 4.641 .035 .070 gp T .364 .549 .006 (.000) COMP Group (gp) 43.447 .000 .412 .002 Time (T) 16.484 .000 .210 gp T 27.356 .000 .306 (.002) RL Group (gp) .175 .677 .003 (2.862) Time (T) ---gp T ----___________________________________________

PAGE 163

150 Fluency (WPM) There was a Group (treatment vs. contro l) and Time (pretest vs. posttest) interaction for WPM, F (1, 62) = 5.684, p = .020, 2 = .084. This indicated that the observed differences between the pretest a nd posttest for students in the treatment condition were different from the observed differences for students in the control condition in WPM. The main effect for Group was not statistically significant, F (1,62)=.014, p=.906, which suggested the observe d average scores between students in the treatment condition and in the control c ondition were not large enough to indicate a difference existed between the groups in WP M. However, the main effect for Time, F (1,62)=18.96, p=0.00, was found to be statistic ally significant, wh ich suggested the overall mean score at Time 1 differed from th e overall mean score at Time 2. To indicate relative positions of the sample means, an interaction graph is provided in Figure 6. Figure 6 Group (Treatment vs. Control) and Time (Prete st vs. Posttest) Interaction for Fluency (WPM) 120 125 130 135 140 145 150 155 160 165 PrePost Control Treatment The interaction graph of Group and Ti me for WPM illustrates a disordinal interaction. Relative to Fluenc y (WPM), the data indicated a mean 13.53 points lower for control students than for treatment students at posttest. The size of the interaction effect

PAGE 164

151 could also be expressed using eta squared ( 2) effect size (eta squared small .0-.3, medium .3-.5, and large above .7). The calculated value, 2= .084, indicated a fairly small effect size; however, it was of statistical si gnificance. To further examine the interaction for Fluency (WPM), two t -tests at an alpha level of .05 each were conducted. A dependent samples ttest was conducted for the control group, and no significant difference existed from the pretest ( M = 136.56, SD = 36.08) to posttest ( M = 146.81, SD = 53.18), t (31) = -1.255, p = .219, showing a small effect size of d=.2. A dependent samples t -test was conducted for the treatme nt group, and pretest scores ( M = 125.28, SD = 160.34) were significantly lowe r than the posttest scores ( M = 160.34, SD = 47.52), t (31) = -5.434, p <.001 with a large effect size of d=.8, indicating WPM treatment group’s scores significantly increased from pretest to posttest. In summary, it was found that pretes t and posttest scores for WPM were significantly different between control and treatment groups. The treatment group showed a significant increase from pretest to po sttest with a large effect size; whereas, within the control group there was no significan t increase from pretest to posttest with a small effect size. It could therefore be interpreted that the treatment group made a significant increase from pretest to posttest in their fluency (WPM ), as measured by words per minute on the same instructional read ing level attained at pretest and compared to the control group. Word Recognition (WR) Word recognition (WR) data revealed no statistically significant interaction for Group By Time. In the control group scores reported from pretest ( M =.9819, SD =. 02) to posttest ( M =.9847, SD=. 01), there was a small effect si ze of d=.2. Whereas, the treatment

PAGE 165

152 group scores demonstrated from pretest ( M =.9819 SD =.02 ) to posttest ( M =.9869 SD =.01) a medium effect size of d=.6. This suggested WR was more effective for the treatment group compared to the control group with a small effect from pretest to posttest at the same instructional level attained at pretest. Comprehension (COMP) There was a statistically si gnificant Group By Time inte raction for comprehension (COMP), F (1, 62) = 27.356, p < .001, 2 = .306. This indicated the observed differences between the pretest and posttest for student s in the treatment condition were different from the observed differences for studen ts in the control condition in reading comprehension (COMP). The main effect for Group was statistically significant, F (1,62)= 43.44, p=.000, which suggested the observed average difference between students in the treatment condition and in the control condition was large enough to indicate a difference existed between the gr oups in COMP. In addition, the main effect for Time, F (1,62)=16.48, p=0.00, was found to be st atistically significant, which suggested the overall mean score at Time 1 differed from the overall mean score at Time 2. To indicate relative positions of the sample means, an interaction is provided in Figure 7 to indicate relative positi ons of the sample means.

PAGE 166

153 Figure 7 Group (Treatment vs. Control) and Time (Pre test vs. Posttest) Interaction for COMP 0.74 0.78 0.82 0.86PrePost Control Treatment The interaction graph of Group and Time fo r COMP type illustrates a disordinal interaction. Relative to the r eading comprehension (COMP) sc ores, the data indicate a mean that was .10 points lower for control student s than for treatment students at posttest. The size of the interaction effect could also be expressed using eta squared ( 2) effective size. The calculated value, 2= .306, indicated a medium effe ct size that demonstrated statistical significance. To further examine the interaction for r eading comprehension (COMP), two t -tests at an alpha level of .05 each were conducted. When a dependent samples ttest was conducted for th e control group, there was no significant difference from pretest ( M = .76, SD = .03) to posttest ( M = .75, SD = .03), t (31) = 1.404, p = .170 with a small effect size of d=.3. However, the treatment group posttest scores ( M = .85, SD = .07) were significantly higher than their pretest scores ( M = .77, SD = .04), t (31) = -5.110, p < .001, showing a very la rge effect size of d=1.17. Therefore, indicating that for reading comprehension COMP, the treatment group’s scores significantly increased from pretest to posttest. In summary, it was found that the pretes t and posttest scores for COMP were significantly different between control and treatment groups. The treatment group

PAGE 167

154 illustrated a significant increase from pretest to posttest with a large effect size d=1.17; whereas, within the control group there wa s no significant increase from pretest to posttest with a small effect size d=.3. It could therefore be interpreted that the treatment group made a significant increase from pretest to posttest in their reading comprehension (COMP), as on the same instructional reading le vel attained at pretest when compared to the control group. Highest Instructional Reading Level Analysis was conducted on Fluency (W PM), Word Recognition (WR), and Comprehension (COMP) at the highest Reading Levels (RL) attained at posttest for the control and treatment groups. Four between -groups analyses of variance (ANOVAs) were conducted on highest reading level scor es for WPM, WR, COMP, and RL. Type I error was controlled by using the Bonferroni adjustment of the significant level to .02. The results revealed that between the Gr oups at the highest instructional reading level there were no statically significan t differences for WPM, WR, and COMP. However, it was found that for RL (instructi onal reading level) by Group (treatment vs. control), the treatment group showed a significa nt increase compared to the control group at the highest instructional reading level. The between-groups analyses of variance (ANOVA) indicated the Treatment instru ctional reading level scores RL ( M = 6.58, SD = .1.59), F (1, 62) = 31.28, p <.001, 2 = .335 were significantly higher than the Control RL ( M = 5.77, SD = 1.44) at the highest instructional reading level. This suggested that even though the treatment and control groups showed no significant difference in WPM, WR, or COMP, the treatment group increased si gnificantly in their instructional reading

PAGE 168

155 levels at the highest level a ttained at posttest. Table 11 disp lays the percentages by group of the instructional readi ng changes at posttest. Table 11 Changes in Instructional Reading Level s for Treatment and Control Groups Treatment Control (n=32) (n=32) Attained a Higher Level 81% 12% n=26 n=4 Stayed at the Same Level 19% 88% n=6 n=28 Summary of Finding for Question 1 In conclusion, the treatment group, us ing the interactive singing software Tune Into Reading demonstrated a significant increase w ith large effect sizes in Fluency (WPM) d=.8 and Reading Comprehension (C OMP) d=1.17 as compared to the control group who were singing in the regular music cl ass at the same instru ctional reading level attained during the pretest. In addition, although there were no observed differences noted in the interaction for Word Recognition (WR), the treatment group effect size was larger d=.6 than the control group e ffect size of d=.3. This sugge sted that from pretest to posttest the treatment group had a larger effect for WR than the control group. Furthermore, at the highest Instructional R eading Level (RL) the treatment group showed a significant increase in RL with a medium effect size d=.7 as compared to the control,

PAGE 169

156 whose effect size reported was very small. Ta ble 12 displays the inte ractions and effect sizes for the groups by variables. Table 12 Interactions and Effect Sizes for WPM, WR, COMP, and RL by Groups Variables Time X Group Treatment Control n=32 n=32 WPM d=.8 d=.2 WR NS d=.6 d=.2 COMP *** d=1.17 d=.3 RL *** d=.7 d=.1 Note small significant effect, *** large sign ificant effect, and NS no significant effect. These findings suggests the treatment stude nts of varying read ing abilities that used the interactive singing program, Tune Into Reading, illustrated a significant increase in their Fluency (WPM), Reading Comprehe nsion (COMP), and Instructional Reading level (RL) as compared to their counterparts who were singing in the regular music class. In addition, for the treatment students Word Recognition (WR) indicated a larger effect from pretest to posttest than the control gr oup. Specifically, this s uggests that rereading through singing, using the in teractive singing program, Tune Into Reading, was more effective regardless of the reading levels for treatment students compared to control students. These results can be interpreted as rereading through singing in the music classroom alone, as was the case for the c ontrol students, does not improve WPM, WR, COMP, and RL for the students of varying reading abilities.

PAGE 170

157 Furthermore, at the increased reading le vel reported at posttest, even though the treatment group had a significant increase in th eir instructional read ing level (RL), there was no significant difference between the groups in WPM, WR, or COMP. This suggested that, even though the treatment students increased in their instructional reading level (pretest M=5.45 and posttest M=6.58), their reading scores at the higher instructional reading level in WPM, WR, a nd COMP were lower than their scores at posttest on the same instructional reading. Sp ecifically, as the early adolescents in the treatment condition increased in text difficu lty, their fluency (WPM ), word recognition (WR), and comprehension (COMP) shifted from a fluent expert reader, on the similar level passage attained at pretest, to a surface fluent reader (e.g., Topping, 2006) at a higher level. Question Two: Quantitative Findings for Group by FCAT Reading Levels The findings in this section address the following research question: To what extent does the Tune Into Reading program impact reading sc ores of students who are “below, at, or above” grade level as de termined by the Florida Comprehensive Assessment Test (FCAT) reading scores? The second quantitative research question investigated whether an interaction effect of the repeated reading met hods through singing occurred on the reading performance of students stratified as “Below At or Above ” grade level in the treatment condition as compared to their counterparts in the control condition. The results from the QRI-4 pretest and posttest reading scores were used to determine reading outcomes (WPM, WR, COMP, and RL) fo r the treatment and control groups. Then, students achievement level scores (lev els 1-5) in reading were us ed to stratify the groups as

PAGE 171

158 “ Below, At or Above ” grade level in the treatment condition (students who used the interactive sing to read program) compared to their counterparts in the control condition (students who were singing as part of th e regular music program). The achievement levels as determined by the Florida Department of Education are: (a) Levels 1 or 2 that are considered Below proficiency in meeting grade level benchmarks, (b) Level 3 that is considered At grade level, and (c) Levels 4 or 5 that are considered Above grade level ( FCAT Briefing Book, 2005). Fluency (WPM) for FCAT Levels 1-5 A repeated measures ANOVA at an al pha level of .05 was conducted for FCAT Levels 1-5 in reading scores on WPM (wor ds per minute). Students were initially assessed on the same instructional reading level attained during their pretest. Means, standard deviations, skewness, and kurtosi s for WPM, from pretest to posttest by treatment group (Control n= 56 vs. Treatme nt n=56), are presented in Table13. Table 13 Descriptive Statistics for Fluency (WPM) for FCAT Levels 1-5 Pretest Posttest Group M SD Skewness Kurtosis M SD Skewness Kurtosis WPM Control 115.00 44.32 -0.51 -1.65 124.75 30.09 -0.85 0.21 1 & 2 Treatment 110.25 28.25 0.11 0.02 129.38 51.54 -0.02 -0.01 WPM Control 135.75 28.16 -0.11 -1.32 138.50 34.06 -0.23 -1.37 3 Treatment 123.08 34.66 0.33 -0.94 180.50 58.57 1.02 1.03 WPM Control 164.13 30.70 -1.58 2.44 201.88 65.82 -0.14 -1.54 4 & 5 Treatment 151.50 25.01 -0.35 1.33 186.50 34.72 1.69 3.33

PAGE 172

159 An examination of Table 13, stratified groups as Below (FCAT Levels 1 & 2), At (FCAT Level 3), and Above (FCAT Levels 4 and 5) for Fluency (WPM), suggested students classified as treatment from FC AT Levels 1-3 attained higher reading achievement scores than st udents classified as contro l at posttest on the same instructional reading level at tained at pretest. The groups stratified as FCAT Level Below (1 & 2) exhibited a mean change in WPM fr om 110 at pretest to 129 at posttest; whereas, the control group went up across the two poi nts in time, 115 at pretest to 125 at the posttest. The groups stratified as FCAT Level At (3) exhibited a means change in WPM from 123 at pretest to 181 at posttest; wher eas, the control group in creased across the two points in time, 136 at pretest and 139 at the posttest on the same instructional reading level attained at pretest. However, for groups stratified as FCAT Level Above (4 and 5), the control group appeared to have higher reading achievement scores than the treatment group. The control group FCAT Level Above (4 & 5) exhibited a mean change in WPM from 164 at pretest to 202 at posttest; whereas, the treatment group mean change increased across the two points in time, 152 at pretest and 187 at posttest on the same instructional reading level attained at pretest. However, to suggest that differences would be found in the population, chance must be ruled out as a plausible explana tion for the observed sample differences. A repeated measures ANOVA was conducted with an alpha level of 0.05. The kurtosis and skewness numbers found in the descriptive statistics suggested normality could be assumed for all three of the group distributi ons. In addition, homoge neity of variances might be assumed, as the largest variance ratio was less than 2, which was not large enough to be considered problematic. Furthermor e, since the sample sizes were equal for

PAGE 173

160 each group, the analysis was expected to be relatively robust to violations of the homogeneity of variance assumption. Based on the analysis assumptions, it appeared reasonable to conduct the repeated measur es ANOVA for the FCAT Level groups 1-5 on WPM. A repeated measures ANOVA was conducted in order to determine if a difference exists in the reading scores between groups across time, and for the different FCAT levels. Alpha level was set at .05. Table 14 illustrates the results of the repeated measures ANOVA. Table 14 ANOVA Table FCAT Levels 1-5 for WPM _______________________________ A review of the ANOVA table indicated for FCAT Levels 1-5 on WMP, there was a statistically signi ficant interaction for Group By Time By Level F (3. 55) = 3.28, p =.04, 2 = .197. This indicated that observed diff erences between pretest and posttest for DV F Sig. WPM Group (gp) 0.01 .930 (14255.512) Level (L) 7.83 .001 gp L 0.27 .762 (6443.604) Time (T) 18.02 .000 gp T 3.85 .062 (907.8478) L *T 1.56 .221 (1332.9236) gp*L*T 3.28 .047 (854.65278)

PAGE 174

161 students in the treatment condition were di fferent from the observed differences for students in the control condition within the three FCAT Levels on WPM. To indicate relative positions of the sample means, intera ction graphs for the three Levels by Groups across Time are provided in Figure 8. Figure 8 Interaction Graphs of FCAT Levels 1-5 on WPM FCAT Levels 1 and 2 FCAT Level 3 FCAT Levels 4 and 5 The interaction graph of Group By Time By Level for WMP reveals disordinal interactions for FCAT Levels 1-3. However, for FCAT Levels 4 and 5 the interaction is ordinal. Relative to fluency sc ores (WPM), the data indicate a mean for FCAT Levels 1 and 2 that were 4 points lower for control stude nts than treatment students at posttest. The size of the interaction effect exhibits a calculated value of 2= .017, indicating a small effect size. In addition, the fluency scores (WPM) for FCAT Level 3 indicated control students were 41 points lower than treatment students at posttest. The size of the interaction effect could also be expressed using eta squared ( 2) effective size. The calculated value, 2= .240, indicated a small effect si ze. However, relative to fluency scores (WPM), the data indicate a mean for FCAT Levels 4 and 5 that was 15 points 100 105 110 115 120 125 130 PrePost C T 100 120 140 160 180 PrePost C T 100 120 140 160 180 200 PrePost C T

PAGE 175

162 lower for treatment students than for contro l students at posttest. The size of the interaction effect could also be expressed using eta squared ( 2) effective size. The calculated value, 2= .001, indicated a small effect size. To further examine the interaction for fluency (WPM), three t -tests at an alpha level of .05 each were conducted. For FCAT Levels 1 and 2, there wa s no statistically significant difference between the groups. The c ontrol group scores repo rted from pretest ( M =115.00, SD = 44.32) to posttest ( M =124.75 SD= 30.09) illustrated a medium effect size of d=.7. Whereas, the treatment group scores demonstrated from pretest ( M =110.25 SD =28.25 ) to posttest ( M =129.38 SD =51.54) a large effect size of d= 1.1. This suggested the treatment group WPM had a larger effect in their scor es compared to the control group with a medium effect from pretest to posttest on the sa me instructional level attained at pretest. For FCAT Level 3 on WPM scores, the treatment group ( M = 180.50, SD = 58.57) was significantly greate r than the control group ( M = 138.50, SD = 34.06), t (22) = -2.148, p = .043. The control group scores reported from pretest ( M =135.75, SD = 28.26) to posttest ( M =138.50 SD= 34.06) illustrated a small effect size of d=.1. Whereas, the treatment group scor es exhibited from pretest ( M =123.08 SD =34.66 ) to posttest ( M =180.50 SD =58.57) a large effect size of d= 1.4. This suggested the treatment group in FCAT Level 3 outperformed the control group on WPM and had a larger effect in their scores for WPM compared to the cont rol group with a medium effect from pretest to posttest on the same instructional level attained at pretest. For FCAT Levels 4 and 5, there was no st atistical significant difference between the groups. The control group scor es reported from pretest ( M =164.13, SD = 30.70) to posttest ( M=201.88, SD= 65.82) with a small effect size of d=.3. Whereas, the treatment

PAGE 176

163 group scores revealed from pretest ( M =151.50 SD =25.01 ) to posttest ( M =186.50 SD =34.72) a small effect size of d=.4. This suggested that for FCAT Levels 4 and 5 WPM had little effect on the scores for both groups. These findings can therefore be interpreted as when the students are grouped by FCAT Levels in reading, the variable of Fluency (WPM), measuring reading rate, is mo re effective for students in FCAT Levels 1 and 2 ( Below ) and FCAT Levels 3 ( At ) than FCAT Levels 4 and 5 ( Above ). This suggests that when thinking about WPM for the higher performing students, reading rate may not be an important variable. Word Recognition (WR) for FCAT Levels 1-5 A repeated measures ANOVA at an al pha level of .05 was conducted for FCAT Levels in reading scores (FCAT levels 1 -5) on WR (word recogni tion). Students were initially assessed on the same instructional r eading level attained during their pretest. Means, standard deviations, skewness, and kur tosis for WR, from pretest to posttest by treatment group (Control n= 56 vs. Treatme nt n=56) are presented in Table15. Table 15 Descriptive Statistics on Word Recognition (WR) for FCAT Levels 1-5 Pretest Posttest Group M SD Skewness Kurtosis M SD Skewness Kurtosis WR Control 0.97 0.03 -2.54 6.73 0.98 0.01 -0.64 -2.24 1 & 2 Treatment 0.98 0.01 -0.64 -2.24 0.99 0.00 2.83 8.00 WR Control 0.99 0.01 0.72 -0.79 0.99 0.01 -2.54 6.77 3 Treatment 0.98 0.01 -0.00 -1.88 0.98 0.01 -1.15 -0.25 WR Control 0.99 0.01 -0.64 -2.24 0.99 0.01 -2.83 8.00 4 & 5 Treatment 0.99 0.01 0.00 -2.80 0.99 0.00 -2.83 8.00

PAGE 177

164 An examination of Table 15, stratified groups as Below (FCAT Levels 1 & 2), At (FCAT Level 3) and Above (FCAT Levels 4 and 5) for Word Recognition (WR), suggested that there was a hi gher reading achievement scores attained for the students classified as treatment from FCAT Levels 1 and 2 than for the students classified as control at posttest on the same instructional reading level at tained at pretest. The groups stratified as FCAT Level Below (1 & 2) in treatment exhibited a mean change in WR from .98 at pretest to .99 at posttest; wher eas, the control group in creased across the two points in time, .97 at pretest and .98 at the posttest. The groups stra tified as FCAT Level At (3) exhibited no means change in WR for either the treatment or the control group. The treatment group scores for WR were .98 from pretest to posttest, and the control group had a slightly higher scor e of .99 from pretest to posttes t. The groups stratified as FCAT Levels Above (4 and 5) WR scores showed no ch anges for either the treatment or control group at .99 on the same instructional reading level that was attained at pretest. This “ceiling effect” in word recognition suggests this test may be too easy for this group of students (Glass & Hopkins, 1996). However, to suggest that differences would be found in the population, chance must be ruled out as a plausible explana tion for the observed sample differences. A repeated measures ANOVA was conducted with an alpha level of 0.05. The kurtosis and skewness numbers found in the descriptive st atistics suggested normality could not be assumed for the three of the group distributi ons. Specifically for FCAT Levels 1 and 2, there was a leptokurtic kurtosis distribution of 6.53 for the control group at pretest and a leptokurtic kurtosis distribution of 8.00 for th e treatment group at posttest. In addtion,

PAGE 178

165 FCAT Level 3 control group at posttest for WR was negatively skewed -2.34 and had a leptokurtic kurtosis distribution of 6.77. Furt hermore, FCAT Levels 4 and 5, the control and treatment groups at posttest for WR were negatively skewed at-2.84 and had a leptokurtic kurtosis distribution of 8.00. However, Stevens (1996) contended that “deviation from multivariate normality has only a small effect on Type 1 error” (p. 243). In addition, Tabachnick and Fi dell (1989) report that for grouped data with an equal sample size, the reported F test was said to be robus t to violations of normality. Homogeneity of variances might be assu med, as the largest variance ratio was less than 2, which was not large enough to be considered problematic. Furthermore, since the sample sizes were equal for each group, th e analysis was expected to be relatively robust to violations of the homogeneity of variance assumption. Based on the analysis assumptions, it appeared reasonable to c onduct the repeated measures ANOVA for the FCAT Level groups 1-5 on WR. A repeated measures ANOVA was conducted in order to determine if there was a difference in the reading scores across gr oups, between time and for the different FCAT Levels. Alpha level was set at .05. Table 16 sh ows the results of the repeated measures ANOVA for WR.

