USF Libraries
USF Digital Collections

The effect of computer-delivered phonological awareness training on the early literacy skills of students identified as ...

MISSING IMAGE

Material Information

Title:
The effect of computer-delivered phonological awareness training on the early literacy skills of students identified as at-risk for reading failure
Physical Description:
Book
Language:
English
Creator:
Gale, Deanne
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Intervention -- Program Evaluation -- Computer Assisted Instruction -- Elementary School Students -- Phonemes
Program evaluation
Computer assisted instruction
Elementary school students
Phonemes
Dissertations, Academic -- Interdisciplinary Education -- Specialist -- USF
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: The current study examined the effects of two computer-delivered phonological awareness training programs (Earobics Step 1 and Lexia Early Reading) on the early literacy skills of kindergarten and first grade students at risk for reading failure. The study utilized a multi-group pretest-treatment-posttest design. Student participants, who were identified for the study through a school-wide screening, were randomly assigned to one of three groups (i.e., Earobics, Lexia Early Reading, or control), and their progress was monitored throughout a five-week intervention period. Results using an analysis of covariance (ANCOVA) to examine differences in adjusted mean post-test scores indicated that the Earobics program produced better outcomes than the Lexia and control groups as measured by the Dynamic Indicators of Basic Early Literacy Skills. Results of a hierarchical linear modeling (HLM) analysis examining initial status and rates of growth also indicated greater rates of change among the Earobics group when compared with the Lexia and control groups. The Earobics program was shown to be an effective intervention for improving early literacy skills for students at risk for reading failure. Implications of the study for working with early elementary students who show deficits in phonological awareness are discussed.
Thesis:
Thesis (Ed.S.)--University of South Florida, 2006.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Deanne Gale.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 121 pages.

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 - 001793978
oclc - 145585042
usfldc doi - E14-SFE0001531
usfldc handle - e14.1531
System ID:
SFS0025849:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001793978
003 fts
005 20070626135105.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 070626s2006 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001531
040
FHM
c FHM
035
(OCoLC)145585042
049
FHMM
090
LB1555 (ONLINE)
1 100
Gale, Deanne.
4 245
The effect of computer-delivered phonological awareness training on the early literacy skills of students identified as at-risk for reading failure
h [electronic resource] /
by Deanne Gale.
260
[Tampa, Fla] :
b University of South Florida,
2006.
3 520
ABSTRACT: The current study examined the effects of two computer-delivered phonological awareness training programs (Earobics Step 1 and Lexia Early Reading) on the early literacy skills of kindergarten and first grade students at risk for reading failure. The study utilized a multi-group pretest-treatment-posttest design. Student participants, who were identified for the study through a school-wide screening, were randomly assigned to one of three groups (i.e., Earobics, Lexia Early Reading, or control), and their progress was monitored throughout a five-week intervention period. Results using an analysis of covariance (ANCOVA) to examine differences in adjusted mean post-test scores indicated that the Earobics program produced better outcomes than the Lexia and control groups as measured by the Dynamic Indicators of Basic Early Literacy Skills. Results of a hierarchical linear modeling (HLM) analysis examining initial status and rates of growth also indicated greater rates of change among the Earobics group when compared with the Lexia and control groups. The Earobics program was shown to be an effective intervention for improving early literacy skills for students at risk for reading failure. Implications of the study for working with early elementary students who show deficits in phonological awareness are discussed.
502
Thesis (Ed.S.)--University of South Florida, 2006.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 121 pages.
590
Adviser: Linda Raffaele Mendez, Ph.D.
653
Intervention
Program Evaluation
Computer Assisted Instruction
Elementary School Students
Phonemes
Program evaluation.
Computer assisted instruction.
Elementary school students.
Phonemes.
690
Dissertations, Academic
z USF
x Interdisciplinary Education
Specialist.
773
t USF Electronic Theses and Dissertations.
0 856
u http://digital.lib.usf.edu/?e14.1531



PAGE 1

The Effect of Computer-Delivered P honological Awareness Trai ning on the Early Literacy Skills of Students Identif ied as At-Risk for Reading Failure by Deanne Gale A thesis submitted in partial fulfillment of the requirements for the degree of Education Specialist Department of Psychological and Social Foundations College of Education University of South Florida Major Professor: Lind a Raffaele Mendez, Ph.D. Kelly A. Powell-Smith, Ph.D. Robert Dedrick, Ph.D. Date of Approval: February 22, 2006 Keywords: Intervention, Program Eval uation, Computer Assisted Instruction, Elementary School Students, Phonemes Copyright 2006, Deanne Gale