PAGE 179

166 Table 16 ANOVA Table for FCAT Levels 1-5 on WR ___________________________________ A review of the ANOVA table indicated that the data revealed for FCAT Levels 1-5 on WR, there was no statistically signifi cant interaction for Group By Time By Level F (3, 53)=0.50, p =.619. In addition, there was also no Group By Level, Time By Group, or Level By Time interactions. However, the main effects for Time F (3, 53)=04.16, p =.053 and Level F (3, 53)=10.92, p =.003 were statistically si gnificant. This can be inferred as the means for the three Levels were different from Time 1 to the overall mean score at Time 2. The group effect size for FCAT Levels 1 and 2 was 2 =.067, a small effect size. However, within the group, the treatment st udents showed a large effect size of d=1.0 compared to the control group with a small effect size of d=.3. DV F Sig. WR Group (gp) 0.00 .992 (0.6253501) Level (L) 10.92 .003 gp L 0.00 .998 (0.l574536) Time (T) 4.16 .053 gp T 0.76 .391 (0.0000736) L *T 1.16 .323 (1113.786) gp*L*T 0.50 .619 (0.0000880)

PAGE 180

167 In addition, for FCAT Level 3, th e group’s effect size was small 2 =.091. However, within the group, the treatment st udents showed a medium effect size of d=.5 compared to the control group with a small e ffect size of d=.3. Furthermore, for FCAT 4 and 5, this level too had a small effect size for WR 2 =.036. However, within the group, the treatment students showed a medium eff ect size of d=.6 compared to the control group with a small effect size of d=.2. These fi ndings might be interpreted as the variable of WR was more effective for the treatment students than the c ontrol students on the same instructional reading attained at pretest. Comprehension (COMP) for FCAT Levels 1-5 A repeated measures ANOVA at an al pha level of .05 was conducted for FCAT Levels in reading scores (FCAT levels 1 -5) on COMP (reading comprehension). Students were initially assesse d on the same instructional re ading level attained during their pretest. Means, standard deviations, skewness, and kurto sis for WR, from pretest to posttest by treatment group (Control n= 56 vs. Treatment n=56) are presented in Table17. Table 17 Descriptive Statistics on Reading Comprehension (COMP) for FCAT Levels 1-5 Pretest Posttest Group M SD Skewness Kurtosis M SD Skewness Kurtosis COMP Control 0.75 0.00 0.00 0.00 0.75 0.00 0.00 0.00 1 & 2 Treatment 0.78 0.06 1.44 0.00 0.83 0.07 -0.64 -2.24 COMP Control 0.76 0.04 3.46 12.00 0.75 0.00 0.00 0.00 3 Treatment 0.77 0.04 2.82 8.06 0.87 0.10 -0.40 -0.58 COMP Control 0.78 0.04 -1.32 0.88 0.75 0.07 1.62 2.47 4 & 5 Treatment 0.77 0.05 2.83 8.00 0.86 0.05 -1.63 1.61

PAGE 181

168 An examination of Table 17, stratified groups as Below (FCAT Levels 1 & 2), At (FCAT Level 3), and Above (FCAT Levels 4 and 5) fo r Comprehension (COMP), suggested higher reading achievement scores attained for the students classified as treatment from FCAT Levels 1-5 than for the students classified as control at posttest on the same instructional reading level attained at pretest. The groups stratified as FCAT Level Below (1 & 2) treatment group exhibited a mean change in COMP from .78 at pretest to .83 at posttest; whereas, the c ontrol group COMP scores showed no change across the two points in time .75 at pretest and posttest. FCAT Levels Above (4 and 5) the treatment group exhibited means change in COMP from .77 at pretest to .87 at posttes t; whereas, the control group COMP scores decreased slightly across the two points in time .76 at pretest and .75 posttest. However, to suggest that differences would be found in the population, chance must be ruled out as a plausible explana tion for the observed sample differences. A repeated measures ANOVA was conducted with an alpha level of 0.05. The kurtosis and skewness numbers found in the descriptive st atistics suggested normality could not be assumed for the two of the group distributions FCAT Level 3 and FCAT Levels 4 and 5. FCAT Level 3 normality appeared to be que stionable for the control and treatment groups. The control and treatment groups di splayed a positive skew ness and leptokurtic kurtosis distributions for COMP at pretest, specifically the co ntrol (sk= 3.46, ku=12.00) and the treatment group (sk=2.82, ku=8.06). Fo r FCAT Levels 4 and 5, the treatment groups displayed a positive skewness of 2.83 and had a leptokurtic kurtosis distribution of 8.00 for COMP at pretest. However, St evens (1996) contended that “deviation from multivariate normality has only a small effect on Type 1 error” (p. 243).

PAGE 182

169 Homogeneity of variances might be assu med, as the largest variance ratio was less than 2, which was not large enough to be considered problematic. Furthermore, since the sample sizes were equal for each group, th e analysis was expected to be relatively robust to violations of the homogeneity of variance assumption. Based on the analysis assumptions, it appeared reasonable to c onduct the repeated measures ANOVA for the FCAT Level groups 1-5 on COMP. A repeated measures ANOVA was conducted in order to determine if a difference existed in reading scores across groups, betw een time and for the different FCAT levels. Alpha level was set at .05. Table 18 shows the results of the repeated measures ANOVA on COMP. Table 18 ANOVA Table for FCAT Levels 1-5 on COMP ________________________________ DV F Sig. WPM Group (gp) 0.18 .674 (0.41365448) Level (L) 11.06 .000 gp L 0.02 .897 (0.0017548) Time (T) 11.80 .002 gp T 20.16 .002 (0.00272309) L *T 0.66 .523 (0.00124236) gp*L*T 1.03 .366 (0.00189312)

PAGE 183

170 A review of the ANOVA table indicated th e data revealed for FCAT Levels 1-5 on COMP no statistical si gnificant interaction for Group By Time By Level, F (3, 53)=1.03, p =.366. In addition, there were also no Group By Level or Level By Time interactions. However, there was a statisti cally significant Time By Group interaction F (3, 53)=20.16, p =.002. This suggests the observed diffe rences between the pretest and posttest for students in the treatment c ondition were different from the observed differences for students in the control conditi on in COMP. To indica te relative positions of the sample means, an interaction gra ph combining the three FCAT Levels by Groups across Time is provided in Figure 9. Figure 9 Interaction Graph of the Groups By Time on COMP 0.74 0.78 0.82 0.86PrePost Control Treatment The interaction graph of Group By Time for COMP illustrates a disordinal interaction. Relative to reading comprehension (COMP) scor es, the data indicate a mean that was 10 points lower for the control stud ents than for the treatment students at posttest. The size of the interaction effect c ould also be expressed using eta squared ( 2) effective size. The calculated value, 2= .349, indicated a medium effect size. To further

PAGE 184

171 examine the interaction for reading comprehension (COMP), three t -tests at an alpha level of .05 each were conducted on the three FCAT Levels. For FCAT Levels 1 and 2, there was a statistically significant difference in comprehension (COMP), such that the Treatment group ( M = .83, SD = .07) effect size d= .3 was significantly greater than the Control group ( M = .75, SD = .00), t (7) = -3.416, p = .011 d=0, on the same instructional reading level attained at pretest. In addition, FCAT Level 3 on COMP scores, demonstrated the treatment group COMP posttest scores ( M = .87, SD = .10) effect size d=1.2 were signi ficantly higher than control COMP posttest scores ( M = .75, SD = .00), t (22) = 4.01, p < .001, effect size d=0. Furthermore, for FCAT Levels 4 and 5, the control gr oup scores reported no difference on COMP scores, such that pretest scores ( M = .77, SD = .04) were not signifi cantly different from posttest scores ( M = .75, SD = .07), t (7) = 1.000, p = .351, with effect size of d=.-4 on the same instructional reading le vel attained at pretest.. However, the treatment group showed there was a significant differen ce on the COMP scores between posttest ( M = .86, SD = .05), and pretest ( M = .77, SD = .05), t (7) = -2.714, p = .030, with a medium effect size of d=.6. This suggested treatment group had a significantly higher mean in their reading comprehension (COMP) scores at posttest compared to the control group, for students stratified by FCAT Levels 4 and 5. Th e findings reported suggested that for the treatment students in FCAT Levels 1-5, CO MP was more effective compared to the control groups in FCAT Levels 1-5. Highest Instructional Reading Level on WPM for FCAT Levels 1-5 Three between-groups analyses of vari ances (ANOVAs) were conducted. Type I error was controlled by using the Bonferroni ad justment of the signifi cant level to .02 at

PAGE 185

172 the highest instructional read ing level attained at postte st on WPM by group (treatment vs. control) for each of the FCAT levels ( Below level, At level, and Above level). Means, standard deviations, skewness, and kurtosis at the highest reading level attained at posttest for WPM are reported by group (tre atment vs. control) for each FCAT level ( Below level, At level, and Above level) and presented in Table 19. Table 19 Means, Standard Deviations, Skewness, and Kurt osis at the Highest Reading Level for WPM Scores by Group (Treatment vs. Control) and FCAT level FCAT Level Group M SD Skewness Kurtosis Below Treatment 119.88 26.08 0.91 -0.54 n=16 Control 127.88 50.63 -1.11 0.73 At Level Treatment 138.50 34.06 0.69 0.36 n=24 Control 181.25 61.85 -0.23 -1.37 Above Treatment 184.88 59.22 0.33 1.35 n=16 Control 127.63 26.64 -0.16 -0.10 The students stratified as FCAT level 1 and 2 Below level showed no difference at the increased reading level at posttest for WPM by group (treatment vs. control), F (1, 14) = .158, p = .697, 2 = .011. However, for the student s stratified as FCAT level 3 At level, the treatment group was significantly higher than the control group for WPM, F (1, 22) = 4.399, p = .048, 2 = .167, at the highest reading level. When students were stratified for FCAT levels 4 and 5 Above level in reading, the treatment group was significantly higher th an the control group, F (1, 14) = 6.217, p = .026, 2 = .308, at the highest instructional reading leve l for WPM attained at posttest.

PAGE 186

173 Highest Instructional Reading Level on WR for FCAT Levels 1-5 Three between-groups analyses of vari ances (ANOVAs) were conducted. Type I error was controlled by using the Bonferroni ad justment of the significant level to .02, at the highest instructional reading level attain ed at posttest, on WR by group (treatment vs. control) for each of the FCAT levels ( Below level, At level, and Above level). Means, standard deviations, skewness, and kurtosis at the highest reading level attained at posttest for WR by group (treatment vs. control) for each FCAT level ( Below level, At level, and Above level) are presented in Table 20. Table 20 Means, Standard Deviations, Skewness, and Kurt osis at the Highest Reading Level for WR scores by Group (Treatment vs. Control) and FCAT level For the students stratified as FCAT levels 1 and 2 Below level, there was no difference at the highest r eading level in WR by group F (1, 14) = .000, p = 1.000, 2 = 1.000. In addition, for the students stratified as FCAT level 3 At level, there was no mean difference at the highest reading level at posttest in WR by group, F (1, 22) = .672, p = .421, 2 = .030. Furthermore, when students were stratified for FCAT levels 4 and 5 Above level, there was no mean difference at the increased reading level attained at posttest in WR by group, F (1, 14) = 1.000, p = .334, 2 = .067. FCAT Level Group M SD Skewness Kurtosis Below Control 0.98 0.01 -0.64 -2.24 n=16 Treatment 0.98 0.01 -0.64 -2.24 At Level Control 0.98 0.01 -1.71 2.23 n=24 Treatment 0.98 0.01 -0.19 -2.25 Above Control 0.99 0.01 -2.83 8.00 n=16 Treatment 0.99 0.01 -2.83 8.00

PAGE 187

174 Highest Instructional Reading Level on COMP for FCAT Levels 1-5 Three between-groups analyses of vari ances (ANOVAs) were conducted Type I error was controlled by using the Bonferroni ad justment of the signifi cant level to .02 at the highest instructional read ing level attained at posttest on COMP by group (treatment vs. control) for each of the FCAT levels ( Below level, At level, and Above level). Means, standard deviations, skewness, and kurtosis are reported at the hi ghest reading level attained at posttest for COMP by group (t reatment vs. control) for each FCAT level ( Below level, At level, and Above level) and presented in Table 21. Table 21 Means, Standard Deviations, Skewness, and Kurt osis at the Highest Reading Level for COMP scores by Group (Treatment vs. Control) and FCAT level Below Control 0.75 0.00 0.00 0.00 n=16 Treatment 0.74 0.02 -2.83 8.00 At Level Control 0.75 0.01 -3.46 12.00 n=24 Treatment 0.76 0.06 0.89 0.68 Above Control 0.74 0.04 0.40 -0.23 n=16 Treatment 0.75 0.03 0.00 3.50 For the students stratified as FCAT level 1 and 2 Below level, there was no difference on the highest reading level at po sttest and no mean difference for COMP by group (treatment vs. control), F (1, 14) = 1.000, p = .334, 2 = .067, at the highest reading level attained at posttest. In additi on, for the students stratified as FCAT level 3 At level, there was no mean difference on the increased reading level at posttest COMP by group, F (1, 22) = .428, p = .520, 2 = .019. Furthermore, when students were

PAGE 188

175 stratified for FCAT levels 4 and 5 Above level, there was no mean difference on the highest reading level attained at posttest in for COMP by group, F (1, 14) = .636, p = .438, 2 = .043. Highest Instructional Reading Level on RL for FCAT Levels 1-5 Three between-groups analyses of vari ances (ANOVAs) were conducted. Type I error was controlled by using the Bonferroni ad justment of the signifi cant level to .02 at the highest instructional read ing level attained at posttest on RL by group (treatment vs. control) for each of the FCAT levels ( Below level, At level, and Above level). Means, standard deviations, skewness, and kurtosis at the highest reading level attained at posttest by group (treatment vs. control) for each FCAT level (below level, at level, and above level) are presented in Table 22. Table 22 Means, standard deviations, skewness, and kurt osis for RL at the Highest Reading Level for RL scores by Group (Treatment vs. Control) and FCAT level FCAT Level Group M SD Skewness Kurtosis Below Control 4.50 1.27 0.00 -2.80 n=16 Treatment 5.56 1.29 -0.84 2.14 At Level Control 5.25 0.45 1.33 -0.33 n=24 Treatment 7.46 1.61 0.15 -1.40 Above Control 7.63 1.33 0.37 -0.66 n=16 Treatment 6.63 1.27 2.11 4.17 For students stratified as FCAT level 1 and 2 Below level for RL, the treatment group was significantly higher than the control group, F (1, 22) = 45.000, p <.001, 2 = .763. The effect size for the treatment group was d=.7; whereas the control group was

PAGE 189

176 d=.3, indicating the variable RL for treatment group had a higher effect size compared to the control group. In addition, for the students stratified as FCAT level 3 At level at the higher instructional re ading level, the treatment group wa s significantly higher than the control group, F (1, 22) = 14.474, p = .001, 2 = .397. The effect size for the treatment group was d=.9; whereas the control group wa s d=.0, indicating the variable RL for treatment group had a higher effect size compared to the control group. However, students were stratified for FCAT levels 4 and 5 Above level at the higher instructional reading level attained at posttest, and no m ean difference existed on FCAT RL by group (treatment vs. control), F (1, 14) = 2.966, p = .107, 2 = .175. However, the effect size for the treatment group was d=.5; whereas the control group wa s d=.0, indicating the variable RL for treatment group had a higher effect size compared to the control group. The findings suggest that for the treatment st udents, RL was more effective compared to the control group. Summary of Findings for Question 2 In conclusion when the students were stratified by FCAT Levels 1 and 2 Below and FCAT Level 3 AT, at the same instructional reading level attained at pretest on WPM, there was a statistical ly significant difference be tween groups across time and within levels. Further analysis suggested re ading rate was more effective for treatment students than control students in FCAT Leve ls 1-3. However, for FCAT Levels 4 and 5 on WPM, there was no significant difference be tween the groups across time. In addition, although there were no observed differences noted in the interaction for Word Recognition (WR), the treatment group effect size for each level was larger than the control group effect size. This suggested, from pretest to posttest, the treatment group had

PAGE 190

177 a larger effect for WR than the contro l group. Furthermore, reading comprehension COMP for the treatment group, using the interactive singing software Tune Into Reading demonstrated a significant increase with larg e effect sizes. Finally, at the highest Instructional Reading Level (R L), the treatment groups showed a significan t increase in RL with a larger effect size as compared to the control groups. Table 23 displays the effect sizes for the groups by FCAT Levels on the four variables. Table 23 Effect Sizes for WPM, WR, COMP, and RL by FCAT Level Groups _______________________________________________________________________ Variables FCAT 1 and 2 FCAT 3 FCAT 4 and 5 Treatment Control Treatment Control Treatment Control WPM d=1.1 d=.7 d=1.4 d=.1 d=.4 d=.3 WR d=1.0 d= .4 d=.5 d=.2 d=.6 d=.2 COMP d=.3 d= 0 d= 1.2 d=0 d=.6 d=.-4 RL d=.7 d=.1 d=.9 d=0 d=.5 d=0 These findings suggests the treatment stude nts of varying read ing abilities that used the interactive singing program, Tune Into Reading, illustrated a significant increase in their Fluency (WPM), Reading Comprehe nsion (COMP), and Instructional Reading level (RL) as compared to their counterparts who were singing in the regular music class. In addition, on the variables by level, it appears FCAT Levels 1-3 had a larger effect than FCAT Level 4 and 5 on WPM, and FCAT Levels 1 and 2 showed a larger effect on WR than Levels 3-5. Further examination reveal ed that for reading comprehension COMP, it appeared to be more effective for FCAT Leve ls 3-5 than for Levels 1 and 2. This implies

PAGE 191

178 the different levels, when using the interactive singing software Tune Into Reading, appeared to be more effective for each group differently. Partic ularly, those students needing fluency (WPM) increased in read ing rate; whereas, those needing more opportunities for reading comprehension increased in their scores. This suggests the use of the interactive sing-to-read program provides for its user’s differentiated instruction. Question Three: Qualitative Fi ndings for Peer Interactions Due to the interpretive case study design of this phase of the study (Patton, 2002), this section is devoted to presenting an anal ysis of the data within individual cases, followed by a cross-case analysis. This was c onducted to determine the major themes for each case as well as those themes across the cases. The constructs and themes that emerged from the data were useful in answ ering the research ques tion that guided this phase of the study: How do middle school reader s interact with their peers within the context of their music classroom? Case studies included two groups of early adolescent peers who participated in this study. The two cases consisted of early adolescen ts in a music classroom who used the interactive singing program, Tune Into Read ing, and their counterparts who sang as part of their regular music class. The focus was on the descriptions of p eer interactions during the literacy task of rereading throug h singing. This section provides a focused understanding of how the peers interacted during the literacy task. The individual case studies of each group of peers are presented se parately, first in an effort to demonstrate how peers interact within the different instructional formats provided by the music teacher. After analysis of the data for each of the cases, a cross case analysis is presented to address the similarities and differences across cases. Integr ated within both individual

PAGE 192

179 cases and across the cases were the releva nt statistical findi ngs drawn from the quantitative phase of the study. Prior to the analysis of the individual cases, an overview of the study is provided. This included a description of the par ticipants, my role as a resear cher during this phase of the study, and the theoretical considerations that apply to this interpretive case study. In addition to the overview, an understanding is pr esented of who these literacy learners are through interactions during th e assessment period. This pr ovided a better understanding of the participants and how they see themselves as readers. Overview Participants During the 2006-2007 School Year, a total of 64 middle school students, in 7th and 8th grade, voluntarily agreed to particip ate in this study. These students were members of the fourth quarter Wheel Mu sic Class (March 12, 2007May 31, 2007). The Wheel Music Class was an assigned an el ective class of new cohorts (mix of 6th -8th grade students) each quarter of the sc hool year. However, within this sample of students in this study, there were no sixth gr ade student participants. During the fourth quarter, there were four intact classes of Wheel Music students randomly assigned by classes to a treatment or control condition for the study. When the classification characteristics were compared (as noted in Chapter 3, Table 3) for the treatment and control groups, it suggested the groups were predominantly White low SES students. Male eighth graders represente d a larger proportion for the treatment and control groups than their female counterparts or seventh grader peer s. In addition, only a small percent of the adolescents received suppor t services for learning or language needs.

PAGE 193

180 The Role of the Researcher I was trained to take observational fiel d notes when I collected data for the National Longitudinal Evaluation of Compre hensive School Reform (NLECSR) for the David C. Anchin Center at the University of South Flor ida (USF) and the American Institutes of Research (AIR) during the 20032004 school year. In addition, I have taken two Qualitative Research classes at the University of South Florida. For the current Interpretive Case Study, my role as a research er was participant observer. Initially, I planned to observe more than participate by sitting in the back of the music classroom while observing and taking fi eld notes. The reason for this decision was I did not want to have an adverse impact on th e peer interactions dur ing the literacy task of rereading through singing. Once the study bega n, I realized the impr acticality of this plan to sit in the back of the music classroom while taking field notes on peer interactions. In order to capture the peers’ interactions, I needed to move amongst the students and simultaneously take field notes. Since this research study commenced duri ng a new quarter of the school year, the students were accustomed to the presence of a researcher having never had this class or teacher prior to the study. Therefore, the st udents recognized me, and I was expected in the classroom. My presence did not deter them from their daily routine, as they greeted me by name whenever I came to the classroom. Theoretical Considerations Sociocultural Theory Many variables influence reading perfor mance of early adolescents’ literacy learning within the context of the middle sc hool classroom. Ivey (1999) contends early

PAGE 194

181 adolescent readers are complex and multidim ensional in their literacy learning. CookGumprez, (1986) and Scriber and Cole (1981) suggest sociocultural theories of literacy occur as literacy is used in specific contexts for specific purposes and is socially constructed and constituted. The act of liter acy is embedded in a network of social relations. Moje (1996) suggests th at in the secondary content classroom the social context that shapes literacy practi ces is uniquely complex. Teacher s and students in secondary classrooms move from class to class, teacher to teacher, and with a subgroup of peers. Teachers and students construct meaning about literacy and learning events based on values, beliefs, and knowledge, depending on the contextual situation. Additionally, teachers and students bring meaning to these interactions through their past beliefs, values, and knowledge during social interactio ns (Moje, 1996). St udies that are guided by broad theories as a social construction have focused on how so cial interactions influence literacy l earning (e.g., Myers, 1992). Moje also contends more research s hould investigate classr oom interactions and how they play a part in shaping literacy pr actices. Sociocultural theories informed this case study, especially using Ryan’s (2000) th eoretical work on peer interactions. Peer Interactions Ryan’s (2000) work investig ated the research on peer groups’ interactions, as a context for adolescent achievement, motiva tion, engagement, and socialization. In her analysis on the research of p eer group socialization for the early adolescent Ryan (2000) theorizes peers generally inter act three ways with one anothe r. During early adolescence, the peer group becomes a prominent contex t for development (Brown, 1990). The school and classroom provide opportuni ties for peers to interact th roughout the day. Ryan (2000)

PAGE 195

182 reports “peer interactions consume significan tly more time in adolescence compared to any other time in childhood” (p. 107). These interactions with p eers can concern both academic (e.g., achievement) and nonacademic matters (e.g., engagement, motivation, self-efficacy, and interest). Ryan (2000) sugge sted three ways early adolescents generally experience peer interactions within the context of middl e school: through information exchange, modeling, and peer pressure. Information exchange occurs when adoles cents have a discussi on with their peers (Berndt, 1999). In an experimental study with eighth-grade students, Berndt, Laychak, and Park (1990) found that when adolescents had to make an academic decision, such as attending a rock concert or study for a test, they initially responded differently from one another. However, after discussing this di lemma with their peers, their answers were similar to their peers. This form of inte raction could influence the early adolescent’s choice to partake in the literacy task presented by the teac her if it was used effectively. Modeling is another form of adolescent peer interaction. This interaction refers to individual changes in cognition, beliefs, or affect, which are a result of adolescents observing their peers (Ryan, 2000) Observing a specific behavi or a peer performs or listening to a peer voice, a certain belief can induce an adolescent to change their stance or adopt their peers’ behavi ors or beliefs. Schunk and Zi mmerman (1996) reported peer modeling influenced self-efficacy beliefs. In their study, they found early adolescents who verbalized difficulty with a task and then observed thei r peers have success with the same task, then believed they could complete the task. The early adolescent, when faced with a literacy task, may have success by observi ng their peers. Peer pressure is the third way the early adolescent interacts with their peers.

PAGE 196

183 Peer pressure takes on the role of so cial reinforcement (Ryan, 2000). Brown, Lohr, and Eicher (1986) found beliefs and be haviors that are disc ouraged by the groups are not likely to be displayed; whereas, beli efs and behaviors that are positively received by the group are more likely to surface. Theref ore, participation in the literacy tasks that the peer group positively received through this interaction could have a positive effect on the group’s beliefs and decisions to participate by the group members. Peer pressure may also play a role in how the peer group influences motivation. Brown, Lohr, and McClenahan (1986) re ported peer pressure regarding school involvement is significantly correlated w ith self-reported behaviors and attitudes regarding school. Ryan (2000) recommended furthe r research on peer in teractions within a domain specific classroom may fill in the gaps in the literature. The above named recommendations from the research of Moje (1996) and Ryan (2000) are used to frame this study’s qualitative component. Ryan’s theory on three genera l categories of peer interactions framed the interpretive case study, along with Moje’s recommendations that research on interactions within the setting of the content cl assroom should be studied to inform practice as to how literacy learning could be shaped. Assessments Prior to and after the experimental treatment, all 64 students were individually assessed in their reading performance in Fluency (WPM), Word Recognition (WR), Reading Comprehension (COMP), and Instru ctional Reading Level (RL). The students were told this assessment would not be a part of their personal records, and any information obtained was confidential so no one in school would ever see any results

PAGE 197

184 from the assessments with their names attached to the scores. However, it was reinforced to the students that it was important to try their very best when reading. I continued with an explanation pertai ning to the assessment process and asked their permission to proceed with the assessmen t. It was explained that they would have two passages (sometimes more) to orally read and while they were reading I would be taking notes. On completion of their oral r eading of the passages, they would be asked comprehension questions about what they re ad. Interestingly, what I observed and what the students provided through unprompted self-reports confirmed Ivey’s (1999) contentions about the complexity of this population of heterogeneous middle school students of varying reading abilities FCAT Reading Level Scores The scores of Florida FCAT levels 15 in reading are indicators, according to the state, of reading ability and performance. Le vels 1 or 2 are considered below grade level; whereas, Levels 3 through 5 are considered at or above grade level in reading. There were 56 students out of 64 (28 in treatm ent and 28 in control groups) who had FCAT level reading scores in this study. Eight of the students did not have 2006 FCAT level reading scores for various reasons (e .g., relocated from another state). Of the 56 students 40 (71%) were consid ered meeting grade level proficiency or above grade level, according to their FCAT leve l reading score. Students at this level are considered proficient readers and are not n ecessarily provided any individual support in their reading skills or strategies. Specifical ly, in this study 40 st udents out of 56 were determined to be proficient (FCAT level read ing scores of 3 through 5) in their reading, according to results from th e FCAT high-stakes tests.

PAGE 198

185 However, when given the QRI-4 reading assessments, their instructional reading levels illustrated only 19% of the students (11 out of 56) at pretest we re on grade level or above in reading. A combined mean score of instructional grade level in reading was 7.5 for all participants at FCAT levels 3 thr ough 5. At posttest only 27% of the students (15 out of 56) were on grade level or above in re ading with a combined mean score of grade level reading at 8.1 for all pa rticipants in both treatment and control groups. This suggested only 15 students out of the 40 st udents, determined by their FCAT level reading scores, were in fact meeting grade leve l proficiency in their reading at posttest. Student participants who scored at FCAT level reading score 1 or 2 are considered below grade level in their readi ng. In this study, 16 stude nts (29%) out of 56 scored at a level 1 or 2 Specifically, these students were determined to be below grade level proficiency in their reading, according to the results from thei r FCAT reading level scores. However, when given the QRI-4 read ing assessments, co-scored for reliability (see Chapter 3), their instructional reading leve ls showed that 80% (45 out of 56) of the students (combining all FCAT leve ls) at pretest were below gr ade level in reading with a combined mean score of grade level reading at 4.94. At posttest 73% ( 41 out of 56) of the students were below grade level with a combin ed mean score of gr ade level reading at 5.49. This suggested that for the 41 students who were reading belo w grade level only 16 of the students are receiving reme diation in their reading. It is therefore suggested that for the 56 students in this middle school 27% ar e proficient readers; whereas, 73% are reading below grade level. These descriptive findings concur with the statistical results for the students stratified by their FCAT levels 3 (At level) and 4 and 5 (A bove level).The statistical

PAGE 199

186 results showed that the groups had different pretest scores on their instructional reading levels (RL) than the control group. Specifi cally, for the treatmen t group on Level 3 ( At ) had significantly higher pretes t scores p= .019 than the control; whereas for control group Level 4 and 5 (Above) had significantly higher pretest scores then the treatment student p= .002. However, the statistical findings repo rted for both groups (treatment vs. control) on their pretest instructional reading leve l scores found no significant difference on the pretest scores p= .677. In addition, the descriptive findings sugge st according to the students’ FCAT levels 3, 4, and 5 (at or above level), 71% (40 out of 56) of students were determined as meeting or above grade level in their r eading based on FCAT level reading scores. However, when assessed using the QRI-4, onl y 27% were performing at or above grade level in their reading. This suggests that usi ng FCAT reading level scores as benchmarks to determine instructional reading level do not appear to correlate to scores from the QRI-4 assessment. Amrein and Berliner (2002), overall, conte nd that “there is no compelling evidence from a set of states wi th high-stakes testing polices that those policies result in transf er to the broader domains of knowledge and skill for which highstakes test scores must be i ndicators” (p.54). Therefore, the us e of a high-stake test scores alone can not account for the many variables associated with unders tanding the reading process and relating that to th e characteristics of this group of early adoles cent literacy learners and their fluent reading behaviors. Coincidentally, these results align with the report from Th e National Assessment of Educational Progress (NAEP) (2005) that contends 73 % of eighth grade students perform below or at a basic level in their reading achievement. In addition, consistent

PAGE 200

187 with NAEP results, Biancarosa and Snow (2006) reported, to the Carnegie Corporation, over 70% of adolescents struggle with thei r reading in some manner and, therefore, require differentiated and strategic instruction. Fluency: Absence of Prosodic Reading Fluency is a necessary aspect of successful reading as it allows readers to read with speed, accuracy, and proper expression (N ational Reading Panel, 2000; Rasinski, 2004). Rasinski (2004) conte nds reading fluency is a “bridge between two major components of readingwording, decoding, and comprehension. At one end of the bridge, fluency connects to accuracy and au tomaticity in decoding. At the other end, fluency connects to comprehension through pros ody, or expressive inte rpretation” (p. 1). The students in both groups read their r eading passages orally with speed and a high level of accuracy in word recognition, and yet they struggled with comprehension during the pretest assessment. Their oral readi ng was absent of volume, tone, pitch or any expression. There was no pausi ng at punctuation, rereading for clarification, or selfcorrections made in 53 out of 64 students. In terestingly, 75% of the students asked prior to reading the passage, “ how fast do you want me to read,” or “I need to read this fast,” (Assessment Notes April 2, 2007). Indicating for th is group of literacy learners, proficient fluent reading was related to speed. My response to all th e students was I want you to read at a pace so you can understand what you are reading and be ab le to answer the questions when you finish. Regardless of this suggestion at pretest, as noted in the statistical findings, there was no significant difference between the groups in fluency (WPM), word recognition (WR), comprehension (COMP), or in structional reading levels (RL).

PAGE 201

188 However, at posttest, the treatment group of students outperformed their counterparts significantly from pretest to pos ttest in reading comprehension COMP and instructional reading levels RL. The oral r eading of the students in the treatment group, although fast, had expression, pitch, and volume, unlike their counterparts. Specifically, 81% of the treatment students, or 26 out of 32, read their passage making self-correction, pausing at punctuation, and rer eading phrases or sentences. Whereas, the control group of students, only 28% or (9 out of 32) of th ese students incorporated these same prosodic elements in their reading. This was particul arly noted in the sta tistical finding showing a significant difference between the groups in COMP reading comprehension, p< .001, and instructional reading levels, p<.001. Rasinski (2004) contends when reading, prosody is incorporated in the rereading with accu racy and automaticity then the student’s comprehension will improve. Students’ SelfReports on Their Reading Disposition Self-reports by the students during th e assessment sessions provided an opportunity for me to hear the students’ pe rceptions of how they see themselves as readers and their personal relationship with the reading process. There were two selfreports that had an overwhelming frequency of responses from the students regarding their relationship with and disposition for th e reading process. The first self-report was a dislike for reading. An example of this was, “I hate to readther e are no books in this school I like” (Assessment Notes, April 1, 2007 ). Specifically, 39% of the students (25 out of 64) made a statement similar to this response before the reading assessment even began. Interestingly, this concurs with Ivey ’s (1999) findings in her case study on the

PAGE 202

189 early adolescent’s attitude towards read ing and book choice in the middle school classrooms. The second self-report with the highest frequencies related to how cognizant the students were of their difficulties in reading comprehension. An example of this was, “I can never remember what I read” (Assessmen t Notes, April 1, 2007). Specifically, 30% of the students or 19 out of 64 students made reference to not being able to remember what they read. This was evident in the scores on reading co mprehension for both treatment and control groups at pretest and pos ttest at the higher instructional reading level. The means for both groups was the sa me in reading comprehension (COMP) at pretest, 77%. The reading comprehension (COM P) significantly increased at posttest to 85% on the same instructional reading for the treatment group. However, at posttest on the highest instructional read ing level, COMP was again th e same for both treatment and control groups at 78%, even though the trea tment students had increased their reading level (RL) over a year comp ared to the control group. Biancarosa and Snow (2006), in their report to the Carnegie Corporation, contend that reading comprehension is an area of c oncern for the early adol escent. Some students may have trouble with decoding words accurately and with automaticity; whereas, others may read words fluently, but they do not re member what they read. In addition, other students may know comprehensi on strategies but do not ha ve sufficient practice or opportunities for use. Biancaros a and Snow suggest this may be a result of limited understanding, support, and practice for strategi es used to develop comprehension in the various content areas. Topping (2006) suggest s that “even fluent readers will show dysfluency with text beyond their in dependent reading levels” (p.106).

PAGE 203

190 Treatment and Control Group Data Analysis The analysis of the data required qualitative analysis procedures. Patton’s (2002) guidelines for content analysis recommend reading through the data at a specific time and making notes in the margins pertaining to sp ecific notions about meanings. Moerman’s (1988) suggestions for conversat ion analysis guided the analys is of peer interactions through conversations. In addition, Miles and Huberman’s pattern analysis (1994) was used to code data and look for emerging patterns. Observational field notes were taken dur ing each 50-minute cl ass session, twice a week for each of the four classes assigned to the treatment or control condition. Field notes were taken on a pad of paper during the Wheel Music Class periods noting time, place, attendance, and all the peer interac tions during the observation. These observations focused on describing the relationship, if any, be tween the literacy ta sk and music teacher assigned (rereading through singing) and focu sed on interactions (peer talk, peer modeling, and peer social reinforcement) among students who were singing using the interactive program, Tune Into Reading, versus the peer intera ctions among students who sang in the traditional music class. Strauss (1993) recommended that to a ssist with this difficult process for beginners, researchers should develop a codi ng paradigm. The paradigm, which applies to this study, consisted of: (1) the literacy task (reread ing through singing) assigned by the music teacher and (2) interactions among the peer groups during the literacy task assigned by the music teacher for the two cases (students using the interactive sing-toread program and students in the regular music class).

PAGE 204

191 Following a theory suggested by Ryan (2000) there are generally three ways that early adolescents experience peer interacti ons within the contex t of middle school: (a) through information exchange (discussi on), (b) modeling (p eer observation and imitations), and (c) peer pressure (social rein forcement). I used thes e three categories as preliminary coding categories and as a framework to focus my observations. Information exchange refers to discussi ons and talk amongst the peers, capturing direct quotes from the various conversations that the peers excha nged during the literacy task: Peer 1 “How did you get the song to slow down,” and Peer 2, “Click on this button” (Observational notes April, 7, 2007). Peer modeling, on the other hand, refers to the act of peers observing one another a nd results in changes in behaviors or understanding within the student (s). This is achieved by describing the interactions during the literacy task that documents th ese changes: [He looke d around the classroom for two minutes then he smiled and went b ack to playing the drums] (Observational notes, April 7, 2007). Finally, peer pressure occurs through social reinforcement, both negative and positive. Descriptions of peers’ accepting or rejecting behaviors exhibited by their counterparts through body language, facial expressions, smiling, or laughing during the literacy task: [T hit the drum wrong… M laughed…and then the class laughed…T turned red and put his head dow n] (Observational notes, April 7, 2007). Ryan (2000) suggested three categories beca me preliminary coding categories. They were then put into a matrix used for data analysis. Field notes were reviewed daily afte r all the observations were completed. Initially, I would read through the notes three times to get a holistic sense of the data collected. Then the notes were bracketed and coded as one of the three peer interaction

PAGE 205

192 categories. Units of data, conversations among peers or paragraphs that described peers observing or applying pressure to other peers, were brackete d and labeled as one of the three peer interaction categories. This was followed by transferring the bracketed notes to a matrix (Appendix B) with the three categ ories. The matrix was used to ensure observations did not stray from the focus of the study. Once the data were transferred, the difficult job of data analysis began. Figure 5 pr ovides an example of the matrix used in this study My first task involved typing the field not es from the observations of the Wheel Music Classes. The notes were typed-up dail y after all classroom observations so the information could remain fresh. Once this task was accomplished, I began the difficult task of reading and analyzing the data. First, I read the field notes from the classes three times to gain a holistic sense of the data. Then, I returned to th e data, bracketed the categories of peer interactions, and labeled them as information exchange, modeling, and peer pressure, so it could be transferred to the peer interaction matrix (Appendix B). I then read each line of the da ta in the matrix and highlight ed units of meaning, patterns where repeated phrases and/or words occurr ed (Patton, 2002). Construct names emerged from these data. The construct names came directly from the data. One example that illustrates how this was done was a phrase that described peer modeling, “In the four corners of the computer lab, small groups of females look at one another and start to laugh softly, as they secretly glanced around the room.” This sentence was highlighted and was bracketed with the cons truct name, Peer Observation. Once in the matrix, the data were further analyzed to determine the elements of peer interactions during the li teracy task. After the elements were identified and assigned

PAGE 206

193 construct names, they were added to th e Construct Key (Appendix C). I used the Construct Key to be consistent with constr uct names from the emerging data but also added any new emerging constructs from th e consecutive observations to the construct key. The elements were then grouped according to the construct names. The elements with the construct names assigned were then cut-up and placed in a folder. The frequency of each construct was tallied to determine whether or not an element was emphasized during the peer inter actions. The frequency calculations were followed by organizing the constructs into categories. E ach category of constructs was placed on a bulletin board and further analysis determined the themes that emerged from these data. These themes were presented first as individual cases, and then a cross case analysis was conducted. I repeated this process for 28 observations (14 observati ons for the treatment case and 14 observations for the control case), and then I analyzed these data again with the finalized Construct Key. To ensure that the qu alitative phase of this study is credible, qualitative researchers with b ackground in literacy were uti lized as a second observer and conducted an analysis check of the data. After the constructs were identifie d, they were grouped accordingly under a construct heading. There were two construct headings that emerged from the data for both the treatment and control group. They were Group Characteristics of Peer Interactions and Peer Interactions During the Instructional Procedures. The frequency for each construct was then calculated to de termine the themes for the cases. These themes are presented first as individual cases then as a cross case analysis.

PAGE 207

194 Integration of the Data Priority was given to the quantitative appro ach. It looked at the st atistical relationship between students who used the sing-to-read pr ogram, Tune Into Reading, compared to students who sang as part of th eir regular music program. Therefore, the analysis for this approach was done first to answer the fi rst two questions of this study. However, concurrently, qualitative case study methods were used to better understand and describe the peer interactions occurring during the l iteracy task assigned by their teacher. The integration of the two types of data occurred within the qualitative findings section of this study. The quantitative results and qualitativ e description were tr iangulated mixing the quantitative results with the qua litative descriptions in orde r to provide a clearer picture and more fully answer the research questions Treatment Group Description of Classroom Computer Lab A single door opens into a small rectangular sound-proof computer lab, located at the back left hand corner of the music clas sroom. Three quarters of the parameter of the room housed 15 permanent computer docking st ations. An empty table, located at the front of the lab and a small table in the middle of the lab, was also used during the intervention for the remaining students. Laptops were placed there from a laptop bunker that stayed during the 7-week intervention. At each comput er station, an individual microphoned sound-proof headset for each of the students was attached to the computers. Students would walk into the lab, sign-in, and retrieve their personal folders to keep a record of songs that they sang and recorded during each session.

PAGE 208

195 Data Analysis The data collected from the treatment group came from 14 classroom observations. The classroom observations took place in the music classroom during the fourth quarter of the 200 6-2007 school year (April 2nd May 15th, 2007), over the 7-week experimental treatment period. As previously noted, four classes were randomly assigned by class to the treatment or control conditi ons. Two classes were combined and became the treatment group, and two classes were combined and became the control group. Observations occurred twice a week for th e treatment group from the week of April 2, 2007May 15, 2007. There were 20 constructs that emerge d from the treatment group observational data. The constructs were tabulated to dete rmine the frequency of each construct in the data. Constructs with a frequency count of fi ve or less were not included in the analysis. Table 24 specifies the 9 constr ucts that emerged from these data with the highest frequency. Refer to Appendix C for a description of the constructs. Table 24 Constructs from the Treatment Group Observational Data Construct Frequency Extrinsic Motivation 34 Peer Observations 29 Peer Hierarchy 18 Peer Support 13 Autonomy 12 Intrinsic Motivation 10 Students’ Perspectives of A lternative Text 9 Safe Risk –Free Environment 8 Disequilibrium 7

PAGE 209

196 There were three themes that emerge d from the treatment group’s classroom observational data. They were developed af ter a thorough analysis of the data, which included reading through the data at least th ree times for a holistic sense of the data, analyzing the data for mean ingful units, developing constructs from the emerging meaningful units, and tallying the constructs for frequency. The themes that emerged were Group Dynamics Motivatio n, and Singability vs. Readability These themes encompassed the essence of peer interacti ons during the treatment groups’ use of the interactive sing to read program, Tune Into Reading Table 25 presents these themes and the frequency with which they occurred in the data collected from the treatment group. Table 25 Themes from the Treatment Groups’ Observational Data Theme Frequency Group Dynamics 68 Motivation 56 Singability vs. Readability 18 Group Dynamics Safe Risk-Free Environment The music teacher was aware that this nonconventional alternative middl e school task of rereadi ng through singing made the students nervous. To help alleviate some of their apprehensions and fears, she would constantly walk around during the sessions and give both verbal (telling them it would be alright) and non verbal (pat on th e back or a warm smile) suppor t. The students did in fact approach the task with some nervous laughs, self-reports of their l ack of singing ability, and intense observation of one another. Th e peers constantly looked around at one

PAGE 210

197 another to be assured this was an acceptabl e activity in which they should partake. The following is an example from the observational notes taken: He looked around the room, focusing on each person for over 20 seconds. Then, his gaze stopped at a small group of males. He watched intently as the males were singing and softly laughing with each other. He shrugged his shoulders and turned back to hi s computer. (Week 1, April 3, 2077). The potency of a belief that one can accomplish the task appeared nested in an environment that was safe and risk-free. Rega rdless of the music teach er’s attempts to be supportive and understanding of the students’ apprehension, there was a need to feel this task was socially acceptable by the group. It wa s only after observing their peers that the early adolescents would feel safe enough to take risks a nd partake in the activity. Peer Observation When the students entered the computer lab each session, they picked up their folders and took a seat at a computer station. Although the computer program generated scores for the students ’ pitch accuracy, the music teacher had the students write their individual scores in their folders. This was because she wanted to use their folders to dialogue with the i ndividual students about their progress. There was no assigned seating during the treatment sessions. The early adolescents came into the lab and sat next to their friends. Intere stingly, not only would the peers sit with their friends, but they also separated themselves by gender. The females in the group sat in each of the four corners of the computer lab; whereas, the seventh grade students sat wherever there was an open seat, although they usually found a seat that corresponded with their ow n gender. In addition, the se venth grade students only sat down and took their seats after the eighth grader s were seated. The center of the lab was

PAGE 211

198 taken over by the eighth grade males. There was one small group of eighth grade males in particular that placed themselves at the cen ter table in the middle of the computer lab. Once the students settled down and were seated, the music teacher would give the students directions for the session. She reminde d them of the procedures for using the program and to record their pitch scores in their folders. After answering any questions, she told them to begin. Although all the studen ts would put their headsets on and select their songs, very few actually started to use the program. In fact, during the first eight minutes each and every student in the lab se cretly glanced around the room and over their shoulders looking at the ot her peers. The following is an example taken from observational notes of the students who were su pposed to have started using the program: In the four corners of the computer lab, small groups of females look at one another and start to laugh softly, as they secretly glanced around the room. A seventh grade male turned his head to th e left looking over his shoulder and then to the right. Two males look at each ot her and then behind where they were sitting, to the middle of the lab. A fema le bends over and turns her body sideways in the chair. Then she scratches her leg and at the same time scans the room. Two males are slouched back in their seats w ith a blank vacant look on their faces and appear to stare into space; however, th eir eyes glance sideways without moving their heads to observe the peers around them. (Week 1, April 5, 2007) The students appeared to comply with the music teacher’s direction to start using the program. They would face their computer s, put their headsets on, and open up the program to a song on the computer. However, it was only after they observed one another