PAGE 2

i Table of Contents List of Tables iii List of Figures v Abstract vi Chapter One: Introduction 1 Overview 1 Description of the Current Study 4 Chapter Two: Review of the Literature 6 Phonological Awareness 10 Instruction in Phonological Awareness 13 Phonological Awareness Intervention 20 Computer Instruction in P honological Awareness 22 The Current Study 32 Chapter Three: Method 33 Participants 33 Measures 33 Letter Naming Fluency 34 Initial Sounds Fluency 35 Phoneme Segmentation Fluency 36 Nonsense Word Fluency 36 Oral Reading Fluency 37 Student Ratings of Computer Training 38 Teacher Ratings of Computer Training 39 Procedure 39 Computer-Administered Phonological Awareness Training Programs 44 Earobics Step 1 45 Lexia Early Reading 47 Analysis 48 Chapter Four: Results 52 Descriptive Information 52 Research Question One: Kindergarten Outcomes 55 Research Question Two: First Grade Outcomes 62 Research Question Three: Kindergarte n and First Grade Development 71

PAGE 3

ii Research Question Four: Kindergarte n and First Grade Comparison 87 Teacher and Student Survey Results 89 Chapter Five: Discussion 95 Major Findings of This Study 95 Kindergarten findings 96 First grade findings 98 Kindergarten and first grade comparison 100 Consumer satisfaction survey 101 Practical Implications 102 Limitations 103 Contribution to the Literature and Di rections for Future Research 103 References 106 Appendices 112 Appendix A: Student Survey 113 Appendix B: Teacher Survey 114 Appendix C: Informed Consent 115 Appendix D: Rotation Schedule 118 Appendix E: Computer Training Checklist 119 Appendix F: Proctor Schedule 120 Appendix G: Student Activity Card 121

PAGE 4

iii List of Tables Table 1 DIBELS benchmarks a nd correlation coefficients 38 Table 2 Earobics Step 1 game descriptions and targeted skills 45 Table 3 Lexia Early Reading game de scriptions and targeted skills 47 Table 4 Demographic information fo r kindergarten and first grade 53 Table 5 Kindergarten DIBELS pre-test scores by group 56 Table 6 Analysis of variance for kindergarten pre-test measures 57 Table 7 Kindergarten DIBELS post-test scores by group 58 Table 8 Analysis of variance and analysis of covariance for kindergarten post-test measures 60 Table 9 ANCOVA Post hoc analysis for kindergarten Initial Sounds Fluency 61 Table 10 ANCOVA Post hoc analysis for kindergarten Phoneme Segmentation Fluency 62 Table 11 First grade DIBELS pre-test scores by group 63 Table 12 Analysis of variance for first grade pre-test measures 64 Table 13 First grade DIBELS post-test scores by group 66 Table 14 Analysis of covariance fo r first grade post-test measures 67 Table 15 ANCOVA Post hoc analysis fo r first grade Letter Naming Fluency 68 Table 16 ANCOVA Post hoc analysis for first grade Phoneme Segmentation Fluency 69 Table 17 ANCOVA Post hoc analysis for first grade Nonsense Word Fluency 70 Table 18 ANCOVA Post hoc analysis fo r first grade Oral Reading Fluency 70

PAGE 5

iv Table 19 Unconditional model for kindergarten ISF 72 Table 20 Conditional mode l for kindergarten ISF 72 Table 21 Unconditional model for kindergarten LNF 74 Table 22 Conditional mode l for kindergarten LNF 75 Table 23 Unconditional model for kindergarten PSF 76 Table 24 Conditional mode l for kindergarten PSF 77 Table 25 Unconditional model for first grade LNF 78 Table 26 Conditional model for first grade LNF 79 Table 27 Unconditional model for first grade PSF 81 Table 28 Conditional mode l for first grade PSF 81 Table 29 Unconditional model for first grade NWF 83 Table 30 Conditional mode l for first grade NWF 84 Table 31 Unconditional model for first grade ORF 85 Table 32 Conditional model for first grade ORF 86 Table 33 Student survey results 90 Table 34 Teacher survey results 92

PAGE 6

v List of Figures Figure 1 Initial status and growth rate of the kindergarten Ear obics, Lexia, and control groups on In itial Sounds Fluency 73 Figure 2 Initial status and growth rate of the kindergarten Ear obics, Lexia, and control groups on Letter Naming Fluency 75 Figure 3 Initial status and growth rate of the kindergarten Ear obics, Lexia, and control groups on Phoneme Segmentation Fluency 78 Figure 4 Initial status and growth rate of the first grade Earobics, Lexia, and control groups on Letter Naming Fluency 80 Figure 5 Initial status and growth rate of the first grade Earobics, Lexia, and control groups on Phoneme Segmentation Fluency 82 Figure 6 Initial status and growth rate of the first grade Earobics, Lexia, and control groups on Nonsense Word Fluency 85 Figure 7 Initial status and growth rate of the first grade Earobics, Lexia, and control groups on Or al Reading Fluency 87

PAGE 7

vi The Effect of Computer-Delivered P honological Awareness Training on the Early Literacy Skills of Students Identif ied as At-Risk for Reading Failure Deanne Gale ABSTRACT The current study examined the effects of two computer-delivered phonological awareness training programs (Earobics Step 1 and Lexia Early Reading) on the early literacy skills of kindergarten and first grade students at ri sk for reading failure. The study utilized a multi-group pretest-treatment-pos ttest design. Student participants, who were identified for the study through a school-w ide screening, were randomly assigned to one of three groups (i.e., Ear obics, Lexia Early Reading, or control), and their progress was monitored throughout a five-week interven tion period. Results using an analysis of covariance (ANCOVA) to examine differences in adjusted mean post-test scores indicated that the Earobics program produced better outcomes than the Lexia and control groups as measured by the Dynamic Indicators of Basic Early Literacy Skills. Results of a hierarchical linear modeling (HLM) analys is examining initial status and rates of growth also indicated greater rates of change among the Ea robics group when compared with the Lexia and control groups. The Earobi cs program was shown to be an effective intervention for improving early literacy skills for students at risk for reading failure. Implications of the study for working with early elementary students who show deficits in phonological awareness are discussed.

PAGE 8

1 Chapter One Introduction Overview Reading is a pivotal skill that is crit ical to a successful academic career. Once students complete the early stag es of learning to read, their later years in education focus on reading to learn. Thus, impaired readi ng skills limit the opport unities students have for learning through independent reading. Beyond the importance of reading within the educational realm, literacy also enables one to have a more enriched personal, social, and professional life. In view of the fact that becoming a proficient reader has such far reaching effects, ensuring that all children be come skillful readers has become a topic of national discussion. The No Child Left Behind Act (Public La w 107-110) has substantially influenced the focus of early reading achievement. One of the pillars of this act is ensuring that all children are proficient readers by the end of third grade. Research has demonstrated that students who are at risk for read ing failure acquire reading skills at a rate that is different from their peers, which places them on a read ing trajectory that is resistant to change (Juel, 1988). Given the research on the persistence of reading difficulty once it is experienced at a young age, it is imperative th at we identify and inte rvene with students who are at risk for reading failure early. This is especially important when considering the research that demonstrates that we ha ve a short period of time in which we can

PAGE 9

2 implement meaningful interventions to change the course of a student s reading trajectory (Lyon & Chhabra, 1996). Substantial research has examined the connection between phonological awareness and reading achievement (Adams, 1990). Phonological awar eness is a critical component of reading that is typically ac quired early in the development of various reading skills. Phonological awareness is mo st often described as the awareness of sounds in spoken words at the word, sylla ble, and phoneme level. Strong phonological awareness skills increase a students understanding of the alphabetic principle, which is a vital aspect of reading development (Torge sen & Mathis, 1999). Several studies have documented the longitudinal relationship be tween early phonological awareness skills and later reading ability (Byrne & Fiel ding-Barnsley, 1995; Torgesen, Wagner, & Rashotte, 1994; Wagner, Torgesen, Rashotte, Hecht, Barker, Burgess, Donahue, & Garon, 1997). Additionally, studies also have revealed that students who experience reading difficulties consistently display impairments in the area of phonological awareness (Al Otaiba & Fuchs, 2002; Fletcher Shaywitz, Shankweiler, Katz, Liberman, Stuebing, Francis, Fowler, & Shaywitz, 1994). The importance of phonological awareness skills in reading development suggests that this is an area that can and should be targeted for early intervention. Although research has reported the signi ficance of phonological awareness, not all reading curricula include explicit training in this area. Some students will develop phonological awareness skills desp ite the absence of direct instruction; however, for those who do not, specific training is necessa ry. Several training programs (both teacherled and computer-delivered) have been devel oped to provide this explicit inst ruction.