PAGE 212

199 that they actually started to use the program. Instead, they looked at one another to see if it was acceptable for them to begin. The focus of the entire group was on a small group of eighth grader males that sat in the middle of th e computer lab, at the center table. This group of three male students appeared not to be cognizant of the rest of the group’s observations. Their conversations and attention remained within the group with the other males they sat with at the table. Peer Hierarchy. This small group of eighth grade males sat directly in the middle table of the computer lab. This group was unlik e the other peers in the lab. They were not quiet, and they did not look ar ound to see what the other peer s were doing. Instead they would talk and laugh with one another. The ot her students watched a nd listened to this small group of eighth grade males intently, and when they spoke and made comments about the task, the rest of the group woul d stop, listen, and follow their lead. The following is an example of this group of peers interacting and how the rest of the students responded: Male 1: This is pretty cool. Male 2: I stink at singing. [Other peers around the room shak e their heads in agreement] Male 1: It’s better than doing work. Male 3: Yeah... not like real work. [Three females look at one anot her and shrug their shoulders] Male 2: Okay, let’s do it! [Male 1: holds up his hand and makes a rock-roll sign] (Week 1, April 5, 2007)

PAGE 213

200 This group of males appeared to be th e dominant characters within the group. There were no interactions w ith the other peer members in the lab; however, once this group of males settled into the task of rereading th rough singing, a domino effect occurred within the lab. When the dominant ma les started to use the interactive sing-toread program, all the students in the lab si multaneously turned to their computers and started working, as if on a silent cue. This occurred at the beginning of the intervention and continued across the entire seven weeks. Peer Support Once the peers settled in and felt comfortable with the interactive program they would come into the computer lab and get right to work, following the lead of the dominant males. This silent cueing sy stem remained intact throughout the seven week intervention as the peers would liste n and model their be haviors and actions according to what they saw and heard from the group of eighth grade males. The interactions among the group of males appeared to be supportive and collaborate. They modeled for the other students this support sy stem within the group as described in the following example: Male 1: How do you slow this music down? Male 3: Click on the tempo key and that will do it. Male 2: Yeah... like this. [Male 2 shows Male 1 how to do it] Male1: Thanks. (Week 2, April 11, 2007) Although the eighth grade males primarily conversed amongst themselves, they modeled a support system for the rest of st udents in the lab. Once they heard the males being supportive and collaborating with each another, several small clusters within the

PAGE 214

201 lab did the same. The climate of peer support was important in order to keep the students continually using the sing-to-read program However, regardless of this supportive behavior, in order to maintain sustainability, the program its elf had to be interesting and motivating for the entire group. Motivation Extrinsic Motivation Often throughout th e intervention the p eers would be heard talking about their pitch accuracy scores. Th e scores for pitch accuracy ranged from 0-100, and the students would often compare thei r scores with one another. These scores measured how accurately the students could sing and record themselves within a given pitch. The students were able to see their vo ice through real time pitch tracking frequency lines as they recorded themselves singing. The objective was to keep these lines within the pitch box above the words in the song. Wh en they completed recording themselves singing, a score would pop-up on the scree n. This would show the students how accurately they met the pitch and rhythm of the song. This game-like quality was interesting and motivating for the students. Fu rthermore, the music teacher had not given the students one particular score that th ey should get (e.g., 80%) with pitch accuracy. Instead, she wanted them to work to their i ndividual highest potential by trying their best. Regardless, the peers would often be heard ch allenging each other as they got their scores as in the following example: Peer1: What did you get? Peer 2: I got a 60. Peer 3: You did better then me…I got a 53. Peer 1: That’s good… that’s a hard song. (Week 2, April 9, 2007)

PAGE 215

202 The peers were excited and motivated by the scores they received. However, it was interesting to note that even though th e conversation by the peers was competitive, it was at the same time also supportive. This may be as a result of the music teacher’s reinforcement of trying their best instead of getting only one score that was acceptable. In addition, the climate of the class that was set by the dominant group of males modeled support and cooperation. Autonomy. The music teacher would place several new songs in the students’ folders contained on the program each week. The interactions amongst the peers focused on what new songs they had and which song th ey were going to sing first. There was a lively discussion each week with the new songs the students could choose from in their folders. Some students would have similar songs; however, the music teacher tried to keep it interesting by varying songs within th e individual students’ instructional reading levels. Although the peers were often heard discussing what songs they got with one another, when it came to deciding on whic h song to sing the choi ce was individual. The following is an example of two peers from the observation notes: Peer 1: What songs did ya get? [Peer 2 shows the list of songs] Peer1: I got that one too! Peer 2: I am gonna do Home On The Range first. Peer 1: Not me I’m gonna do this one. [Peer 1 points to a song -then they both turn to the computers] (Week 3, April 16, 2007)

PAGE 216

203 The peers appeared to be motivated by the different songs they got each week. Although this choice of material was contro lled (the teacher placed songs in their folders), it appeared sufficient to keep th e students interested and motivated in the interactive program. In addition, even though th ey would discuss what songs they each got, they appeared to be comfortable with choosing what song they wanted to work on individually. The opportunity for choice appeared to contribute to holding their interest and keeping the students motivated. Intrinsic Motivation The game-like quality and different materials afforded to the students were contributing fact ors to their continued use of the interactive sing-to-read program. There was a shift, however, that oc curred around the end of the third week of the intervention, when these students’ motivation became internalized. No longer were the discussions about what score they had comp ared to their peers, instead they became individually focused, self-regul ated, and engaged in their own achievement. They would be completely engrossed with their own songs and pitch scores regardless of what was happening around them. One example of this is cited below: He was focused on rereading the song. This was the fourth time he recorded himself singing. He had the screen up that displayed his vocal tract. He used his finger to align where he was off pitch. He went back again and reread the song [He moved back and forth in his seat, nodding his head to the beat, and tapping his foot to the music] Finally, [he sub vocalized] I got a 90! (Week 4, April 25, 2007).

PAGE 217

204 The above example was representative of the students in this intervention. The shift from motivation that was extrinsic, to getting a high score and competing with the other peers, shifted to a form of internal competition. The students became focused on the task and interactive with thei r own learning. As noted in the above example, this student was rereading, applying strate gic processes, and regulati ng his learning for his own purpose. Interestingly, his focus appeared to be on comprehending the rhythm and beat of the song. Singablity vs. Readability Twice a week for seven weeks, the students would enter the computer lab to use the interactive sing-to-read program, Tune Into Reading. In the beginning of the intervention, the music teacher worked with the whole group, giving direct instruction on how they were to use the inte ractive program. Along with di scussing the protocol use of the program, she made only one song accessible for all students during the first two sessions of the intervention. She chose Hot Cr oss Buns because as she explained to the students: I have put only one song in each of your folders on the program. It is the same song for everyone to try. I picked this song because it has a steady beat and the words repeat themselves. Therefore, you will be able to feel comfortable while you are learning to us e the program. (Week 1, April 2, 2007) It appeared the music teacher felt that by using this song because of its low readability level (second grade) and limited change in octa ves, pitch, and rhythms, the students would be able to concentrate on lear ning how to use the program and not have difficulty with singing and recording the song. In addition, the students would have

PAGE 218

205 success in their pitch accuracy scores when they sang and recorded the song because of its easy accessibility for the divers e reading levels within this group. Students Perspectives of the Alternative Text. While watching the students use the sing-to-read program, it appeared the students were adept at using the computer. They could easily manipulate this digi tal text, by adjusting the song s’ speed, page size format, and much more. They would show one another so me of the different difficulties that they had encountered and how to work around them. They were less sure, however, about the genre of rereading through singing, particular ly matching the rhythms, pitch, tempo, and beat of the songs to the words they were singing. Once they were comfortable with using the program, a little more then half of the students (approximately 59%) would skip th e procedure for listening to the background music and rereading the text silently. Then, they would complain to the music teacher about their low pitch accuracy scores. On e session during the third week of the intervention, the music teacher went over the importance of listening to the background music and rereading the song s ilently several times. She to ld the students that whether you are reading a book, or singing a song, you have to have the beat in you head. The only way you get the sounds in your head is by practicing. This helps you to know when you stop and take a breath, or stress certain wo rds. In addition, you would know how fast or slow you should read the words, followed by the sentences, and finally sing the song in its entirety. This will help you understand what you are reading or singing, and it will improve your pitch accuracy scores. However, the students were not convinced that this was related to reading a book as shown in the following excerpt:

PAGE 219

206 Peer 1: Yeah but singing is not like real reading [The group shakes their heads in agreement] Peer 2: Real reading is like a text book Peer 3: Yeah you read the book and answer the questions Peer 4: This is good for singing-but not reading a book (Week 3, April 16, 2007) It appears the students’ view of music as an alternative text was not a task related to reading. Their percep tion of what constitutes real read ing is ingrained in their school experience. However, the music teacher cont inued to make a point about the importance of prosody and rereading, whether yo u are reading a book or singing a song. She suggested to the students they conduc t an experiment. She asked the students to record Hot Cross Buns (a song they we re all very familiar with) first with the background music and then without it. Then, she asked them to reread the song first three times while you listen to the background music, and then record th emselves with the music and then without. After you record yourself the music teacher told the students to write down their scores with the background music and then without. When the students finished, she wrote the scores up on the board and averaged the scores with background music and the scores without. What they found was that the average pitch scores without background music was M=25; whereas, the av erage score with the background music was M=73 for the students. After the stude nts finished, she asked them to make comments on what they found. The following is an example of the comments made by the students:

PAGE 220

207 Peer 1: I guess we need to rere ad stuff to remember how it sounds. Peer 2: The beats help us sing the song and read the words. Peer 3: I guess I can see how this could help... when you are reading books too. Peer 4: Like if you’re reading and you don’t stop you like can’t remember. Peer 5: The music in the background kinda helps you to reach the highs and lows. (Week 3, April 16, 2007) The students were provided with an o pportunity to see the importance of rereading and the prosodic f eatures of text. They recognized the purpose of having the rhythm and beat in their heads helped them not only with their pitch scores but also when reading for meaning. In addition, their percepti ons of what constituted a real reading task, regardless of its alternative format, were brought to the forefront. This was accomplished because the music teacher took the time to s how the students why it was important to reread text and listen to the background mu sic. She provided clear goals and objectives explaining why they were following the procedur es for the literacy task rather than just assigning and telling the students to just follow the directions. As a result, the students understood the purpose and could adjust their view of what and why they were being asked to do during the task. Disequilibrium. As the songs became more difficult the students often complained to the music teacher about how hard it was for them to sing the songs and get a good score. The music teacher sat with the individual students and had them sing the songs so that she could provide assist ance. Often the students would discuss with one another how hard it was to sing the songs. However, no matter how difficult the singing got the

PAGE 221

208 students would persevere. The following is an excerpt taken from the observational notes between two peers: Peer1: I can’t get this song. Peer 2: Yeah it is really long. Peer 1: I need to read it a ton of times before I record it. Peer 2: Yeah… you could slow it down too. Peer 1: I’m gonna try it again Peer 2: Go for it! (Week 4, May 3, 2007) As with any new learning, in reading as the task became more difficult the students need the stamina to continue. The disequilibrium that occurs with all new learning and the perseverance to continue is what makes learning successful. The example above shows the students were awar e that the material was getting more difficult, yet they still opted to continue with the task of rereading through singing. Summary of Results for Treatment Condition Many of the constructs that evolve d from the observational data are well documented in the literacy research (e.g., Guthrie & Wigfield, 2000) as effective practices to meet the needs for the early adolescent learner. Specifically, the students should be motivated and engaged in the liter acy task presented, so they could achieve academically. Interestingly, the shift from an extrinsic form of motivation (motivated because of reward or punishment) to an in trinsic motivation (motivated because they want to do this above anything else) was when it appeared that the students really became engaged in the task.

PAGE 222

209 The opportunity to make a choice of text also supported the student’s motivation. Each week the choice of songs provided au tonomy for the students and appeared to increase their motivation to continue using the program. In addition, although this was an independent task, there was considerable stude nt collaboration and s upport in a safe riskfree environment. It was interesting to note, how ever, the peers’ social system that was in place. The peers’ hierarchy and the passive aggres sive stance taken by the students, as to whether or not they should partake in the tas k, are not as well docum ented in the research (Ryan, 2001). The students “buy-in” was nested in whether the dominant characters supported or rejected the task assigned. This silent cueing system should be considered when instructing this group of literacy lear ners. The outcomes could have been very different if the students influenced by the dominant characters had rejected the sing-toread program. The use of the alternative textual format and the genre it delivered was interesting to see and hear the student’s perceptions as to what constitutes real reading. It was just as interesting to see the perspectives change a bout the task of reread ing and the place that prosody has in understanding text. After th e music teacher showed them how prosody and rereading affects their reading and singing, it was obs erved the students would constantly go back and reread the text befo re recording. Interesti ngly, when the students were provided with clear objec tives for the task by the music teacher, there was a mutual understanding of the expected outcomes. In a ddition, the definition for alternative text in this study changed to include not only the format (digital), the genre (songs), but also the perspectives of the students as explained in more detail in Chapter 5.

PAGE 223

210 These descriptive findings substantiate the statistical results previously reported in this chapter. That is when the students ar e motivated, have choice of text, have diverse and interesting textual formats, opportunities for peer collaborati on, and understand why they are doing the literacy task, their academic achievement will improve (Guthrie & Wigfield, 2000). Specifically the treatment group of students increased significantly from pretest to posttest in fluency (WPM) p<.001, word recognition (WR) p=.009, reading comprehension (COMP) p<.001 at the same inst ructional level attain ed at pretest. In addition, at the increased read ing level from pretest to pos ttest the treatment students increased in their instruction reading le vel M=1.13 years within this seven week intervention. Interestingly, at this increased readin g level as disequilibrium occurred and the students were building the stamina to sing more difficult songs, their mean scores in all areas (WPM, WR, and COMP) declined. Althou gh the students had increased in their instructional reading level from pretest to posttest, when comparing scores on the same instructional reading level attained at pretest to posttest scores at the highest reading level their scores decreased. Specifica lly in: (a) fluency (WPM), at the initial posttest M= 160 wpm to M=147, at the increased reading level posttest, (b) word recognition (WR), initial posttest M= .99, to M= .98 at the increased r eading level posttest, and (c) comprehension (COMP) at the initial posttest M= .85, to M=.75 at the increased reading level posttest. This suggested that, as the early adolescents in the treatment condition increased in text difficulty, their fluency (WPM), word r ecognition (WR), and comprehension, (COMP) shifted from a fluent expert reader at one level to a dysfluent reader (e.g. Topping, 2006) at a higher level.

PAGE 224

211 Control Group Description of Classroom Routine There were 32 students in the contro l group during this study. Students would enter the classroom and choose their drum or other instrument that they used for the lesson. Then, they would get a chair and place it in one of the three semi-circle stadium steps in the classroom (Chapter 3 room desc ription). The music teacher would bring the class together with a beat of her drum in the center stage of the music classroom. The students would echo back the beat and class would begin. Data Analysis The data collected from control group came from 14 classroom observations. The classroom observations took place in the music classroom during the fourth quarter of the 2006-2007 school year (April 2nd May 15th, 2007) over the 7-w eek experimental treatment period. As previously noted, there were two classes randomly assigned by class to control conditions. Two classes were combined and became the control group. Observation occurred twice a week for the co ntrol group from the week of April 2, 2007May 15, 2007. Observational field notes were taken dur ing each class session twice a week, during the 50 minute class periods for each of the 2 classes assigned to the control condition. The field notes were taken on a pa d of paper during the Wheel Music Class periods noting time, place, attendance, and all the major character interactions during the observations. The focus of these observations was to describe the relationship, if any, between the literacy task the music teacher assigns (reread ing through singing) and the

PAGE 225

212 peer interactions (e.g., peer talk, peer mode ling, and peer social re inforcement) among students who are singing in th e regular music classroom.. There were 15 constructs that emerged from the control group’s observational data. The constructs were tabulated to dete rmine the frequency of each construct in the data. Constructs that had a fr equency count less then five were not included. Table 26 specifies the 6 constructs that emerged from these data with the highest frequency. Refer to Appendix C for a description of the constructs. Table 26 Constructs from the Control Group Observational Data Construct Frequency Extrinsic Motivation 34 Alternative Approaches to Singing 24 Dominant and Vulnerable Peer 21 Fake Rereading 18 Disengaged 13 Peer Leaders 8 There were three themes that emerged from the control group’s classroom observational data. They were developed af ter a thorough analysis of the data, which included reading through the data at least th ree times for a holistic sense of the data, analyzing the data for mean ingful units, developing constructs from the emerging meaningful units, and tallying the constructs for frequency. The themes that emerged were Engagement, Group Formats and Readi ng Strategies. These themes encompassed the essence of peer interactions during the control groups’ singing durin g their regular music period. Table 27 presents these themes and the frequency with which they occurred in the data collected from the control group observational data.

PAGE 226

213 Table 27 Themes from the Control Groups’ Observational Data Theme Frequency Engagement 58 Group Formats 29 Reading Strategies 33 Engagement Alternative Approach to Singing. The music teacher wanted her students to be engaged and involved in the Music Wheel Cl ass. She felt learning music theory and different aspects of singing for this group of students would not hold their interest. Specifically, this was because this group of students had not chosen singing as an elective, instead they were assigned to this elective. Therefore, she decided that a handson interactive alternative format would be mo re successful. The music teacher decided to use a drum circle. Not only would it build a sense of community for the students, using drums would also involve them in singing and creating their own music. Initially, a simple drumming sequence was taught to the students, and then this was followed by learning the multiple stanzas of the three songs during the seven week intervention. In order for the students to learn the drummi ng sequence, the music teacher modeled the pattern of beats, and in turn, th e students would echo in response the same pattern. She started very slowly at first and then would incr ease in speed. The students appeared to be engaged as they would follo w her every beat. The following is an example from the observational notes:

PAGE 227

214 The students would hit their drums echoing the music teacher. Their backs were arched up straight. If they lost their place, they would stop and tap their foot until they caught the beat Their faces were serious and intent on following the lead of the teacher. (Week 1, April 5, 2007) The students were engaged and on task as they were echoing the music teacher’s drumming sequence. They were focused and ap peared to comply with the music teacher directions. The use of the alte rnative approach to teaching si nging appeared to hold their interest and keep them engaged and focuse d on the task. There were no interactions among the peers while they were drumming, in stead the group had remained serious and focused. Their attention was on concentrati ng, observing, and listening to the music of the drum sequence played by the teacher. On ly after they finished did you hear the students laugh or make comments to one another. Extrinsic Motivation Once the students were able to perform the simple drumming sequence, they were taught the songs to accompany it. The goal was to keep the rhythm and beat of the song while si nging by drumming. In order to assess if the students had accessed the song and its corresponding drum sequence, the music teacher had the students take turns in small groups a nd perform for the rest of the class. This round robin routine by the small groups was motivating and highly competitive for the students. When all the groups had a turn, th e peer interactions were often comments regarding who performed the best during th e challenge. The following excerpt between two groups was an example of this: Group 1 Peer: We did better than you! [Group 1 members cheered]

PAGE 228

215 Group 2 Peer: No way… you guys messed up big time! [Group 2 members shout yeah yeah] Group 1 Peer: How did we mess up… your crazy [Music teacher stops the interactions] (Week 2, April 9, 2007) The students were motivated by the challeng e of competing with one another. The prize of being the best was the goal. During this competition, each group of students were actively engaged and tried to do their best. Th is was a motivating activity for all of the groups, and the reward was to perform, be the best, and get it right. Group Formats Dominant and Vulnerable Peers Instructional delivery fo r this group of students was primarily accomplished through a whole group format. Each session the students would follow the same procedures. They would sing and play the drums echoing the music teacher. During the session, the music t eacher stopped on occasion and addressed the students if she heard the drumming or singing off key. The students often made comments to one another, blaming them for making a mistake. It was the more dominate students in the group who addressed the mo re vulnerable students. However, these exchanges were not loud enough for the musi c teacher to hear. Instead the exchanges were accomplished secretly and critically as they blamed one another for making mistakes. The following is an example of one dominant peer admonishing a more vulnerable student for making a mist ake while drumming and singing: Peer1: You made the mistake. Peer 2: No I did not… shut-up.

PAGE 229

216 [Peer 2 makes a nasty face at Peer 1] Peer 1: Yes you did I heard it you…jerk [Peer 2 turns red and puts her head down] (Week 3, April 16, 2007). Although the teacher did not single out any one student for making a mistake, the peers would blame one another. The cri ticism occurred often throughout the sessions between dominant and vulnerable peers. However, because it was done secretly, the teacher was not aware of what was happening. In addition, when the vulnerable peer was admonished, the other peers seated near or around the student who had been blamed for making the mistake did not say a word. Instea d, they would look at one another or look away when this happened. Peer Leaders. There was two occasions during the seven week intervention that the peers broke-up into small cooperative gr oups. The task was to create a new drum sequence that would accompany the song they learned in class. The music teacher selected the students for each group and then to ld them they had 30 minutes to complete the task. As the students were getting ready to join their groups, they were told that when they were done they would perform thei r creation for the rest of the class. Once the students were in their groups there was one dominant peer who would take control and lead the rest of the group. The peer leaders were self-designated eighth graders; however, they were not of a partic ular gender. When the peer leaders were in groups, they would take over and direct th e other group members. They organized and managed the other peers, so the task assigned was accomplished. In addition, students in the group who did not cooperate were reprimanded because the goal was to complete the

PAGE 230

217 task assigned. The following is an example of the peer interactions within the small groups: Peer Leader: Okay, let’s st art with a high drum beat. Peer Member 1: Let’s hit it twice on high. Peer Member 2: Sounds good, let’s try it. Peer Leader: M are you with us? Peer Member 3: What if we hit the side like this. [Peer 3 demonstrates for the group] Peer Leader: Okay let’s try it. [She stops and speaks to M again] Peer Member 4: That sounds good. Peer Leader: Okay let’s sing it with the song…Go. (Week 4, May 1, 2007). When the students were in the small c ooperative group formats, there was a peer that assumed the role of leader. This was not an assigned position, instead it was allowed position by the rest of the group. The lead er managed, organized, and kept the group focused to complete the task. Although the leader was either male or female, they were the dominant characters within each of the groups. In addi tion, even though the decision making appeared to be collabora tive pertaining to the creatio n of the drum sequence, the final approval of what and how they woul d perform the drum piece was made for the group by the peer leader.