PAGE 10

3 Teacher-led programs such as Phonological Awareness Training for Reading (Torgesen & Bryant, 1994) and Lindamood Phoneme Sequencing Program (Lindamood & Lindamood, 1998) have been used in a variet y of settings to successfully increase phonological awareness skills. Th ese programs, however, are typically delivered in small groups and require time and staff resources that are not always available in the general education classroom. Therefore, additional types of intervention resources need to be explored. Due to the progress and benefits of tec hnology, an increasing nu mber of computer software programs have been designed to deliver aspects of reading instruction that were once only available from teachers. Advances in technology such as high-quality sound, digitized speech, colorful gra phics, and interactive design have added to the instructional value of computer software programs. A dditionally, computer software programs have benefits such as allowing students to work independently at their own pace and instructional level, providing opportunities for individual feedback and motivation, and repetition of needed skills. Although there has been an increase in the research exploring the utility of computer software in the cl assroom and its effect on reading instruction, there is still considerable need for research in this area. Specifically, with regard to phonological awareness, the Re port of The National Reading Panel stated, More research is needed to determine whether a nd how PA might be taught more effectively using computers (National Institute of Child Health and Human Development, 2000, p. 44).

PAGE 11

4 Description of the Current Study The purpose of the current study was to examine the effect of two computerdelivered phonological awareness training pr ograms on the early literacy skills of kindergarten and first grade students at risk for reading failure. Students in this study were identified as at-risk for reading failure based on an ongoing screening process that the participating elementary school re gularly conducts as a part of a Reading First grant. Specifically, students in the participating school in grades kindergarten through three are assessed four times per year using the Dynami c Indicators of Basic Early Literacy Skills (DIBELS). Kindergarten and first grade stud ents who were identified during the Fall assessment period as needing intensive substantial interven tion based on their performance on the DIBELS were recruited for this study. Students were randomly assigned to either one of two experimental groups or the control group. In one experimental group, the students used the computer-based phonological awareness program Earobics Step 1. In the second experimental group, the students used the computer-based phonological awareness program Lexia Early Reading. A third group (i.e., the contro l group) received no specific intervention designated by the study. Students used their re spective computer programs in the school computer lab 20 minutes daily for 25 days, resulting in a total of eight hours of exposure to the intervention. Several DIBELS measures were administer ed to all students in the study as preand post-tests. Specifically, the Initial Sounds Fluency, Letter Naming Fluency, and Phoneme Segmentation Fluency measures were administered to participating kindergarten students. The Letter Naming Fluency, Phoneme Segmentation Fluency,

PAGE 12

5 Nonsense Word Fluency, and Oral Reading Fluency measures were administered to first grade students. In addition, alternate forms of the same DIBELS measures used for preand post-tests were administered to all stud ents participating in the study weekly to monitor the progress of indivi dual participants. At the c onclusion of the intervention period, students participating in the study and their teachers were given a survey to assess their opinions regarding the computer programs.

PAGE 13

6 Chapter Two Review of the Literature Within the past decade, there has been a national trend in education focused on delivering reading instruction that is based on scientific research in order to assure that all children become proficient readers. Reading is an essential skill that, when absent or impaired, affects not only a childs educatio nal career but also his or her personal fulfillment and life opportunities. Reading in struction has been a cornerstone of our educational system for decades; however, nationa l statistics indicate th at the majority of children are not becoming skillful readers. In the year 2000, the National Center for Educational Statistics reported that only 32 percent of the na tions fourth grade students were proficient in reading (Na tional Center for Education Sta tistics, 2003). In the course of recent years, professionals and experts have been called upon to examine and synthesize the abundance of reading research in order to begin to address this staggering statistic. In 1997, the U.S. Department of Educati on and the U.S. Department of Health and Human Services asked the National Acad emy of Sciences to create a committee to study the effectiveness of in terventions for young children w ho are at risk for reading difficulties. After reviewing research on normal reading development and instruction, risk factors for reading failure, and preventi on of reading failure, the National Research Council Committee on Preventing Reading Difficulties in Young Children reported several recommendations for reading instruct ion and addressing at-risk students. The