PAGE 231

218 Reading Strategies Fake Rereading. During the interven tion there were three songs taught to the group. Initially, the music teacher put the song in its entirety on the overhead projector for the group to read. Then, she went over is olated vocabulary words she felt the students needed explained, so they would understa nd the song. This was followed by the music teacher’s use of modeling each of the stanzas of the song, and in turn, the students would chorally sing and echo back what she sa ng. Finally when the music teacher felt the students were able to sing the song, she would have them play the accompany drum piece to go along with their singing. For each session, the students would reread the song by singing each stanza and playing their drum sequence. The drummi ng supported the students’ singing by providing the prosody needed to keep the rhythm, volum e, and pitch of the song. However, even though the students knew the word to the songs as they reread through singing each session, often the teacher would stop the group and state she c ould not hear their voices. The following example is taken from the obs ervation notes as the peers were charged with rereading (re-singing) the song: The students were playing the dr um sequence and moving their lips as if they were singing. Of the 20 students in the group, only about 6 were actually singing. The music teacher woul d stop the students and address the group reminding them to sing. However, this fake rereading through singi ng continued. (Week 4, May 1, 2007) However, this fake rereading throug h singing continued. Although the students knew the words to the song and the correspond ing drum pattern, when it came time to

PAGE 232

219 reread (re-sing) the song, they would play the drums but not sing. The more comfortable they became with what was required, the more they did not have to focus on what was happening. Even with addition of new songs or drum patterns, the learning became routine, and the students fa ked rereading (re-singing) th e songs (Tovani, 2000). Disengaged. The songs and drum patterns were taught to the student by having the students echo what the mu sic teacher played and sang. Th e students would first listen to the teacher and then chorally sing and echo the back the stanza and the corresponding drum pattern. Towards the end of the seven week intervention, of ten the students would be seen daydreaming as they went through th e motion of singing and playing their drums. Some students would be silently whispering to one another, and still others would be playing around and making up their own drum patterns. The following is an example from the observational notes on how the st udents were performing during the daily session towards the end of the intervention: He would hit the drum while looking at the door. She was singing and looking straight ahead however, when the teacher spoke she became startled. They were whispering to one another exchanging ideas about what to wear to the dance. The two young males were laughing softly pretendi ng to play their drums with another pattern. (Week 5, May 3, 2007) It appeared the students had shifted from being very focused and engaged initially to more automatic in their response to th e music teacher reread ing through singing. They disengaged from the task they were performing. The learning became routine and presented the students with no struggle, challenge, or motivat ion to continue. They were off task and not engaged with the task the music teacher had them perform.

PAGE 233

220 Summary of Results for the Control Group The students in the control group appeared to be motivated and on task when they were using the alternative approach to rere ading through singing, initially. The use of the drums appeared to hold the students interest In addition, the opportunity to create their own drum sequence was extrinsically motivating. They were motivated to be the best and to sing the songs and reproduce the drum patterns correctly. This was apparent with the light-hearted interactions as they competed with each other during the round robin performances as the music teacher assessed their learning. This competitive banter appeared to be ex trinsically motivating and engaging for the students. It kept them focu sed and aligned with the objec tives of the lesson. However, what started out as light hearte d competition soon turned to critical analysis between the peers. They would blame one another for mi stakes made during the performance. This was done secretly without the music teach er’s knowledge of the interactions. The interactions became uncaring and unsupportive as they would blame each other for not performing correctly. Although th e music teacher did not single out a peer for making a mistake, through he r actions she reinforced ther e was only one right way to perform the drum pattern and sing the s ong. This was done by stopping the class and trying the procedure again, ofte n. The dominant peers blamed the more vulnerable peers while the other students would avoid becomi ng involved or supporti ng the peer needing support. These dynamics occurred consistently during the interventi ons, and even though the music teacher wanted the students to ha ve a sense of community by using the drum circle, this appeared not to be the case.

PAGE 234

221 There were students in the control gr oup that immediately took on the role as peer leaders, even without that role being de signated to them. It appeared to be assumed by the other members in the group that this person was in charge, and there was no questioning this position. In addition, deci sions made in the small group format did appear to be collaborative; how ever, the final decision of what to do and how it was to be accomplished was determined by the peer leader. This position was an excepted and allowed by the rest of the peers. The use of rereading through singing was accomplished by having the students echo chorally back the drum patterns and the song lead by the music teacher. Although the students were engaged and motivated in itially, they became disengaged with this routine. The data suggested the students becam e complacent in the task when the learning became unchallenging. These descriptive findings concur with th e statistical findings previously reported in this chapter. That is when the students become unmotivated, and disengaged, and the classroom environment does not provide oppor tunities for peer collaboration, their academic achievement will not improve (Gut hrie & Wigfield, 2000). Specifically, the control group of students did not increase significantly from pretest to posttest in fluency WPM; p=.219, word recognition WR; p=.379, reading comprehension COMP; p=.170 at the same instructional level at tained at pretest. In additi on, at the highest reading from pretest to posttest, the contro l students did not increase in th eir instruction reading level from M=5.58 at pretest to M= 5.77 at postt est within the seven week intervention.

PAGE 235

222 Cross Case Analysis After a thorough analysis of the data for each of the cases, there were similarities and differences across cases. In order to capture the multi-dimensional and complex nature of peer interactions, these similaritie s and differences were described through the theme Social Systems that appear to capture the esse nce of peer inte ractions for the treatment and control groups. Specifically, the social structure of the peers through their interactions appeared to influence the task of rereading throu gh singing. Within this theme, there were four constructs embedded. They were: (a) Peer Positions, (b) Instruction Expectations (c) Alternative Approaches to Tasks and (d) Reading Strategies Therefore, analysis for this section wi ll describe how the two groups displayed similar and dissimilar characteristics of peer interactions within this theme and across these constructs. Social Systems Peer Positions. The treatment and control group had in place a social system that positioned some of its peer members in the role of dominance over the other peers. These dominant characters held this position, and the other peer me mbers allowed them to assume it. Both groups contained this two class system where a small group or individuals lead the rest of the group pertaining to acceptabl e social behavior. Interestingly, however, within the treatment group and the control gr oup, the interactions from the dominant peers with the other peer members were accomplished very differently. Within the treatment group, the peers modele d behaviors or talk that resulted in the rest of the peers imitating their behaviors. There were no discussions with the other

PAGE 236

223 peer members only among the small group of eighth grade males. Whereas, within the control group, the dominant peers of various gender (male or female) specifically directed the other members to conform to a certain behavior they deemed socially acceptable. This may have been as a result as to how these dominant peers interpreted what was expected of them through the instructional delivery provided. Instructional Expectations. The role of dominant peers remained constant within the treatment and control groups, during the intervention. However, the instructional expectations afforded to the groups by th e music teacher were very different. These expectations appeared to be interpreted by the dominate peers and then reinforced through their interactions with the rest of the peer group. What appeared to be expected of the peer s in the control group was that there was only one right way to perform the singing and drumming. These expectations were modeled to the peers as the music teacher would stop the singing and drumming several times daily during each session and tell the st udents that some people are off key, try it again. In turn, the dominant peer would ad monish the vulnerable peer for making a mistake. However, the treatment group was exp ected to try their best. The music teacher would often remind the students to try their best and not to worry about their pitch scores. The dominant males in the treatment group would encourage and support one another, modeling collaboration to the othe r peers. Therefore, it appear ed that within both groups the dominant peers interpreted what was expect ed of them and then in turn reacted to these expectations thro ugh their interactions. The control group peers were expected to perform correctly and accurately as modeled by the music teacher. Since there wa s only the right or wrong way to perform

PAGE 237

224 the song and drum sequence, the interacti ons might have been interpreted by the dominant peers and conveyed to the rest of the group in this manner through their interaction. However, the treatment gr oup was allowed differentiation through the instructional delivery. Therefore, the dominant peers appeared not to be compelled to take on the task of reinforcing group accu racy; instead they became a group member while still maintaining th eir position among the group. Alternative Approaches. Both groups found the alterati ve approaches to learning motivating and engaging. This was apparent during the interactions within the groups. The light-hearted competitions through inter actions were documented in the data as the groups led by the dominant peers would eith er through discussion or modeling set the climate of motivation for rest of the group. However, as the inte rvention continued, a shift occurred within both of the groups as to their motivation for these alternative approaches to the task of rereading through singi ng, as did the role of the dominant peers. The peers within the control group became disengaged towards the task, including the dominant peers within the group as th e sessions progressed in time. The data suggested that around the four th week as the sessions continued the peers would daydream, talk, and entertain each other dur ing the sessions. The motivation levels shifted from highly motivated to complacent. In contrast, the treatment group of students’ motivation shifted from external motivation to internal for all th e peers, including the dominant peers. They became engaged in the task and self-regulat ed in their learning. This might have occurred as a result of how the strategic process for reading unfolded. Reading Strategies. Fluency instruction for the students was the same in both groups. The music teacher used repeated readings of the songs, while embedding

PAGE 238

225 prosodic features of text. The students reread (re-sang) their songs three or more times each session and each group was supported with the prosodic features of songs through background music, rhythm, tempo, pace, and vol ume. The control group used their drums and followed a modeled example of prosody from the music teacher; whereas, the treatment group prosodic elements were c ontained in the background music from the Tune Into Reading program. This process as reported in the literature (e.g., Samuels & Farstrup, 2006) should improve the student s’ automaticity (WPM), accuracy (WR), reading comprehension (COMP), and instructi onal reading levels (RL) for both of the groups. However, this was not the case for these two groups. Summary Cross Case Analysis The descriptive findings did support the st atistical results previously reported; however, these findings did not c oncur with the findings cited in the literature in entirety. Within the groups, the treatment group displaye d a statistically signi ficant difference in fluency (WPM) p<.001, word recognition (W R) p=.009, and reading comprehension (COMP) p<.001; whereas within th e control group, they did not in fluency (WPM) p=.219, word recognition (WR) p=.379, or r eading comprehension (COMP) p=.170 on the same instructional level at tained at pretest. This sugge sts that there may be other contributing factors for the students to be fluent readers. One potential factor, as suggested in the descriptive findings, is an environment that is safe and supportive and instruction differentiated to meet th e needs of all of the students. In addition, the statistical findings show that when the treatment and control groups were compared the treatment, stude nts had a significant increase in reading comprehension (COMP) at p<.001 and instru ctional reading level (RL) p<.001 as

PAGE 239

226 compared to the control group. According to Ra sinski (2004), it is the prosodic features of text genre that assist in reading comprehension; ther efore, these findings suggest prosody was a contributing factor in the in crease of reading comprehension for the treatment students. Yet, this appeared not to be factor for the control group. The descriptive findings sugg est the treatment group in ternalized their lear ning as the sessions continued; however, the contro l students disengaged from tas k. This could be interpreted to mean that prosody needs to interact with th e learning and not be passive so that reading comprehension can occur. Therefore, although both groups were following the protocol for reading fluency improvement, the students singing through rereading alone, as in the case of the control, did not improve reading comprehension, which is the goal of fluency instruction. Chapter Summary In this chapter, I answered the three rese arch questions after an in-depth analysis of the statistical and observati onal data from students in th e treatment condition, using the interactive sing-to-read program, Tune Into Reading, and their counterparts singing as part of their regular music program. The st atistical analysis was conducted on the first two questions initially investigating the difference in reading outcomes in fluency (WPM), word recognition (WR), reading co mprehension (COMP), and instructional reading levels (RL). I administered pretest and a posttest measured by the Qualitative Reading Inventory-4 (QRI-4) (Leslie & Caldwell, 2006) and compared the posttest scores with the pretest scores to determine if students in the expe rimental group gained significantly over their counter parts in the control group. In itially, the students were assessed at posttest with a r eading passage on the same instru ctional level attained during

PAGE 240

227 the pretest. This was followed with statistical analysis at the highest instructional reading level attained by the students. Then, the students were grouped according to their FCAT reading level scores. FCAT level reading scores (level 1-5) range from highest level (5) to lowest level (1). The treatment and control groups were stratif ied according to their FCAT level as: (a) Level 4 or 5 Above grade level, (b) Level 3 At grade level, and (c) Level 1 or 2 Below grade level. Once the students were grouped, percentages of students for each group were calculated for the students at each level. The percentages showed an equal distribution of FCAT level reading scores between the two groups; however, eight students were missing FCAT level reading scores (four treatm ent and four control). Therefore, only 56 out of 64 students’ data were analyzed. Finally, question three investig ated the peer interactions that occurred during the study intervention. There were 14 rereading th rough singing sessions for the students in the treatment and control groups that occurr ed for 50 minutes each, twice a week, over a seven-week period. During these sessions, obs ervational notes were taken on the peer interactions that occurred during these sessions. Th ese observations focused on describing the relationship, if any, between th e literacy task the music teacher assigned (rereading through singing) and the peers’ inte ractions during the task. There were two cases in this study. The treatment students were singing using the interactive program Tune Into Reading, and the control students were reading through singing in the traditional music class. Priority was given to the qua ntitative approach because it looked at the statistical relationship between th e sing-to-read program, Tune Into Reading used by the treatment

PAGE 241

228 group compared to their counterparts in th e control group rereadi ng through singing in the regular music class. The reading outco mes from the students of varying reading abilities were measured by the QRI-4. The anal ysis for this approach was executed first to answer the first two questions of this st udy. However, concurrently qualitative case study methods were used to better understand an d describe the peer in teractions occurring during the literacy task assigne d by their teacher. The integration of the two types of data occurred within the qualitative findings sect ion of this chapter and used a triangulation strategy to interpret the findings. This integrated the statistical results with the descriptive findings in order to answer the research questions of the study. The study findings indicated that the middl e school students of varying reading levels significantly improved in their reading fluency scores through the use of the interactive sing-to-read program Tune Into Reading, compared to the group who were rereading through singing in the regular music classroom. In addition, prosody appeared to have a direct connection to reading comprehension. Furthermore, the use of the interactive program provided oppor tunities for differentiated reading level achievement. Finally, group dynamics highly influenced the early adolescent’s motivation, engagement, participation, and successf ul outcomes in reading fluency.

PAGE 242

229 CHAPTER FIVE: DISCUSSION Chapter Five provided a disc ussion of the study results. There were five sections within this chapter. The fi rst section summarized the study. The second section described the conclusions and implications derived fr om the research findings. The third section discussed the contributions this study makes to the existing body of knowledge on reading fluency with middle school students of varying reading abilities. Along with the discussion on reading fluency, a discussion of the findings related to the sociocultural interactions during the literacy task be tween the peers was included. Recommendations for practice derived from the research findings and the study’s conclusions and implications were in the f ourth section. Finally, the fift h section provided suggested recommendations for future research. Summary of the Study Fluency research suggests a fluent read er is one who can read a text with automaticity, accuracy, and proper expression, wh ile viewing comprehension of text as the ultimate goal (LaBerge & Samuels, 1979). The methodology most noted in the literature to support fluency inst ruction is the process of rereading text, three or four times. Rereading affords the students quicker, more accurate, and better sounding reading. The literature on fluency also sugge sts a fluency model should be provided so students can hear proficient or al reading that captures all the elements of what fluent reading sounds like. Rasinski (2004) conte nds that utilizing a text with naturally

PAGE 243

230 embedded features of prosody, such as poe try, speeches or singi ng will assist with building fluency in the readers. However, th e assumption is often made that by the time most students enter middle school, they are fluent readers and comprehenders across a variety of texts (Alvermann, 2001). This is es pecially true of those students deemed proficient readers, determined by their yearly standardized test results. As a result, fluency instruction is often only provided to the students deemed less than proficient in their reading, according to the high-stakes test results. The purpose of this concurrent mixed me thods study was to i nvestigate rereading through singing with two groups of hete rogeneously grouped middle school students within a music classroom. The two groups were randomly assigned by class to a treatment group (n=32), that used an interactive sing-to-read program, Tune Into Reading (Electronic Learning Products, 2006), or to a control group (n=32) that were rereading through singing as part of their regular musi c program. All 64 participants were members of an assigned elective Wheel Music Class classroom during the fourth quarter of the 2006-2007 school year (April 2nd May 15th, 2007) over a seven week experimental treatment period. The Qualitative Reading Inventory-4 (QRI-4) (Leslie & Caldwell, 2006) was utilized to measure from pretest to postte st the performance in fluency, WPM (measured by words per minute), word recognition, WR (measured by oral reading accuracy), reading comprehension COMP (measured by implicit and explicit questions after the reading), and instructional reading level, RL (measured by combining scores from word recognition and comprehension questions) befo re implementation. Initially, the students were assessed at posttest with a reading passage on the same instructional levels attained

PAGE 244

231 during the pretest. This was followed by analys is at the students’ highest instructional reading level. Concurrently, this investigation provided a description of the peers’ interactions in both groups during the literacy task assigned by the music teacher. The intent of this study was to address the fo llowing research questions: Quantitative Research Questions 1. To what extent is the reading perf ormance of word recognition, fluency, comprehension, and instructional reading level, as measured by the QRI-4, of students using the Tune Into Reading program, different from their regular music curriculum counterparts? 2. To what extent does the Tune Into Reading program differently impact the reading scores of students who are “b elow, at, or above” grade level as determined by the Florida Comprehensive Assessment Test (FCAT) reading scores? Qualitative Reading Question1. How do middle school readers interact w ith their peers within the context of their music classroom? Question one addressed the differences in reading performance for the students using the interactive sing-to-read program, Tune Into Reading, compared to the students who were rereading through singing in thei r regular music class. This comparison measured the students in their fluency, WPM (measured by words per minute), word recognition, WR (measured by oral reading accuracy), reading comprehension COMP (measured by implicit and explicit questions af ter the reading), and instructional reading

PAGE 245

232 level, RL (measured by combining scores from word recognition and comprehension questions). Question one findings revealed the treatment students of vary ing reading abilities that used the interactive singing program, Tune Into Reading, illustrated a significant increase in their Fluency (WPM), Readi ng Comprehension (COMP), and Instructional Reading level (RL) as compared to their c ounterparts who were singing in the regular music class. In addition, for the treatment students, Word Recognition (WR) indicated a larger effect from pretest to posttest than the control group. Specifically, this suggests that rereading through singing, usin g the interactive singing program, Tune Into Reading, was more effective regardless of the reading levels for treatment students compared to control students. These results can also be interpreted as re reading through singing in the music classroom alone, as was the case for the control students, does not improve WPM, WR, COMP, and RL for the students of va rying reading abilities. Therefore, these findings suggest that regardless of their read ing levels early adoles cents benefited from fluency instruction. Furthermore, at the highest reading level reported at pos ttest, although the treatment group had a significant increase in thei r instructional readi ng level (RL), it was reported there was no significant difference be tween the groups in WPM, WR, or COMP. The descriptive findings suggest ed the treatment students we re interactive with their learning. They appeared to assimilate and accommodate the new learning from the text while they were rereading through singing. Howe ver, as the new material became harder they shifted to a state of di sequilibrium (Piaget, 1964). Th is might be interpreted as reading fluency is not a static condition; instead, it is flui d and continually developing.

PAGE 246

233 Toppings (2006) suggests an early adolescent can be a fluent reader at one level and yet display dysfluency at a higher level. Thi nking about reading fluency using Toppings theory, it then might be inferred for this gr oup of literacy learners that reading fluency should be thought of as a strategic process, rather than a skill acquired through repeated practice alone as reported in th e literature (e.g., Samuels, 2006). Finally, it was reported that within the groups, the tr eatment group illustrated a significant increase in fluency (WPM), word recognition (WR), reading comprehension (COMP), and instructional reading level (R L); whereas, within the control group, there was no significant increase from pretest to pos ttest in any of these areas. This suggests that within the groups, during the literacy task of rereading through singing, something happened within the classroom culture of the treatment group that was different from the control group. Further analysis re vealed the peers’ social inte ractions within the treatment group’s classroom culture might have contribut ed to the significant increases in all variables. Particularly, the p eer interactions appeared to be supportive and collaborative. These results suggested the act of literacy was embedded with in this network of social relations. Moje (1996) contends that in the secondary conten t classroom, it is the social context that shapes the literacy pr actices for the early adolescent. Question two used the same scores from the QRI-4; however, the students were grouped by their 2006 FCAT Reading Level achievement scores. The FCAT reading scores (levels 1-5) range from highest level (5) to lowest level (1). The treatment and control groups were stratified according to their FCAT Levels as: (a) Level 4 or 5 Above grade level, (b) Level 3 At grade level, and (c) Level 1 or 2 Below grade level. Once the

PAGE 247

234 students were grouped, a comparison was ma de between the groups on each dependent variable, looking at how each FCAT Level was differently impacted. The results reported the intervention was more effective for the treatment students that used the sing-to-read program, Tune Into Reading program compared to the control group. Interestingly, it was also noted that fo r the treatment students at the various FCAT Levels, the program used afforded them oppor tunities to improve differently in the reading components each level individually needed. Specifica lly, for the students grouped as Below grade level in their FCAT scores, the intervention was more effective in improving reading rate WPM and word accuracy WR. However, for the students grouped as Above grade level, the results reported re ading comprehension COMP was more effective. These findings suggest for the tr eatment students that used the sing-to-read program, Tune Into Reading, this interactive sing-to-read program was effective in meeting the differentiated needs for each level. However, when the FCAT Levels were used as benchmarks for the initial pretest of the groups, there was a discrepancy between the reported FCAT Levels and results of pretest scores from the QRI-4. Specifically, wh en all the participants were given the QRI4 pretest, there was no significant differe nce between the groups. Conversely, for the students stratified by th eir FCAT levels 3 ( At level) and 4 and 5 ( Above level), it was reported the groups had different pretest scores on their instru ctional reading levels (RL) than the control group. Specifically, showing for the treatment group on Level 3 ( At ), they had significantly higher pretest scores than the cont rol; whereas for control group Level 4 and 5 ( Above ), they had significantly higher pretest scores than the treatment students. This suggests using FCAT reading level scores as benchmarks to determine

PAGE 248

235 instructional reading level does not appear to correlate with the scores from the QRI-4 assessment. Question three investigated peer inter actions that occurred during the study intervention. There were 14 rereading through singing sessions for students in treatment and control groups that occurred for 50 mi nutes each, twice a wee k, over a seven-week period. During these sessions, observational notes were taken on the peer interactions that occurred during these sessions. These observations focused on describing the relationship, if any, between the literacy task the musi c teacher assigned (rereading through singing) and the peers’ interactions during the task. There were two cases in this study. The treatment students were si nging using the interactive program, Tune Into Reading, and the control students were reading through singing in th e traditional music class. As noted previously in the findings, it was suggested that during the literacy task of rereading through singing, the classroom culture and the occurrence of social interactions might have contribut ed to the significant increase s in all variables within the treatment group; however, this appeared not to be the case within the control group. Suggesting that within the tr eatment group during these soci ocultural interactions, the classroom culture supported academic improvement The findings suggested the treatment groups ’ classroom culture appeared to be safe, risk-free, motivating, and collaborative; whereas within the control group, the classroom culture was initially motivating, engaging, and competitive. In addition, it was found dominant peers within the treatment gr oup had no direct discussions with the other peers. Instead, they modeled support and collab oration with one another, and other peer

PAGE 249

236 members followed. The dominant peer interac tions with other peers within the control group however were through direct discussion as they told th e peers to conform to the literacy task assigned. This sugge sted the data revealed learni ng for the students in the treatment group progressed from engagement to assimilation, followed by self-regulation and interaction with text. The control group fi ndings revealed their learning shifted from initial engagement to fake reading to disengagement. These findings suggest these sociocultural interactions played an im portant role in improving fluent reading performance as noted in the treatment group scores. Discussion: Conclusions and Implications Addressing Early Adolescents Differing Fluency Development Biancarosa and Snow (2006) reported to the Carnegie Corporation, over 70% of adolescents struggle with their reading in some manner, and therefore, require differentiated and strategic in struction. Furthermore, they contend that when thinking about reading fluency for the early adolescents there are a range of literacy needs to be met for this population. Some students may s till need support with reading the words; whereas other students can read the wo rds accurately but need support with comprehension. Still, other adolescents ma y know the strategies but not have had sufficient practice within the cl assroom. What they need is instruction an d support that addresses the differing literacy needs for all students. As previously noted in the findings, for th e treatment students that used the singto-read program, Tune Into Reading, the program was effective in meeting differentiated needs for each level. For students grouped as Below grade level in their reading, the intervention was more effective in the r eading areas of Fluency (WPM) and Word

PAGE 250

237 Recognition (WR); whereas, for students grouped as Above grade level, the intervention was more effective on reading comprehension (COMP). This suggested the sing-to read program was effective for the reading areas each group of students needed and therefore addressed the range of differing needs. However, even when given the uniqu e individual differences among early adolescent literacy learners, curriculum de livery is often a one-size-fits-all practice (Alvermann, 2001; Ivey, 1999; Moore, 2000). Th erefore, the integration of content literacy to meet the diverse n eeds for this population is cha llenged through the contextual structure and curriculum delivery. This was evident in the classroom structure and curriculum delivery of the treatment and control groups. Small groups provided classroom structur e to the treatment group. The students worked in the computer lab in small gr oup communities, and curriculum delivery was accomplished through individual computer usage. However, the classroom structure for the control group was through a whole group format, and curriculum delivery was provided to the entire bo dy of students present. The findings suggested the treatment gr oup using the interactive sing-to-read program, Tune Into Reading, individually had higher read ing outcomes compared to the control group who were singing within a whole group setting. The inference is that in order to meet and address the reading flue ncy needs for this population of literacy learners, instruction and deliver y of curriculum needs to meet the individual needs of the students. This suggests that for fluency in struction to be succ essful, the curriculum delivery should provide opportuni ties for individual work.