PAGE 14

7 NRC Committee recommended that instruction in the areas of the alphabetic principle, reading sight words, reading words by mapping speech sounds to parts of words, achieving fluency, and comprehension (p. 6) be included in every primary-grade classroom. Recognizing the need to addre ss at-risk students early, the committee also recommended that children have access to early childhood environments that promote skills that have been identified as predicto rs of later reading success (Snow, Burns, & Griffin, 1998). In 1997, the National Reading Panel (NRP) was assembled by the Director of the National Institute of Child Health and Human Services in consultation with the Secretary of Education at the requ est of Congress. The NRP, consis ting of 14 experts in the field of reading, was asked to complete a comprehensiv e review of various approaches used to teach children to read, how these approaches should be used in the classroom, and directions for future resear ch. Building upon the work of the National Research Council Committee on Preventing Reading Difficu lties in Young Children, the NRP initially screened over 100,000 studies in the areas of alphabetics, fluency, and comprehension. Subsequently, regional public hearings were held to assess th e needs of teachers, parents, students, and policymakers. Following the regional hearings, the NRP divided into subgroups to address research in the following areas: Alphabetics, Fluency, Comprehension, Teacher Education and Read ing Instruction, and Computer Technology and Reading Instruction. The subgroups revi ewed studies that met a set of rigorous research methodological standard s. A formal statistical meta-analysis was completed for topic areas that contained a sufficient number of studies, and a quali tative analysis was completed for the remaining topics. The NR P met over a period of two years to prepare

PAGE 15

8 the results, and in the year 2000, they submitted Report of the National Reading Panel. Teaching Children to Read: An Evidence-based Assessment of the Scientific Research Literature on Reading and its Impl ications for Reading Instruction (National Institute of Child Health and Human Development, 2000) The Report of the National Reading Panel (National Institute of Child Health and Human Development, 2000) included findings on the most effective instructional practices in what were considered the five essential elements of reading instruction: phonemic awareness, phonics, fluency, vocabulary, and text comprehension. In the area of phonemic awareness, the NRP found that teaching phonemic awareness to children improves their reading more than reading in struction that does not directly address phonemic awareness. Phonemic awareness inst ruction was found to be most effective when it was taught explicitly and systemati cally, focused on one or two types of phoneme manipulations, and was taught in small groups. Similarly, systematic phonics instruction was found to benefit students in grades kindergarten through si x. Synthetic phonics instruction (i.e., teaching students to convert letters to sounds and blend sounds into words) was found to have a positive effect on students with lear ning disabilities, lowachieving students, and students from low socioeconomic backgrounds. The area of fluency was found to be one of the critical factors in reading co mprehension. Across a variety of grade levels, repeated oral readi ng with teacher, peer, or parent guidance was found to have a positive impact on word r ecognition, fluency, and comprehension. Vocabulary knowledge was found to be another cr itical factor in reading comprehension. While there was little resear ch on the best methods of vocabulary instruction, the National Reading Panel found that vocabular y should be taught both directly and

PAGE 16

9 indirectly and include repetition and multiple exposures to vocabulary terms. Instruction can be enhanced through learning in rich contexts, incidental learning, and computer instruction. Finally, with regard to comprehension, th e National Reading Panel found that students benefited from explicit in struction in a comb ination of reading comprehension techniques such as recall, question answering, question generation, and summarization of text. These findings of the National Reading Panel have been recreated into both teacherand parentfriendly literature in an effo rt to support evidence-based reading practices in schools and homes. In 2001, President George W. Bush a nnounced his proposal for educational reform within the framework of the No Child Left Behind Act (NCLB Act). Included in this act were his plans for increased accountability for st ates, school districts, and schools; more choice for parent s and students; more locali zed flexibility for federal education dollars; and a stronger emphasis on reading, particularly for younger children. The NCLB Act was passed into law in Janua ry 2002 and has had a remarkable influence on education. One aspect of the Presidents ag enda with regard to reading is to ensure that all children can read by th e end of third grade. The Reading First initiative was subsequently put into place to help achieve this goal. This initiative increased the federal investment in scientifically-based reading instruction programs in the early grades. Additionally, it provides state grants whic h are divided among local communities on a competitive basis. The grant monies are intended to provide resources for screening and diagnostic assessment to identify students in grades kindergarten thr ough three who are at risk for reading failure (NCLB Executive Summary, 2002).