PAGE 251

238 Assessing Fluent Readers in the Middle School Fluency is a necessary aspect of successful reading, as it allows the readers to read with speed, accuracy, and proper e xpression (National R eading Panel, 2000; Rasinski, 2004). The National Reading Panel (2000) reported they found sufficient evidence that guided oral reading through repe ated reading will have a positive impact on fluency and comprehension. However, the lit erature on reading fluency often focuses on the beginning reader’s initial stage of literacy acquisition or on the older adolescent reader who has difficulty learning to read. Th is focus has placed reading fluency in a deficit view, which focuses on remediation at the decoding level, ra ther than creating a direct link to comprehension (Clay, 1985). St ayter and Allington (1991) suggest that “we have failed to consider some of the broader ramifications of an emphasis on fluency, especially with older and more developed read ers” (pp.143-144). This appears to be true when fluency instruction could support both the struggling and more developed reader’s, as was found in this study with the increased reading outcome s for all of the students of varying reading levels, using the interactive sing-to -read program, Tune Into Reading. Assessing proficient fluent reading for th is group of literacy le arners proves to be a difficult task, as little has been addressed about needs of early adolescent middle school reading fluency of varying read ing abilities. Therefore, asse ssments that are used with beginning readers (e.g., Dynamic Indicators of Basic Early Literacy Skills DIBELS; Good & Kaminski, 2002) or high-stake test scores (FCAT) are utilized to determine proficient fluent a dolescent readers.

PAGE 252

239 The ORF (oral reading fluency) and RF (re tell fluency) are assessments currently used in this middle school. These are subtests and a part of the DIBLES assessments used with older students. The concern with the ORF and RF tests according to Allington, (2006) is: During the ORF test the student is given one minute to orally read a passage, while the examiner counts the number of wo rds correctly read within the minute. During the RF test the student s reads orally for one minut e and then the student is asked to retell what he or she can recall from the passage. While the student is retelling the story, the teacher counts th e number of words uttered by the student (p.40). This might explain why the students in this study equated fluent reading with speed and not to comprehending text while decoding. As noted during the pretest assessment, 75% of the stude nts asked prior to reading, “How fast do you want me to read?” or “Do I need to read this fa st?” (Assessment Notes April 2, 2007). Although they were told to read at a pace that they would be able to understand and answer the comprehension questions, after reading the pa ssage, the students still read quickly with no expression and no pauses or stopping at punc tuation during their read ing at the pretest. Suggesting, for these students,’ their understandi ng of what it meant to be a fluent reader was equated to reading the words with speed, instead of reading for meaning. A reasonable conclusion reached is assessment th at does not take into consideration deep comprehension (internalizing material), but only surface comprehension (word level speed and accuracy), can prove to be problem atic for determining pr oficient reading of early adolescents of varying reading abilities.

PAGE 253

240 High-stakes testing scores are also used to determin e a fluent and proficient reader. The FCAT levels 15 in reading ar e used as indicators of reading ability and performance, according to the State of Florida. Levels 1 or 2 are considered below grade level; whereas, Levels 3 through 5 are consider ed at or above grade level in reading. The results of this yearly assessment can have a dramatic impact on the early adolescent literacy learner with the possi bility of retention, class placement, and specifically instructional practices provided to the students. The score obtained from this high-stakes test place early adolescents below, at, or a bove their classmates in reading, and it is assumed early adolescent students who may or may not have passed the test will receive instructional strategies needed to prepare th em to be fluent readers and comprehenders. The FCAT reading level scores were used in this study for two reasons: (a) to address the second research question of this study con cerning the comparison of the relationship with reading performance and FCAT levels, and (b) to approximate the appropriate beginning reading levels pr ior to the QRI-4 pretest assessment. The primary purpose of the FCAT in readi ng is to assess student achievement of higher-order thinking skills (Florida Depart ment of Education, 2005); therefore, it was assumed a student who attained higher FCAT level scores in reading (levels 3-5) would be at or above grade-level in reading. Ho wever, when the FCAT level reading scores were used to determine the benchmarks fo r administering the QRI-4 at pretest and posttest, there appeared to be a much lowe r than anticipated relationship between the FCAT level reading scores and scores obtained during the QRI-4 assessments. In particular, when the students were st ratified by FCAT reading levels 3-5 (as At or Above grade level in reading), it was found that FCAT reading level scores reported

PAGE 254

241 71% (40 out of 56) of the students were mee ting grade level or above in their reading. However, when these same students were given the QRI-4 reading assessments, the results demonstrated only 19% of the student s (11 out of 56) at pretest were on grade level or above in reading, and at posttest only 27% of the students (15 out of 56) were on grade level or above in reading. This sugge sted only 15 students out of the 40 students determined by their FCAT level reading sc ores were in fact meeting grade level proficiency, according to their QRI-4 scores in reading at posttest. These findings suggest the correlation assumed was not found between scores on the FCAT and scores from the QRI-4 used to determine proficient fluent readers. Therefore, it might be inferred that the use of a high-stake test scor es can not account for the many variables associated with understand ing the reading proce ss when relating that to the characteristics of this group of early adolescent literacy lear ners and their fluent reading behavior (McCombs, Kirby, Barne y, Darilek, & Magee, 2005; Rothstein, 2000). Amrein and Berliner (2002), overall, contend that “there is no compelling evidence from a set of states with high-stakes testing polices that those policies result in transfer to the broader domains of knowledge and skill for wh ich high-stakes test scores must be indicators” (p.54). Furthermore, the findings revealed that for the 56 students in this middle school, 27% were proficient readers, and 73% are reading below gr ade level, according to their QRI-4 scores. These results align with th e report from The Na tional Assessment of Educational Progress (NAEP) (2005) that re ports 73 % of eighth grade students perform below or at a basic level in their reading achievement.

PAGE 255

242 In addition, Rothstein (2000) questions whether an annual test of a students’ knowledge, at just one point in time, could provide an accurate assessment of fluent reading for this population of literacy learners. This was particularly true in this study, as the treatment students shifted fr om a fluent reader on one leve l to a surface fluent reader on a higher level. The findings reported sugge st treatment students were in a state of disequilibrium that mirrored Piaget’s Theory of Cognitive Development (1964). Based on Piaget’s theory (1964), as the students assimilated the higher level reading material and were building the schema for this new information, so they could accommodate it, they were in a state of dise quilibrium. Suggesting their fluency growth in reading was fluid and changing as each new cognitive task presented itself and required students to build the cognitive stam ina for the new more difficult reading tasks. This is interpreted as the use of an annual assessment to determine a fluent proficient reading for the early adolescent is problema tic because it does not allow for the ever changing state of fluent reading as found in this study. Toppings (2006) contends reading fluency is not “an entity, or a benchmarkable competence, or a static condition” (p.106). In addition he adds, “Even expert readers will show dsyfluency when confronted with an unfamiliar topic that provides challenge greatly beyond the students’ independent reading level” (Toppings, 2006, p.106). This appears to contradict some of the literatur e on reading fluency assessment, particularly when fluency is measured as a discrete ski ll (reading rate and word accuracy). However, it appeared to be a strategic proc ess for the students in this study.

PAGE 256

243 The Role of Prosody in Reading Fluency Reading prosody is the music and rhythm of oral language. Specifically, when a student demonstrates expressive oral read ing by using pace, volume, pitch, and rhythm, this is indicating behaviors of prosodic readi ng. However, there is not a consensus in the field concerning the role prosody plays in read ing fluency. The reading literature suggests fluent readers exhibit behaviors that ble nd reading accuracy, automaticity, and prosody (Samuels, 1979). Some scholars contend it is th e prosodic elements in reading that has a direct connection to reading comprehens ion (e.g., Allington, 2006; Rasinski, 2004); whereas, other scholars (e.g., Torgesen & Hudson, 2006) view reading prosody as not having any direct relationship to comprehe nding text. Instead, they suggest decoding (word accuracy) with automa ticity (reading rate) are the direct connection to comprehension. While there is no debate amongs t the reading community as to the need for fluent readers to be efficient decoders, in order to comprehend text, the stance taken that word level reading with speed alone improves comprehension can be problematic for the fluent middle school decoders. Biancar osa and Snow (2006) contend most early adolescents do not have difficu lty reading fluently at the word level; instead, the difficulty arises with th eir reading comprehension. The Qualitative Reading Inventory-4 (QRI-4) (Leslie & Caldwell, 2006) was used in this study as a pretes t and posttest measure to dete rmine the students’ fluency, word recognition, comprehensi on, and instructional reading levels for the groups. An instructional reading level is calculated by using the combination of a score in word accuracy and reading comprehension. Therefore, to determine the instructional reading level for the students in this study, it combined their word accuracy scores with their

PAGE 257

244 comprehension scores. As previously noted by Biancarosa and Snow, the students in this study appeared not to have difficulty with r eading the words; the difficulty arose with comprehending what they read. The word accuracy (WR) scores reporte d showed no statistically significant difference between the groups from pretest to posttest. In addition, the WR scores indicated both groups were at independent le vel in how accurately they could read the words in the text. In fact, when the student s were group by FCAT Levels for the students At and Above grade level in reading, they had reached a “ceiling effect” (Stevens, 2002). This suggested that for these students WR had gone as high as it coul d go at this level. However, because their comprehension scores were not as high, the students could not be moved to a higher instru ctional reading level. In additi on, the reading rate WPM for both groups met an acceptable criterion (140 WP M) for the students of this age group (Rasinski, 2004). This might suggest the vari ables of word accuracy in reading (WR) and the reading fluency rate of speed (WPM) may not be contributing factors for early adolescents when thinking about important components for fluent reading leading to comprehension. Based on The Automaticity Theory, LaBerg e and Samuels (1974) define fluent reading as the ability to decode and compre hend text at the same time. Their theory suggests cognition has only a limited capaci ty to process information. Therefore, decoding (at the word level) can become au tomatic, and the focus cognitively can be on the complex process of comprehending text. Through guided and repeated reading, both decoding (automaticity and accuracy) in word recognition and comprehension are developed.

PAGE 258

245 Samuels (1979) further defines repeated reading as a fluency-building strategy that consists of timed rer eading of a short passage seve ral times (at least 3 times), checking for accuracy ( word recognition), automaticity (words per minute) and with prosody (expression). Furthermore, the step s recommended for an effective fluency instructional model are: (a) to provide a mode l for student’s expressi ve fluent reading, (b) to give the students a passage to read (approximately 150 words) 3 times, and (c) to have the students orally read the passage asse ssing for accuracy, automaticity, and expression (Rasinski, 2004). Repeated reading is most authentic when the practiced material is eventually performed orally, such as plays, poetry recita tion, or in this study singing lyrics to songs (Rasinski, 2004; Stayter & Allington, 1991). This form of repeated exposure through singing assists the reader with fluency th rough prosodic reading. The singing performed by the students appears to exaggerate the la nguage of reading, as the students find their voice in the rhythm and the bounce of the musi c. The reader uses appropriate volume, rhythm, pitch, tone, and phrasing (prosody), wh ile singing the song lyrics; therefore, they give evidence of actively constructing meaning from th e passage (Rasinski, 2004). The findings of this study concur with Rasinski (2004), in part. Distinctively, prosody when rereading through singing appeared to have a direct connection to reading comprehension (COMP) and increasing the inst ructional reading leve l (RL). However, the practice of rereading, through singing by following the protocol recommended in the literature alone, did not produce the same findings as what ha s been previously reported (Samuels, 1979). If that were indeed the case, then both groups should have increased in their reading comprehension and instruc tional reading level because both groups

PAGE 259

246 followed the recommended procedures for fluency instruction. Nevertheless, the treatment students significantly outper formed the control group in reading comprehension (COMP) and instructional re ading level (RL). This suggests repeated practice of rereading through singing by the co ntrol group of varying reading levels did not improve their reading comprehensi on or instructional reading levels. In addition, the treatment group appeared to interact with the prosodic elements of text, rather than just being passively immersed in the prosodic elem ents through repeated practice as noted in the control group. Speci fically, the treatment group applied reading strategies to comprehend the pr osody of the songs which result ed in an increase in their reading comprehension over the control group. As noted in the excerpt from the data: He was focused on rereading the song. This was the fourth time he recorded himself singing. He had the screen up that displayed his vocal tract. He used his finger to align where he was off pitch. He went back again and reread the song again. [He moved back and forth in his seat, nodding his head to the beat, and tapping his foot to the music] Finally, [he sub vocalized] I got a 90! (Week 4, April 25, 2007). The student (representative of his peers) was being strategic and metacognitve as he interacted with the text. He was interactiv e with the text as he applied strategies for effective comprehension of text. However, he appeared to be inte racting and applying strategic processes to the prosodic elements of the text. As noted, he would trace his

PAGE 260

247 finger on the vocal tract line and then reread and re-sing the song as his body moved to the rhythm and beat of the music. The above example exemplifies an interaction with the prosodic elements of the text. Therefore, based on The Automaticity Theory (LaBerge & Samuels, 1974), it appeared th at not only was the student decoding automatically, the strategic processes appeared to focus on comprehending the prosody of the text. This suggested that for reading comp rehension to increase and see transfer effects for reading comprehension to other reading material (e.g., QRI-4 assessments), these middle school students needed to be interactive and co mprehend the song lyrics and the prosodic elements of the text they were reading through singing. The individual interactive sing-to-read program, Tune Into Reading, was used individually, and as noted in the above ex cerpt, the student mani pulated the text to understand the song and its prosodic elements. He used various readi ng strategies to see where he could improve, as he traced the voice frequency lines, reread the song and then recorded himself again until he reached his goal. The practicality of this alternative format assisted him in comprehending text, unlike a linear text that can not be stopped, started, or slowed down. In addition, the continuous background music assisted him because he did not have to use his cognitive cap acity to remember the rhythm or beat of the song that was being automatically supplied. Therefore, he could focus on comprehending the prosodic elements while being guided automatically by the background music. In addition to the signifi cant increase in reading co mprehension (COMP) scores, the treatment group significantly outperformed the control gr oup in their instructional reading level (RL). As previous ly noted, instructional reading level is calculated by using

PAGE 261

248 the combination of a score in word accur acy and a score in reading comprehension. Mariotti and Homan (2005) s uggest that to determine the percent correct for word recognition, the teacher counts th e errors and subtracts it from the total number of words in the passage, then divides by the total number of words contai ned in the passage (p.76). The formula is noted as: total number of words in the passage-errors = word recognition percent correct total number of words in the passage To determine the comprehension percent correct the teacher subtracts the errors from the total number of questions, and then divides th at number by the tota l number of questions. The formula is noted as: total number of questions -errors =comprehension percent correct. total number of questions Once this is accomplished, Mariotti and Ho man (2005) suggest a criterion is used to indicate instructional r eading levels of the stud ents. Two well known scholars developed criteria for determining instructi onal reading levels, Betts (1946) and Powell (1971). Betts criteria suggest that there is a st andard baseline of scores across grades that can be interpreted descriptivel y incorporating the pr osodic elements of oral text reading. Powell criteria adjust the baseline in word recognition and comprehension for passage difficulty by passage reading le vels (Mariotti & Homan, 2005). When looking at the criteria separately, it does not appear to totally address the needs of interpreting instruct ional reading level for these early adolescents of varying reading ability. However, possibly combining the criteria might address the elements

PAGE 262

249 necessary to capture the behavi ors for instructional reading levels for early adolescent readers. Specifically, this could be accomplis hed by using Betts criter ia that descriptively captures the prosodic elements and Powell’s crit eria that adjusts the baseline for word recognition. Therefore, it would address wh at was found in this study; the prosodic elements in reading played a significant role in increasing instructional reading level and comprehension for the treatment group. In addition, I concur with Mariotti and Ho man when they state the most important function of and Informal Reading Inventory (I RI) is qualitative descriptive interpretations of behaviors in reading, along w ith the quantitative criteria th at need to be taken into consideration when determini ng instructional reading levels These behaviors such as pausing at sentence, self-cor recting, using tone, and other prosodic elements were found as indicators of comprehending te xt. In particular, as found in this study at pretest during the reading assessments, both groups of read ers read their assessment passages orally with speed and a high level of accuracy in word recognition, yet they struggled with comprehension during the pretest assessment. Their oral reading was absent of volume, tone, pitch or any expression. There was no pausing at punctuation, rereading for clarification, or self -corrections made in 53 out of 64 students or 83% of the groups. However, at posttest, the treatment group of students outperformed their counterparts significantly from pretest to pos ttest in reading comprehension COMP and instructional reading level RL. The oral r eading of the students in the treatment group, although fast (180 wpm), had expression, pitc h, and volume, unlike their counterparts. Specifically, 81% of the treatment students or 26 out of 32 read their passage making self-correction, pausing at punc tuation, and rereading phrases or sentences. Whereas, in

PAGE 263

250 the control group of students, only 28% or 9 out of 32 of these students incorporated these prosodic elements in th eir reading. This suggests pr osodic elements of reading appear to have a direct conn ection to reading comprehension. Sociocultural Interactions Vygotsky (1978) contributed to the con ception that liter acy is a social construction, specifically, vi ewing cognition as a profound so cial phenomenon. Initially, learning is socially construc ted, and then as the higher mental processes take shape, learning becomes internalized. If this perspective is embraced, it could be interpreted as social experiences through soci ocultural interactions shape thinking and interpretations of the world. The treatment and control groups had a so cial system in place that positioned some of its peer members in the role of dominance over other peers. These dominant peers took this position, and the other peer members allowed them to assume it. Both groups appeared to have this two class syst em, where a small group or a few individuals lead the rest of the group, determining what was considered acceptable social behavior. Interestingly, however, within the groups the interactions from the dominant peers with the other peer members was accomplished very differently. A small group of eighth grade males were the dominant peers within the treatment group. They modeled behaviors or talk that resu lted in the rest of the peers imitating their behaviors. Their talk was supportive and coll aborative with one a nother; however, there were no discussions with other peer memb ers only amongst this small group. This is consistent with Ryan’s (2000) definition; modeling is a form of adolescent peer interaction. This interaction refers to indivi dual changes in cognition, beliefs, or affect,

PAGE 264

251 which are a result of adolescen ts observing their peers. Obse rving a specific behavior a peer performs or listening to a peer voice, a certain belief can indu ce an adolescent to change their stance or adopt their peers’ behaviors or beliefs. Schunk and Zimmerman (1996) reported peer modeling influenced se lf-efficacy beliefs, as was found in this study. The students, after observing the dominant peer s in the treatment group, initially used the sing-to-read program, Tune Into Reading, and continued its use while showing support and cooperation with one another as modeled by these dominant peers. Whereas within the cont rol group, the dominant peers were male or female individuals and their in teractions were direct discussi on with the other peer members. The dominant peers directed the other member s to conform to certain behaviors they deemed socially acceptable. This might have been a result of how these dominant peers interpreted what was expected of them through the instruct ional expectations provided by the music teacher’s modeling. Information ex change occurs when adolescents have a discussion with their peers (B erndt, 1999). This form of in teraction could influence the early adolescent’s choice to partake in the lit eracy task presented by the teacher if it was used effectively. However, Ryan (2000) contends it also has an adverse influence if the peers use this form of interactions to control other peers to conform to socially acceptable behavior. As noted in the findings, the dominant peers w ithin the control group directed and tried to intimidate and control the more vulnerable peers into conf orming to the instructional expectations. This appeared to have an a dverse effect on the other peer members. As noted in the findings, the other vulnerable peer s did not come to aid of the peer that was being admonished; instead, they would look at one another or look away.