PAGE 17

10 The importance of early identification a nd intervention with students who are at risk for developing reading difficulties is supported in an article by Good, Simmons and Smith (1998). This article discussed tw o common characteristics among children with reading difficulties. The first characteristic a reading progress trajectory that deviates significantly from peers without reading difficulties, suppor ts the importance of early identification. This trajectory indicates that students who have poor initial reading skills in early grades are likely to continue with these p oor reading skills in later grades. This may be partially due to poor initial reading sk ills leading to fe wer opportunities for exposure to print and eventual discourageme nt and frustration felt by the student, which in turn, contributes to reduced exposure to print. Once a student is far enough along this trajectory, catching up to aver age reading peers becomes very difficult. The second characteristic, phonological de ficits, pinpoints a specific area to target for early identification and interventi on. Students who are low in phonological awareness skills are at risk for developing reading difficul ties. Additionally, phonological awareness is a skill that can be assessed and taught at an ea rly age, making it an ideal area to target for early identification and intervention. Phonological Awareness Phonological awareness is defined by Torgesen and Mathis (1999) as ones sensitivity to, or explicit awareness of, the phonological structure of words in ones language (p. 2). Adams (1990) discussed several levels of phonological awareness ranging from an awareness of rhyme to being able to switch or substitute the various parts of a word. More specifically, phonological awareness is a broad term that includes skills such as identifying and making oral r hymes, identifying and working with syllables

PAGE 18

11 in spoken words, identifying and working with onsets and rimes in spoken syllables, and identifying and manipulating in dividual sounds in words (phonemic awareness). Before children can learn to read words, they need to understand that language is made up of individual sounds or phonemes. The National Reading Panel r ecognized this, and included it as one of five critic al areas of read ing instruction. The long-term association of phonologi cal awareness and reading has been documented in two studies. In order to explore this relations hip between phonological processing and reading skills, Torgesen, Wagner and Rashotte (1994) completed a longitudinal study with 244 students in six elementary schools. The students were randomly selected, spoke English fluently, a nd were screened to detect articulation difficulties. At the beginning of kindergarten, first, and second grade, all students participating in the study were given 22 tasks measuring five areas of phonological processing, reading and prereading skills, and general verbal ability. When examining the stability of phonological skil ls over time, it was found that individual differences in phonological skills were stable from kindergarten through second grade. A further analysis of relationships between phonological skills and reading revealed that phonological awareness in kinderg arten had a significant rela tionship with first grade word-reading skills (coefficien t = .67). In addition, prer eading skills in kindergarten were found to have an effect on the development of phonological skills (coefficient = .23); however, the effect was moderate compar ed to the effect of phonological skills on reading. To expand upon the previous study, Wagner et al. (1997) continued to assess phonological skills, word-level reading, and verb al aptitude with the same sample of

PAGE 19

12 students through fourth grade. At the end of fourth grade, 216 st udents remained from the previous 244. The results of this analysis indicated th at individual differences in phonological skills continued to remain stab le through fourth grade. Additionally, individual differences in phonological awar eness largely influenced subsequent individual differences in word-l evel reading each year through fourth grade. The results of this study add to the previous study in the finding that these differences in phonological awareness and their influence on reading abilit y are not limited to early reading development. Research also supports the importance of phonological awareness by showing that students with reading difficulties are consistently impaired in this area. A review of the literature by Al Otaiba and Fuchs (2002) f ound that the majority of students who were unresponsive to early reading intervention had phonological awareness deficits. After a thorough search of the literature, the authors reviewed 23 studie s that met several criteria: published in a peer-reviewed j ournal, included students from preschool to third grade, included students at risk for learning disabilities, contained interventions ta rgeting early literacy skills, and had study outcomes that a ddressed reading development. Within the 23 studies, the following areas were frequently examined to determine their correlation with student unresponsiveness: phonologica l awareness, phonological memory, rapid naming, intelligence, attention or behavior, orthographic processing, and demographics. Seventy percent of the studies found phonological awareness to have an apparent link to treatment unresponsiveness and s ubsequent reading difficulties. The other variables were found less frequently and mo re inconsistently.

PAGE 20

13 Additionally, Fletcher and colleagues (Fletcher, Shaywitz, Shankweiler, Katz, Liberman, Stuebing, Francis, Fowler, & Shaywi tz, 1994) examined nine variables related to reading proficiency among students with an identified learning di sability in reading (defined by a discrepancy between abili ty and achievement on norm-referenced assessments), students with low achievement in reading (defined by ability potential at or above the 9th percentile and achievement in reading below the 25th percentile on normreferenced assessments), and average reader s. One hundred ninety-nine students aged 7.5 to 9.5 years were included in this study. Results indicate d that children in both the learning disability group and the low achieveme nt group were consistently more impaired in phonological awareness than children who were identified as average readers. Instruction in Phonological Awareness Specific training in phonological awarene ss, either before reading instruction begins or during reading instruction, consiste ntly accelerates reading growth for children who receive it (Torgesen & Mathis, 1999). The National Reading Panel concluded that phonemic awareness instruction is most eff ective when it is taught explicitly, lasts between five and eighteen hours total (with av erage 25-minute sessions), and is taught in small groups (National Institute of Child Health and Human Development, 2000). The National Research Council Committee on Pr eventing Reading Difficulties in Young Children also recognized the importance of th is skill and stated in their report that Explicit instruction th at directs childrens attention to the sound structure of oral language and to the connections between speech sounds and spellings assists children who have not grasped the alphabetic princi ple or who do not apply it productively when they encounter unfamiliar printed words (Snow, 1998, p. 6).