PAGE 265

252 Peer pressure can also take on the role of social reinforcement (Ryan, 2000). Brown, Lohr, and Eicher (1986) found beliefs and behaviors that are discouraged by the groups are not likely to be displayed; whereas beliefs and behaviors positively received by the group are more likely to surface. Participation in the literacy tasks involving the treatment peer group positively modeled through the dominant peer interactions had a positive effect on the group’s beliefs and d ecisions to participate by all of the group members. Whereas, within the control group se tting, what appeared to happen was that peer pressure was applied by the dominant peers, and it was not positively received. Therefore, they disengaged from the task while trying to escape the pressure. These findings suggest the role of the dominant pe ers and sociocultural interactions have a significant influence in the reading perfor mances of the group. Specifically, it was found the treatment group showed a significant incr ease in all areas of reading fluency; however, the control group di d not. This might be interpreted as the sociocultural interactions modeled through the dominant peers in the treatment group of support and collaboration was positively interpreted by th eir peers, and the results were higher in performance of read ing within the group. Contributions of the Study Although previous research has identified characteristics of effective reading fluency instruction, the focus has been on be ginning readers or older struggling readers. This focus has involved interpreting fluent reading as having a connection to reading comprehension at the word level (LaBerge & Samuels, 1974). As found in this study, that was not the case for middle school readers in the treatment group. Inst ead, it appeared to be the prosodic elements of text that had th e direct connect to r eading comprehension. As

PAGE 266

253 future fluency studies at the middle school le vel are completed, the results from this study will provide additional information to the role prosody plays for fluent readers of varying reading ability in the middle school. In addition, there have been very few studies conducted on middle school readers of varying reading ability, following the pr otocol for effective fluency instruction. Furthermore, few studies have been conducted that embed literacy stra tegies naturally in a content class, while looking at transfer effects of comprehension to other reading material. In this study, the literacy task of rereading through singing maintained the integrity of the subject matter of the music class, while embedding the literacy elements for fluency instruction. The results reveal ed rereading through singing for treatment students transferred to a read ing assessment, showing a sign ificant increase in reading comprehension. These results contribute to th e concept of embedded literacy instruction and transferability of reading perf ormance for other reading tasks. Finally, very few studies have integrated the role of peer interactions during a specific literacy task, while m easuring their reading performa nce in reading. The findings revealed for both treatment and control group a two class system within the classroom settings. These social systems were led by the dominant peers, and the rest of the members allowed this and followed their lead. However, how the dominant peers interacted influenced the other peer member s. The results suggested these interactions determined how all the peers responded to the literacy task. These findings help to clarify the role that peer in teractions might have in the middle school.

PAGE 267

254 Recommendation for Practice Fluency Instruction As noted in the findings, the role of pr osody appears to have a direct connection to reading comprehension for treatment student s. However, the instructional emphasis of expressive reading tends to decrease fo r students once they leave primary grade (Allington, 2006). Therefore, there is a need to incorporate models and practice of prosodic elements of reading text for students of all levels, especially within the middle school environment. This can be accomplished by having teachers ha ve more read alouds and provide students with more opportunities to orally read so they are able to practice the prosodic elements of text. In addition, fluency instruction needs to be differentiated to meet the developing needs of these students. As noted in the fi ndings, the treatment group using the interactive singing software, Tune Into Reading, when grouped by FCAT Levels was effective in meeting the differentiated needs for each level. Additionally, when the treatment students reached a higher level in their reading, their fl uency decreased. This suggests fluency in reading is not stagnant; it is instead fluid and ever changing with the different tasks middle school students face (Topping, 2006). Sugge sting, fluency is a strategic process rather than a skill. As well, the expectations that students in the middle school enter the context of the school environment as fluent r eaders should be revaluated, as this was not the case with this group of early adolescents Furthermore, opportunities for individual practice, rather than a whole group one-size-fits-all model, should be considered. It was found that studen ts in the treatment group made a cognitive shift from assimilating the reading informati on to interacting and

PAGE 268

255 internalizing their learning. This, in part, ap peared to be because they had opportunities for individual practice. In addition, the alte rnative text students us ed in the treatment added to their comprehension improvement. Both groups found the altera tive approaches to learni ng motivating and engaging. This was apparent during the interacti ons within the groups. The light-hearted competitions through the peer interactions we re documented in the data. The groups led by the dominant peers would, either through di scussion or modeling, set the climate of motivation for rest of the group. However, as the intervention continued, a shift occurred within both of the groups as to their motivation for these al ternative approaches to the task of rereading through singing, as did the role of the dominant peers. For the alternative textual format, the Tune Into Reading program that the treatment group used was not only motivati ng and engaging but eas ily manipulated. The students could adjust the program, and this ap peared to assist them in comprehending the prosodic elements of the text. In addition, th e perception of what alternative was changed during the course of the inte raction. The treatment students appeared to perceive this musical textual format as one that assisted th en in their learning. This suggests defining and using alternative textual formats should incl ude, not only the delivery of the text and the genre it provides, but the perceptions it develops. This perception changed from a fun game-like alternative text to a text the student could use to comprehend the reading material.

PAGE 269

256 Assessing Reading Fluency Since it appeared prosody had a direct connection to reading comprehension, assessments should assess the students with this element in mind. I concur totally with Mariotti and Homan (2005) when they suggest qualitative descriptive interpretations of behaviors in reading, along w ith the quantitative criteria need to be taken into consideration when determining instructional reading levels a nd fluent reading behaviors. These behaviors such as paus ing at sentence, self-corr ecting, using tone, and other prosodic elements were found as indicator s of comprehending text in this study As well, the measurement tools curren tly being used (e.g., ORF and RT) appear detrimental to interpreting reading fluency fo r these students, as the students and their teachers are interpreting fluency reading at a surface word level. Furthermore, using the FCAT reading levels scores appears probl ematic, as the use of these scores as benchmarks did not correlate to the QRI-4 read ing levels. Therefore, not all the students that might benefit from furthe r instruction in their reading fluency are actually getting instruction they need. In addition, the findings revealed all the stude nts in the treatment group benefited from fluency instruction. Howe ver, when using these scores within the school setting, only those students suggested by these scores are receiving fluency instruction, when all could benefit. Sociocultural Interactions and Influences on Instruction The role of dominant peers remained c onstant within the treatment and control groups, during the intervention. Nonetheless, in structional expectations afforded to the groups by the music teacher were very different within each setti ng. These expectations

PAGE 270

257 appeared to be interpreted by the dominate peers and then reinforced through their interactions with the re st of the peer group. What appeared to be expected of the p eers in the control gr oup was only one right way to perform the singing and drumming. These expectations were modeled to the peers as the music teacher would stop the singi ng and drumming several times daily during each of the sessions and tell the students some people were off key, and to try it again. In turn, the findings revealed the dominant peer would admonish the vulnerable peers for making a mistake. Neverthele ss, the treatment group was exp ected to try their best. The music teacher would often remind the students to try their best and not to worry about their pitch scores. The dominant males in the treatment group would encourage and support one another, modeling collaboration to the other peers. Therefore, it appeared that within both groups, the dominant peers in terpreted what was expected of them, and then in turn, reacted to these expe ctations through their interactions. The control group peers were expected to perform correctly and accurately as modeled by the music teacher. Since there wa s only the right or wrong way to perform the song and drum sequence, the interacti ons might have been interpreted by the dominant peers and conveyed to the rest of the group in this manner through these interactions. However, the treatment gr oup was allowed differentiation through the instructional delivery. Therefore, the dominant peers appeared not to be compelled to take on the task of reinforcing group accu racy; instead, they became a group member while still maintaining thei r position among the group. This suggests instruction should meet the needs of the individual students and individual accomplishments should be rewarded.

PAGE 271

258 Recommendations for Future Research The findings of this study reported the ro le of prosody appears to have a direct connection to reading comprehension. Conve rsely, the fact that there were only 64 students and the individual students of this study were not each randomly assigned to a treatment or control condition limits the generalizablity of the findings to this group of students. In addition, the dur ation of the study was seven weeks in length. Future researchers may consider increasing the sa mple size and lengthening the study period to obtain additional data for reading fluency. Additionally, as to random se lection for this population, th e sample characteristics were predominantly White eighth grade lo w SES males. There were no sixth grade students, a limited number of seventh graders, or students that required additional support in their learning. Future research should investigate a greater diversity in the classification characteristics of the students of this study. The content class was a music class, and the strategies taught were appropriate for this content area. The findings suggested th ere was a transfer effect from the embedded literacy taught to another literacy task. Fu ture research might investigate embedded literacy to see if this tr ansfer effect holds between other content classes. Finally, the mixed method design of this study was effective in capturing the reading performances and the descriptive findings. However, the case study used observational field notes only to capture th e peer interactions but did not include videotapes or tape recordings of these interactions. Futu re researchers might want to utilize these in their research designs fo r the purposes of capturing more in-depth

PAGE 272

259 understanding of peer interac tions and how this relates to the literacy task that the students are involved

PAGE 273

260 References Alfassi, M. (2004). Reading to learn: Effect s of combined strategy instruction on high school students. Journal of Educational Research, 97, 171–84. Alexander, P. A. (1998). The nature of di sciplinary and domain learning: The knowledge, interest, and strategic dimens ions of learning from subj ect-matter text. In C. Hynd (Ed.), Learning from text across conceptual domains (pp. 263-287). Mahwah: Erlbaum. Alexander, W.M. (1968). A survey of organizational patterns of reorganized middle schools. (final report number 7-0-026). Gain sville, FL: University of Florida. Alexander, W.M., & George, P.S. (1981). The exemplary middle school. New York: Holt, Rinehart, and Winston. Alexander, P. A., & Jetton, T. L. (2000). Learni ng from traditional and alternative texts: New conceptualization for an information age. In A. Graesser, M. Gernsbacher, & S. Goldman (Eds.), Handbook of discourse processes Mahwah: Erlbaum Allington, R. L. (2006). Fluency: Still waiting af ter all these years. In S.J. Samuels, & A. E. Farstrup (Eds.), What Research Has to Say About Fluency Instruction. Delaware: International Reading Association. Allington, R. L. (2001). What really matters for strugg ling readers: Designing researchbased programs New York: Addison-Wesley.

PAGE 274

261 Alvermann, D. E. (2002). Effective l iteracy instructi on for adolescents. Journal Literacy Research, 34, 189-208 Alvermann, D.E. (2003 ). Seeing themselves as capable and engaged readers: Adolescents and re/mediated instruction Learning Point Associates. North Central Regional Laboratory, Institute of E ducation Sciences, U.S. Department of Education, contract number ED-01-CO-0011. Alvermann, D.E., & Moore, D. W. (1991). S econdary school reading. In R. Barr, M.L. Kamil, P. Mosenthal, & P.D. Pearson (Eds.), Handbook of Reading Research 2, 951-983. New York: Longman. Alvermann, D.E., & Phelps, S. P. (2005). Content reading and literacy: Succeeding in today’s diverse classroom (4th Ed.). Boston: Pearson Education. Amrein, A.L., & Berliner, D.C. (2002). Highstakes testing, uncertainty, and student learning. Education Policy Analysis Archives 10(18). Retrieved (11/11/2006) from http://epaa.asu.edu/epaa/v10n18/ Ary, D., Jacobs. L.C., & Razavieh, A. (1996). Introduction to research in education (5th ed.). Fort Worth: Harcourt Brace College Publishers. Bandura, A. (1997). Self-Efficacy: The exercise of control. New York: Freeman. Barr, A.S. & Glifford, C. W. (1929). The vocabulary of American history. Journal of Educational Research, 20, 103-121. Beane, J., & Brodhagen, B. (1996). Doing Curriculum Integration Madison: University of Wisconsin.

PAGE 275

262 Bean, T.W. (2000). Reading in the content ar eas: Social constructi vist dimensions. In M.L. Kamil, P.B. Mosenthal, P.D. Pearson, & R. Barr (Eds.), Handbook of reading research (Vol. 3, pp. 629-644). Mahwah: Erlbaum. Becker, H. J. (1987). Addressing the needs of different groups of early adolescents: Effects of varying school and classroom organizational practices on different social background and abilities. (Report No. 16). Mary land: VSP Industries. Berndt, T. J. (1999). Friends’ influen ce on students’ adjustment to school. Educational Psychologist, 34 15-29. Berndt, T. J., Laychak, A. E., & Park, K. (1990). Friends’ influence on adolescents’ academic achievement motivation: An experimental study. Journal of Educational Psychology, 82 664-670. Biancarosa, C., & Snow, C. E. (2006). Readi ng next—A vision for action and research in middle and high school literacy: A Report to Carnegie Corporation of New York (2nd ed.).Washington, DC: Alliance for Excellent Education. Biggs, M.C., Homan, S.P., Dedric k, R., Minick, V., & Rasinski T. A. (in press). Using and interactive singing program: A compar ative study of middle school struggling readers. Reading Psychology An International Quarterly. Binet, A. (1904). New investigations upon th e measure of the intellectual level among school children. In The Development of Intelligence in Children. Baltimore: Williams and Wilkins. Bogdan, R. & Biklen, S. K. (2003). Qualitative research for education: An introduction to theories and methods (4th Ed.). Boston: Pearson Education Group.

PAGE 276

263 Bond, G. L. & Bond, E. (1941). Developmental reading in high school. New York: MacMillian. Bough, M. (1983). Middle school curriculumdo es it meet the need of the transescent child? Contemporary Education, 54, 272-274. Brimm, R.P. (1963). The junior high school. Washington, D.C.: The Center for Applies Research in Education, Inc. Broudy, E. (1975). The trouble with textbooks. Teachers College Record, 77, 13-34. Brooks, K., & Edwards, F. (1978). The middle sc hool in transition: A research report on the status of the middle school moveme nt. Lexington: College of Education University of Kentucky. Brown, B. B. (1990). Peer groups and peer cultu re. In S.S. Feldman & G.R. Elliot (Eds.). At the threshold: The developing adolescent (pp.171-196). Cambridge: Harvard University Press. Brown, B. B., Lohr, M.J., & McClenahan, E.L. (1986). Early adolesce nts, perceptions of peer pressure. Journal of Early Adolescence, 6, 139-154. Bulgren, J. A., Schumaker, J. B., Deshler, D. D., Lenz, B.K., & Marquis, J. (2002).The use and effectiveness of a comparison routine in diverse secondary content classrooms. Journal of Educational Psychology, 94 356–71. Bushwell, G. T., & John, L. (1931). The vocabulary of arithmetic. (Supplemental Educational Monographs No. 38). Chica go: University of Chicago Press. Butzlaff, R. (2000). Can music be used to teach reading. Journal of Aesthetic Education, 34(3), 167-178.

PAGE 277

264 Callahan, R. (1962). Education and the cult of efficiency. Chicago: University of Chicago Press. Calkins, L. M. (2001). The art of teaching reading New York: Addison-Wesley. Carnegie Council on Adoles cent Development (1989). Turning Points: Preparing American Youth for the 21st Century New York: Carnegie Corporation of New York. Chall, J.S. (1996). Stages of reading development (2nd Ed.) Fort Worth: Harcourt Brace. Clay, M. (1985). The early detections of reading difficulties Portsmouth: Heinemann. Cody, R. P. & Smith, J. K. (2006). Applied statistics and the SAS programming language. (5th Ed.). Upper Saddle River, New Jersey: Pearson. Conley, M. W. (1988). Knowledge about pedagogical content: How practicing teachers come to understand content reading instruction. Paper presented at the annual meeting of the National Read ing Conference, Tucson, AZ. Cook-Gumperez, J. (1986). Intr oduction: The social construction of literacy. In J. CookGumperez (Ed.), The social construction of literacy, (pp. 1-15). Cambridge, England: Cambridge University Press. Coser, L.A., Kadushin, C., & Powell, W.W. (1982). Books: The culture and commerce of publishing. New York: Basic Books. Cottle, T. J. (2001). Mind fields-Adolescents consciousne ss in a culture of distraction New York: Peter Lang. Creswell, J. W. (2003). Research design qualitative, quantitative, and mixed methods approaches. (2nd Ed.) Thousand Oaks: Sage Publications.

PAGE 278

265 Creswell, J. W., Plano Clark, V. L., Guttm an, M., Hanson, W. (2003). Advanced mixed methods research design. In A. Tashahhori & C. Teddlie (Eds.), Handbook on mixed methods in the behavioral and social sciences, (pp. 209-240) Thousand Oaks: Sage Publications. Crocker, L. & Algina, J. (1986). Introduction to classical an d modern test theory. New York: Holt, Rinehart and Winston. Cronbach, L. J. (1951). Coefficient alpha and the internal st ructure of tests. Psychometrika, 16, 297-334. Cross, D., & Paris, S. (1987). Assessment of reading comprehension: Matching test purposes and test properties. Educational Psychologist, 22, 313-332. Cuban, L. (1992). What happens to reforms that last? The case of the junior high school. American Educational Research Journal 29, 227–251. Curtis, F. D. (1938). Investigation of vocabulary in textbooks of science for secondary schools. Boston: Ginn. Deci, E.L. & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human behavior New York: Plenum. deLeon, A. G. (2002). Moving Beyond Storybooks: Teaching Our Children to Read to Learn. Carnegie Reporter Fall, Vol. 2, No. 1. Dewey, J. (1910). How we think. Boston: D.C. Heath. Dillon, J. T. (1983).Cognitive complexity and duration of classroom speech. Instructional Science, 12, 59-66. Dolch, E. W. (1928). Combined word studies. Journal of Educational Research, 17, 11-19.

PAGE 279

266 Dolan, T., Harrison, C., Gardner, K. (1979). Th e incidence and contex t of reading in the classroom. In E. Lunzer & K. Gardner (Eds.). The effective use of reading, (pp. 108-138). London : Heinemann. Dole, J.A., Duffy, G.G., Roehler, L.R., & P earson, P.D. (1991). Moving from old to new: Research on reading comp rehension instruction. Review of Educational Research, 61 239-264. Donahue, P.L., Voelkl, J.R., Cam pbell, J.R., & Mazzeo, J. (1999). NAEP 1998 reading report card for the nation and states. Washington, DC: Department of Education Dowhower, D. L. (1987). Effects of repe ated reading on second grade transitional readers’ fluency and comprehension. Reading Research Quarterly 22, 389-406. Dowhower, D. L. (1991). Speaking of pr osody: Fluency’s unattended bedfellow. Theory into Practice 30, 158-164. Draper, R., Smith, L. K. Smith, Hall, K. M., & Siebert, D. (2005). What’s more important-literacy or content? Conf ronting the literacy-content dualism. Action in Teacher Education, 27(2), 12-21. Dunlap, W. P., Cortina, J. M., Vaslow, J. B ., & Burke, M. J. (1996). Meta-analysis of experiments with matched groups or repeated measures designs. Psychological Methods 1, 170-177. Durkin, D. (1978-1979). What cl assroom instruction reveal about reading comprehension instruction. Reading Research Quarterly, 15, 481-533. Eccles, J. S., Lord, S., & Midgley, C. (1991) What are we doing to adolescents? The impact of stage–environment fit on a dolescents’ experiences in schools and families. American Journal of Education. 99, 90-101.

PAGE 280

267 Eccles, J. S., & Wigfield, A. (2002). Motivational beliefs, values, goals. Annual Review of Psychology. 53 109-132. Eichhorn, D.H. (1973). The Boyce medical study. Educational dimensions of emerging adolescent learner. Springfield: Educational Leadership Institute, Inc. Eichhorn, D.H. (1966). The middle school. New York, NY: Center for Applied Research. Electronic Learning Systems. (2006). Carry -a-Tune Learn to Sing Program. Tampa, Florida. Elish-Piper, L., & Tatum, A. W. (2006). A ddressing the literacy needs of adolescent students: Listening to their voice. The NERA Journal, 42(1), 1-12. Englert, C. S., & Palincsar, A. S. (1991). Rec onsidering instructional research in literacy from a sociocultural perspective. Learning Disabilities Research & Practice, 6, 225-29. Fisher, D., & Ivey, G. (2005). Literacy and language as learning in the content-area classes: A departure from “Every Teacher a Teacher of Reading”. Action in Teacher Education, 27(2). San Diego, CA. Flesch, R. F. (1940). A new readability yardstick. Journal of Applied Psychology, 32, 221-233 French, D. (2003). A new vision of authentic assessment to overcome the flaws in high stakes testing. Middle School Journal, 35, 1-12. Fry, E. (1977). Fry’s Readability Gra ph: Validity, and Extension to Level 17. Journal of Reading 21, 249.

PAGE 281

268 Gates, A.L. (1935). The improvement of reading: A program of diagnostic and remedial methods. New York: MacMillian. Gillet, J., Temple, C. & Crawford, A. (2004). Understanding reading problems (6th Ed.). Boston: Pearson. Glass, G.V., & Hopkins, K. D. (1996). Statistical Methods in Education and Psychology. (3rd Ed.). Boston: Allyn and Bacon. Glaser, B.G. & Strauss, A.L. (1967). The discovery of grounded theory. Chicago: Aldine Publishing Company. Goetze, M., Cooper, N., & Brown, C. (1990). Recent research on singing in the general music classroom. Bulletin of the Council for Re search in Music Education, 104, 16-37. Good, C.V. (1927). The supplemental reading assignment. Baltimore: Warwick & York. Gordon E.E. (2003). Learning sequences in music: Skill, Content, and Patterns. Chicago: GIA Publications. Gray, W.S. (1919). The relationship between study and reading (Proceedings of the Annual Meeting of the National Educ ation Association). Washington, DC: National Education Association. Gray, W.S. (1937). The nature a nd organization of basic instru ction in reading. In G. M. Whipple (Ed), The teaching of reading. (Thirty-sixth Yearb ook of the National Society for the Study of Education Pa rt I). Bloomington: Public Schools Publication Co.

PAGE 282

269 Gray, W.S. (1939). Reading. In G. M. Whipple (Ed), Child development and curriculum. (Thirty-eighth Yearbook of the National Society for the Study of Education Part I). Bloomington: Public Schools Publication Co. Green, J.C., Caracelli, V. J., & Graham, W. F. (1989). Toward a conceptual framework for mixed methods evaluation designs. Educational Evaluation and Policy Analysis, 11(3), 255-274. Greenberg, M.T., Siegel, J.M., & Leitch, C.J. (1982). The nature and importance of attachment relationships to pare nts and peers during adolescence. Journal of Youth and Adolescence 12(5), 373-386. Guthrie, J.T., & Wigfield, A. (2000). Enga gement and motivation in reading. In M. Kamil, P.B. Monsental, P.D. Pearson & R. Barr (Eds.) Handbook of Reading Research, 3, 403-422. Guthrie, J. T., & Davis, M. H. (2003). Mo tivating struggling read ers in middle school through an engagement mode of classroom practice. Reading and Writing Quarterly, 19, 59-84. Hall, G. S. (1908). Social significance of adolescence. Journal of Psychology, 19, 149-212. Hall, L., Boone, N., Grashel, J. & Watkins, R., Eds. (1997). Strategies for Teaching: Guide for Music Methods Classes. Reston: Music Educator s National Conference. Havighurst, R. J. (1972). Developmental tasks and education. New York: Davis McKay Co., Inc. Herber, H. L. (1970). Teaching reading in the content area. New Jersey: Prentice-Hall.

PAGE 283

270 Hodgkinson, H. (1986). The right school for the right kids. Educational Leadership. 45(5), 10-14. Huey, E. B. ((1908). The psychology and pedagogy of reading. New York: MacMillian. Inhelder, B. & Piaget, J. (1958). The growth of logical th inking from childhood to adolescents. New York: Basic Books. Ivey, G. (1999). A multicase study in th e middle school: Complexities among young adolescent readers. Reading Research Quarterly, 34 172-192. Ivey, G., & Broaddus, K. (2001). “Just plain reading”: A survey of what makes students want to read in middle school classrooms. Reading Research Quarterly, 36(4), 350-377. James, W. (1923). Talks to teachers on psychology: And to students on some life’s ideals. New York: Henry Holt. Johnson, H.J. (1984). A synthesis of research fi ndings on middle level education. In J. H. Lounsbury, (Eds.) Perspectives : middle school education, 1964-1984. OH: National Middle School Association. Johnson, B., & Turner, L.A. (2003). Data co llection strategies in mixed methods research. In A. Tashahhori & C. Teddlie (Eds.), Handbook on mixed methods in the behavioral and social sciences ,( pp.297-320) Thousand Oaks: Sage Publications. Johnston, J.H., Markle, G. C., & Stingley, L. (1982). What research says to the practitioner about peer rela tions in the classroom. Middle School Journal, 13(3), 22-26.

PAGE 284

271 Judd, C.H. & Buswell G. T. (1922). Silent reading a study of various types (Supplemental Education Monographs, No. 23). Chicago: University of Chicago Press. Juvonen, J., Vi-Nhuan, L., Kaganoff, T., Augustin e, C., & Constant, L. (2004). Focus on the wonder years: Challenges facing the American middle school. A Report to the Rand Corporation. Santa Monica, CA. Kamil, M. (2002). Adolescence and literacy: Reading for the 21st century A Summary of the Research: National Assessment of Education progress. Kilpatrick, W. H. (1919). The project method. New York: Teachers College Press. Kottmeyer, W. (1944). Improving reading instructions in St. Louis schools. Elementary School Journal, 45, 33-41. Kuhn, D., Black, J., Keselman, A., & Kaplan, D. (2000). The development of cognitive skills to support inquiry learning. Cognition & Instruction, 18, 495-523. Kuhn, M.R., & Stahl, S. A. (2003) Fluency: A review of development and remedial practices. Ann Arbor: Center for the Im provement of Early Reading Achievement. LaBerge, D., & Samuels, S.J. (1974). Toward a theory of automatic information processing in reading. Cognitive Psychology, 6 293-323. Lamb, S.J., & Gregory, A.H. (1993). The rela tionship between music and reading in beginning readers. Educational Psychology 13 19-27. Lee, C. D. (2001). Is October Brown Chin ese? A cultural modeling activity system for underachieving students. American Educational Research Journal, 38, 97-141.

PAGE 285

272 Lenz, B. K., & Deshler, D. D. (2003). Teaching content to all: Evidence-based inclusive practices in middle and secondary schools Boston, MA: Allyn and Bacon. Leslie, L., & Caldwell, J. (2006 ). Qualitative Reading Inventory-4. Boston, MA.: Allyn & Bacon. Leu, D.J., Jr. (2000). Literacy and technol ogy: Deictic consequences for literacy education. In an information age. In M.L. Kamil, P.B. Mosenthal, P.D. Pearson, & R. Barr (Eds.), Handbook of reading research, 3 743-770. New York: Longman. Lipsitz, J. ( 1980). Growing up forgotten: A review of research and program concerning early adolescence. New Brunswick, NJ: Transaction Books. Lounsbury, J.H. (1984). Perspectives: middle school education, 1964-1984. Ohio: National Middle School Association. Lounsbury, J.H., & Vars, G.F. (1978). A curriculum in the middle school years. New York : Harper and Bow. Mac Iver, D. J. & Epstein, J.L. (1991). Responsive practices in the middle grades: Teacher teams, advisory groups, remedi al instruction, and school transition programs. American Journal of Education, 99(4), 587-622. Marger, R. F., (1968). Developing attitudes towards learning. Belmont: Fearon. McCallister, J. M. (1930). Reading difficulties in studying content subjects. Elementary School Journal, 31, 191-201. McGuire, Mary Sweatt (2004) Ma king the lyrics sing for strugg ling readers: An insider's view. Ed.D. dissertation, Boise State Univ ersity, United States -Idaho. Retrieved June 13, 2007, from ProQuest Di gital Dissertations database.

PAGE 286

273 McCallister, J. M. (1932). De termining the types of readi ng in studying content subjects. School Review, 40, 115-123. McCombs, J. S., Kirby, S. N., Barney, H., & Magee, S. ( 2005). Achieving state and national literacy goa ls, a long uphill road: A Rand Re port to Carnegie Corporation of New York (2nd Ed.).Washington, DC: Alliance for Excellent Education. McKee, P.G. (1934). Reading and literature in the elementary school. Boston: Houghton Mifflin. McPartland, J.M. (1987). Balancing high quality subject-matt er instruction with positive teacher-student relations in the middle grades: Effects of departmentalization, tracking and block scheduling on learning environments (Report No. 15), Maryland: VSO Industries. Merriam, S.B. & Associates (2002). Qualitative research in practice: Examples for discussion and analysis. California: Jossey-Bass. Mariotti, A.S., & Homan, S. P. (2005). Linking reading assessment to instruction: An application worktext for elem entary classroom teachers. (5th Ed.). Mahwah, New Jersey: Lawrence Erlbaum Associates. Miles, M. & Huberman, A.M. (1994). Data analysis: An expanded sourcebook. Thousand Oaks, California: Sage. Moje, E.B. (1996). “I teach students, not s ubjects”: Teacher-stude nt relationships as contexts for secondary literacy. Reading Research Quarterly, 31(2), 172-195. Moje, E.B., Young, J.P., Readence, J. E., & Moore, D. W. (2000). Reinventing adolescent literacy for new times: Perennial and millennial issues. Journal of Adolescent & Adult Literacy, 43(5), 400-410.

PAGE 287

274 Moll, L. C., Gonzalez. N. (1994). Critical issues: Lessons from the research with language minority children. JRB: A Journal of Adolescents & Adult Literacy, 26, 439-456. Moore, D. W. (1996). Contents for literacy in a secondary school. In D.J. Leu, C.K Kinzer, & K.A. Hinchman (Eds.). Literacies for the 21st century: Research and practice. Chicago: National R eading Conference. Moore, D. W., Readence, J. E., & Rickleman, R. J. (1983). An histor ical exploration of content area reading instruction. Reading Research Quarterly, 18(4), 419-438. National Center for Education Statistics, Na tional Assessment of Educational Progress: The Nation’s Report Card, Web site 2005. Retrieved January 7, 2007, from http://nces.ed.gov/nationsreportcard National Reading Panel Report. (2000). Teaching children to re ad: An evidence-based assessment of scientific-r esearch literature on readi ng and its implications for reading instruction. Washington, D.C.: National Institute of Child Health and Human Development. No Child Left Behind Act, Department of Education, Washington, D.C. Retrieved January 7, 2007 from http://www.ed.gov/nclb/landing.jhtml Nunnally, J. C. (1978). Psychometric theory (2nd Ed.). New York: McGraw-Hill. O'Brien, D. G., Stewart, R. A., & Moje, E. B. (1995). Why content literacy is difficult infuse into the sec ondary school: Complexities of curriculum, pedagogy, and school cukture. Reading Research Quarterly, 30, 442-463.

PAGE 288

275 O’Shea, L.J., Sindelar, P.T., & O’Shea, D. J. (1995). The effects of repeated reading and attentional clues on reading fluency and comprehension. Journal of Reading Behavior, 17, 129-142. Onwuegbuzie, A. J. (2003). Expanding the framew ork of internal and external validity in quantitative research. Research in the Schools, 10(1), 71-89. Pearson, P.D. (1985). Changing the face of reading comprehension instruction. Reading Teacher, 38, 724-738. Pearson, P.D., & Gallagher, M.C. (1983). Th e instruction of reading comprehension. Contemporary Educational Psychology, 8, 317-344. Pearson, P.D., & Fielding, L. (1991). Comprehens ion instruction. In R. Barr, M. Kamil, P. Mosenthal, & P.D. Pearson (Eds.), Handbook of Reading Research 2, 815-860. New York: Longman. 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. Educa tion, contract number ED-01-CO-0011. Phelps, S. (2005). Ten years of research on adolescent literacy, 1994-2004: A review. Learning Point Associates. North Centra l Regional Laboratory, Institute of Education Sciences, U.S. Department of Education, contract number ED-01-CO0011. Piaget, J. (1964). Development and Learning. In R.E. Ripple and V. N. Rockcastle, Piaget Rediscovered. New York: Cornell University Press.

PAGE 289

276 Pikulski, J.J. (1991). The tran sitions years: Midd le school. In J. Flood, J.M. Jensen, D. Lapp, & J. Squire (Eds.), Handbook of Research on Teaching English Language Arts ( pp. 303-319). New York: Macmillan Publishing Company Pikulski, J.J., & Chard, D.J. (2005). Fluenc y: Bridge between decoding and reading comprehension. The Reading Teacher, 58 (6), 510-519. Pressley, L.C. (1923). The technical vocabulary of public school subjects. Bloomington: Public School Publishing Co. Rasinski, T.V. (2004). Cr eating fluent readers. Educational Leadership, 61(6), 46-51. Resnick D. P. & Ressnick, L. B. (1977). The nature of literacy: An historical exploration. Harvard Educational Review, 47, 370-385. Rice, J.M. (1893). The public-school system of the United States. New York: Century. Ritter, B. T. & Lofland, W. T. (1924). The relationship between reading ablity as measured by certain standard tests and th e ablity required in interpertation of printed matter involving reason. Elementary School Journal, 24, 529-546. Rosenshine, B. (1980). How time is spent in elementary classrooms. In Denham and Leiberman (Eds.). Time to learn. Washington: Department of Health, Education and Welfare, National Institute of Education. Rothstein, R. (2000). How standa rdized test drop the ball. New York Times, September, 2000 Ryan, A. M. (2000). Peer group as a context fo r the socializati on of adolescents’ motivation, engagement, and achievement in school. Educational Psychologist, 35(2), 101-111.

PAGE 290

277 Sackes, P (2000). Standardized minds: The high price of America’s testing culture and what we can do to change it. Perseus Publishing: Cambridge, MA. Salvin, R.E (1988). Synthesis of research on grouping in elementary and secondary schools. Educational Leadership, 46(1), 67-77. Samuels, S.J. (1979). The me thod of repeated readings. The Reading Teacher, 41, 756-760. Samuels, S.J & Farstrup, A.E. (2006). What research has to say about fluency instruction. Delaware: International Reading Association. Sample, K.J. (2005). Promoting fluency in adolescents with reading difficulties. Interventions in Schools and Clinics, 40(4). 243-246. Schunk, D.H., & Zimmermann, B. J. (1916). Mo deling and self-efficacy influences on children’s development and self-regulation. In J. J uvonen & K. Wentzel (Eds.). Social motivation: Understanding children’s school adjustment, (pp. 154180).Cambridge, England: Cambridge University Press. Schreiber, P.A. (1980). On the ac quisition of reading fluency. Journal of Reading Behavior. 12 177-186. Schreiber, P.A. (1987). Prosody and structure in children’s syntactic processing. In R. Horowitz and S. J. Samuels’s (Eds.) Comprehending oral and written language. (pp. 243-270). New York: Academic Press. Schreiber, P.A. (1991). Understanding prosody’s role in reading acquisition. Theory into Practice 30, 158-164. Schreiber, P.S., & Read, C. (1980). Children’ s use of phonetic cues in spelling, parsing, and maybereading. Bulletin of the Orton Society 30 209-224.

PAGE 291

278 Schunk, D. H. & Rice, J. M. (1993). Strate gy fading and progress feedback: Effects of self –efficacy and comprehension among students receiving remedial reading service. Journal of Special Education 27 257-276. Scribner, S., & Cole, M. (1981). The psychology of literacy. Cambridge, MA: Harvard University Press. Shaywitz, B.A., Pugh, K. R., Jenner, A. R., Fulb right, R. K., Fletcher, J. M., Gore, J. C., & Shaywitz, S. E. (2000). The neurobiology of reading disability (dyslexia). In M. L. Kamil, P. B. Mosenthal, P. D. Pearson, R. Barr (Eds.). Handbook of Reading Research, 3 141-151. Mahwah New York: Erlbaum. Simmons, R.G., & Blyth, D.A. (1987). Moving into adolescence: Th e impact of pubertal change and school context. Hawthorn: Random House. Smith, N. B. (1919). Correlations of ability in reading with the general grades in high school. School Review, 27, 493-511. SPSS Inc. (2006). SPSS Base 15.0 for Windows SPSS Inc., Chicago IL. Spring, J. (1986). The American school. New York : Longman. Stake, R.E. (2000). Case studies. In N.K. Denzin & Y.S. Lincoln (Eds.), Handbook of Qualitative Research, (2nd Ed.) (pp. 509-536). Thousand Oaks: Sage Publications. Stayter, F.Z., & Allington, R.L., (1991). Fluency and the understanding of text. Theory into practice, 30(3), 143-148. Stanovich, K.D. (1980). Effects of explic it teaching and peer tutoring on the reading achievement of learning disable and low performing students in regular classrooms. Reading Research Quarterly 16, 32-71.

PAGE 292

279 Stanovich, K.D. (1984) The interactive compen satory model of reading. A confluence of Developmental, experimental and educational psychology. Remedial and Special Education, 5(3), 11-19 State of Florida Department of Education ( 2001). Understanding th e FCAT reports 2006. Retrieved January 7, 2007, from http://www.firn.edu/doe/sas/fcat.htm Stevens, J.P. (1999). Intermediate statisti cs: A modern approach. Mahwah, NJ: Erlbaum. Stevens, J.P. (2002). Applied multivariate statistics for the social sciences (4th Ed.). Mahwah, NJ: Lawrence Erlbaum Associates. Tabachnick, B. G., & Fidell, L. S. (1989). Using multivariate statistics, (2nd Ed.). CA: Harper Collins Publishers. Tanner, J. M., (1962). Growth at adolescence. 2nd ed., Oxford England: Blackwell Scientific Publications. Tashakkori, A., & Teddlie, C. (2003). Mixed methodology: Combining qualitative and quantitative approaches. (Eds.). Thousand Oaks, CA: Sage Publications. Taylor, N. E., Wade, M. B., & Yekovich, F. R. (1985). The effects of test manipulation and multiple reading strategies on th e reading performance of good and poor readers. Reading Research Quarterly, 20, 506-574. Therrien, W.J. (2004). Fluency and comprehensi on gains as a result of repeated reading: A meta-analysis. Remedial and Special Education, 25 (4), 252-261. Thorndike, E.L. (1917). Reading as reasoning: A study of mist akes in paragraph reading. Journal of Educational Psychology, 8, 276-282. Thorndike, E.L. (1921). The teachers word book. New York: Columbia University Teachers College.

PAGE 293

280 Topping, K.J. (2006). Building reading fl uency: Cognitive, behavioral, and socioemotional factors and th e role of peer-mediated learning. In S.J Samuels, A.E. Farstrup, (Eds.), What research has to say about fluency instruction (pp. 106-129). Delaware: Internationa l Reading Association. Tovani, C. (2000). I read it but I don’t get it. Portland, ME: Steinhouse. Vacca, R. T. & Vacca, J. L. (2002). Content area reading: Literacy and learning across the curriculum (7th ed.). Boston MA: Allyn & Bacon. Van Til, W.G., & Lounsbury, J. (1961). Modern education for the junior high school years. New York: Boobs-Merril. Vygotsky, L. (1978). Mind and society. Cambridge, MA. Harvard University Press. Wade, S.E., & Moje, E.B. (2000). The role of te xt in classroom learning. In M. L. Kamil, P. B. Mosenthal, P. D. Pearson, R. Barr (Eds.). Handbook of reading research (Vol. 3, pp. 609-627). Mahwah New York: Erlbaum. Wagner, E. (1938). Reading and ninth grade achievement. (Teachers College Contributions to Education, No. 756). New York: Columbia University Teachers College. Webster, P. R. (2002). Computer-based t echnology and music teaching and learning. In R. Colwell& C. Richardson (Eds.), The New Handbook of Research on Music Teaching and Learning. (pp. 416-439). Wentzel, K.R. (1996). Social and academic mo tivation in middle school: Concurrent and Long term relations to academic effort. Journal of Early Adolescences. 16(4), 390-406.

PAGE 294

281 Whipple, G. M. (1920). New materials of instruction. (Nineteenth Yearbook of the National Society for the Study of Educa tion Part I) Bloomingt on: Public School Publishing Co. Wiles, J., & Thomason, J. ( 1975). Middle school research 19681974: A review of substantial studies. Educational Leadership, 32, 412-423. Wiley, K.B. (1977). The status of pre-college science, mathematics, and social science educational practices in th e U.S schools: An overview and summaries of three studies. Washington DC: U.S. Government Printing Office. Witty, P. A. (1948). Current role and eff ectiveness of reading among youth. In N. B. Henry (Ed.), Reading in the high school and college (Forty-Seventh Yearbook of the National Society for the Study of E ducation Part I) Bloomington: Public School Publishing Co. Worthy, J., Moorman, M. & Turner, M. (1999). Wh at Johnny likes to read is hard to find in school. Reading Research Quarterly, 3(1), 12-27. Zimmerman B. J. (2000). Attaining self-regula tion: A social –cogniti ve perspective. In M. Boekaerts, P.R., Pintrich, and M.H. Zeidner. Handbook of self-regulation (Vol. 41, pp. 64-70). San Diego CA: Academic Press.

PAGE 295

282 APPENDICES

PAGE 296

283 Appendix A: Qualitative R eading Inventory-4 (QRI-4)

PAGE 297

284 Appendix A: (Continued)

PAGE 298

285 Appendix A: (Continued)

PAGE 299

286 Appendix B: Qualitative Matrix An Example of the Qualitative Classroo m Observation Notes Transferred to the Categorical Matrix for the Peer Interactions Information Exchange Peer discussion/talk direct quotes from conversations during the literacy task Peer 1“ How did you get the song to slow down” Peer 2 “ Click on this button” Modeling Peer Observation/ through descriptions of interactions during the literacy task He looked around the classroom started to smile and went back to playing the drums Peer Pressure Social reinforcement/ descriptions through looks / comments/ laughs during the literacy task T hit the drum wrong, M laughed and then the class laughed T turned red and put his head down,

PAGE 300

287 Appendix C: Construct Key Peer Interactions:During the Literacy Task of Rereading Through Singing Construct Heading Construct Definition Group Characteristics Extrinsic Motivation Students engaging in a task because of a reward or punishment. Group Characteristics Intrinsic Motivation Students engaging in a task for their own personal learning. Group Characteristics Peer Observations Peers’ observing each other that influences behavior changes. Group Characteristics Peer Hierarchy Social system in the classroom that positions some members of the peer group above other. Group Characteristics Peer Support Peers providing or showing support for one another Group Characteristics Dominant and Vulnerable Peers Peer positions that place dominant peers over the more vulnerable peers. Group Characteristics Students’ Perceptions of Alternative Text How the students understand and perceive the alternative text. Group Characteristics Disequilibrium A cognitive state that occurs as new and different information occurs in the learning. Group Characteristics Fake Rereading/Singing The students appear to be singing the songs however, they are not. Group Characteristics Peer Leaders Students in the group take on the role of leadership over the other students. Instructional Procedures Alternative Approaches to Singing Teaching approaches different to practice singing that used the drums. Instructional Procedures Safe Risk-Free Environment A setting where the students feel comfortable enough to take a risk.

PAGE 301

About the Author Marie C. Biggs is currently an assistan t professor at St. Petersburg College of Education. Her background is in the area of reading research with an emphasis on working with struggling “at ri sk” readers. Over the last twenty years she has held positions as a reading specialist for K-12 students and faculty positions as an instructor in Literacy Education and Assessment at Wheel ock College in Boston and St. Petersburg College in Tampa, and the University of S outh Florida where she taught undergraduates and graduate preservice teacher s. Also, in her fieldwork as a professional development coordinator and consultant through Tufts Un iversity in Boston she provided outreach services and taught graduate courses in read ing to in-service teachers in various school systems throughout New England. Currently, her research interests are focu sed on the use of A lternative Texts for Struggling Adolescent Readers grades 4-12. In 2004, she conducted a pilot study using an interactive singing software program, Carry-a-Tune with middle school struggling readers. This was followed by becoming the lead field researcher in a seven-site study of 4th through 12th grade struggling readers in the Tampa area, who were using the Carry-aTune program. In addition, she was the coordinator for a multi-district study with 1,300 ELL across three districts in Florida. Her im plications and findings from the research have been presented throughout the United States at conferences that focus in the field of Literacy, Music, Technology, and Educational Leadership.


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 2200397Ka 4500
controlfield tag 001 001988959
005 20090216141253.0
007 cr bnu|||uuuu
008 090216s2007 flu s 000 0 eng
datafield ind1 8 ind2 024
subfield code a E14-SFE0002184
035
(OCoLC)307532767
040
FHM
c FHM
049
FHMM
090
LB1139.23 (Online)
1 100
Biggs, Marie.
0 245
Reading fluency through alternative text :
b rereading with an interactive sing-to-read program embedded within middle school music classroom
h [electronic resource] /
by Marie Biggs.
260
[Tampa, Fla] :
University of South Florida,
2007.
500
Title from PDF of title page.
Document formatted into pages; contains 287 pages.
Includes vita.
502
Dissertation (Ph.D.)--University of South Florida, 2007.
504
Includes bibliographical references.
516
Text (Electronic dissertation) in PDF format.
520
ABSTRACT: Singing exaggerates the language of reading. The students find their voices in the rhythm and bounce of language by using music as an alternative text. A concurrent mixed methods study was conducted to investigate the use of an interactive sing-to-read program Tune Into Reading (Electronic Learning Products, 2006) as an alternative text, embedded within a heterogeneous music classroom. Measured by the Qualitative Reading Inventory-4 (QRI-4) (Leslie & Caldwell, 2006), the fluency, word recognition, comprehension, and instructional reading level of the treatment students were compared to their counterparts who sang as part of the regular music program. Concurrently, this investigation also provided a description of the peers' interactions during the literacy task assigned by the music teacher. The intent of this study was to address the following three research questions.Furthermore, the use of the interactive program provided opportunities for differentiated reading level achievement. Finally, group dynamics highly influenced the early adolescent's motivation, engagement, participation, and successful outcomes in reading fluency.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
590
Advisor: Susan Homan, Ph.D.
653
Singing
Rereading
Fluency
Early adolescents
Alternative text
690
Dissertations, Academic
z USF
x Childhood Education
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.2184