PAGE 21

14 Brennan and Ireson (1997) examined the effects of formal and informal phonological awareness training programs on kindergarten students. Thirty-eight kindergarten children (mean age five years, four months) divided into three classrooms were participants in the study. Students in one class receiv ed a systematic and explicit phonological awareness training program as a part of their literacy curriculum; one class used a structured program designed to stimul ate reading and writing skills that did not contain explicit phonological aw areness activities; and a third control group used letter characters to teach the ch ildren letter names and sounds with no explicit phonological awareness activities. These programs were in corporated into the curriculum for eight months over the span of the kindergarten ye ar. Pretests were administered at the beginning of the school year and posttests were administered in May. An analysis of variance was conducted to determ ine if there were any sign ificant differences between the groups on any of the pretest items and none were found. The experimental group, which received the explicit, systematic phonological awareness training, achieved a significantly higher mean score on a word reading assessment, p < .05; word segmentation, t = 2.08, p < .05; syllable segmentation, t = 2.09, p < .05; deletion of initial phoneme, t = 3.88, p < .0005; phoneme segmentation, t = 4.85, p < .0005; and phoneme synthesis, t = 4.41, p < .0001 when compared to the informal phonological awareness training group. Additionally, several studie s (Byrne & Fielding-Barn sley, 1995; Kozminsky & Kozminsky, 1995; Schneider, Kuspert, Roth & Vise, 1997) have demonstrated longitudinally that children who receive expl icit phonological awareness instruction prior to or during the early stages of formal r eading instruction outperform children who do not

PAGE 22

15 receive explicit instruction. Byrne and Fiel ding-Barnsley (1995) ex amined the long-term effects of an explicit phonemic awareness trai ning program on later reading skills. The 64 students in this study were trained during preschool in sm all groups (i.e., 4-6 students) for 25-30 minutes once per week for 12 weeks. The training program focused on identifying pictures with an initial or ending sound, listening to jin gles or poems with a particular phoneme repeated in an initial or ending position of a word, identifying and coloring pictures that began or ended with a particular sound, and playing card games based on identifying similar in itial and ending sounds. A control group of 64 preschool students also was included in the study. The students were assessed at the end of first grade in the areas of word identification, spelling, alphabet naming, and phoneme identity. Sixty-four of the original 64 expe rimental group students and 54 of the original 64 control group students were assessed at the end of first grade. By grade two, 62 students in the experimental group and 53 students in the control group remained. Students in grade two were assessed on (a) na ming numbers, (b) reading number words, (c) reading pseudowords, regular words, and ir regular words, (d) print exposure, and (e) listening and reading comprehension. Byrne and Fielding-Barnsley (1995) found th at at the end of first grade, the experimental group significantly outperformed the control group on measures of pseudoword reading, t (22) = 3.40, p < .01. Marginal differences also were noted on the measure of reading regular words, t (22) = 1.63, p = .06. No differences were found between the two groups on measur es of irregular word readin g or spelling. Scores for students across both groups reached near ceil ing levels for the measures of alphabet naming and phoneme identity. Therefore, di fferences were not found between the two

PAGE 23

16 groups on these measures due to restriction of range. At the end of second grade, the experimental group performed significantly better on the measures of pseudoword reading, t (22) = 2.84, p < .01. Additionally, upon further analysis, there was a larger difference between the two groups on two-sy llable pseudowords than on one-syllable pseudowords. There were no differences be tween the groups on a measure of regular word reading; however, further analysis rev ealed that the experimental group performed better on the least common words from the word list. The rationale behind the relevance of this analysis was that the words that are less common are more likely to require decoding skills as opposed to sight word recognition. The experimental group also outperformed the control group on the measure of reading comprehension, t (22) = 1.73, p = .05). Further evidence for the positive longitudinal effects of explicit phonological awareness training was demonstrated in a study by Kozminsky and Kozminsky (1995). Their training program was implemented with 35 children during their kindergarten year. The eight-month training program focused on listening, identifying and creating rhymes, breaking sentences into words, breaking wo rds into syllables and phonemes, blending phonemes and syllables into words, counti ng syllables and phonemes in a word, and phoneme manipulation. The activities include d individual, small group, and large group activities and ranged between 90 and 120 minutes per week. The activities were completed by both the classroom teacher and a research assistant. A control group of 35 students also was included in the study. Th e students in the experimental and control groups were assessed on their phonological awareness skills with the Lindamood Auditory Conceptualization (LAC) test a nd the Phonological Aware ness Test (PAT).

PAGE 24

17 The LAC was administered at the beginning of the kindergarten year, at the end of the kindergarten year, and at the end of first grad e. The PAT was administered at the end of the kindergarten year followi ng the training program. Partic ipants were assessed on their reading comprehension skills using the Read ing Comprehension Test, which is a groupadministered multiple choice test that incl udes tasks ranging from matching words and pictures to reading 200-word passages and answering comprehension questions. This assessment was administered to both groups at the end of first grade and the end of third grade. Kozminsky and Kozminsky (1995) f ound that the experimental group outperformed the control group on the Phonologi cal Awareness Test at the end of the kindergarten year, t (59) = 3.35, p < .05. No significant difference was found between the experimental and control groups at the end of kindergarte n on the Lindamood Auditory Conceptualization test. However, the mean scores of the experimental group were 0.81 standard deviations higher than the control group on this assessment at the end of first grade ( M = 51.79, SD = 16.22 and M = 37.38, SD = 16.69, respectively). The measure of reading comprehension also indicated significantly better performance by the experimental group at the end of both first and third grade, F (1,45) = 6.00, p < .05. Similarly, Schneider, Kuspert, Roth and Vise (1997) completed two longitudinal studies exploring the effects of phonol ogical awareness training on phonological awareness, early literacy, reading, and spel ling skills. Two hundred five kindergarten children participated in the training group, and 166 children made up a control group. The average age for all the children was five years, seven months. All children were assessed at the beginning of their kindergarte n year on general intell ectual ability, early

PAGE 25

18 literacy skills (i.e., letter knowledge, reading of real words and pseudowords), and phonological processing skills (i .e., phonological awareness, phonological memory, and rapid naming). The same measures were ad ministered during the kindergarten year following training as a posttest. The students also were assess ed at the beginning of first grade on metalinguistic transfer. These te sts consisted of task s including matching pictures based on similar beginning a nd ending sounds, segmenting words into phonemes, identifying word length, and phoneme manipulation. Reading and spelling were assessed at the end of first and sec ond grade. The reading assessment measured both decoding and comprehension; and the spel ling assessment measured both frequently used words and rare, irregular words. Th e seven-month training program included daily practice in metalinguistic games that lasted fo r 15-20 minutes and were instructed by the classroom teacher. The training program incl uded instruction on listening, identification of rhymes, syllable segmentation and analysis, identification of initial phonemes, and phoneme analysis and synthesis. Results indicated that when the th ree phonological awareness tasks were combined into a single factor, the trai ning group presented with better phonological awareness skills than the control group on the posttest administered at the end of the training group (Schneider et al., 1997). Significant differe nces were not found between the groups on the measures of metalinguistic transfer that were administered at the beginning of first grade. The researchers hypothesized that there was not consistency among the training groups and analyzed the records kept by the teachers who were implementing the training program. This an alysis revealed that only nine of the 22 teachers implementing the training program di d so with consistency and completed the

PAGE 26

19 entire program. The other 13 teachers be gan the training program consistently and subsequently fell behind schedule and did not finish the program to its completion. The researchers reanalyzed the me talinguistic transfer data to compare differences between the consistently trained and inconsistently trained groups. They found that there were significant differences between the consiste ntly trained group a nd control group on the initial phoneme, word length, a nd phoneme analysis tasks ( p < .05); and no differences between the inconsistently trained students and the control group. Additionally, the consistently trained group pe rformed better on measures of reading and spelling at the end of first grade (p < .05). This same effect did not remain true for the measures of reading and spelling at the end of second grade, on whic h there were no significant differences between any of the groups. Schneider et al. (1997) initiated a similar, second study in order to address some of the limitations found within the first study. The study procedure was generally the same with minor adjustments to improve the integrity of the traini ng program. The time required of the training was decreased to ten minutes to assist with helping the teachers fit the program into their daily routine. A dditionally, members of the research team met with the teachers at least one time per week to provide feedback and discuss strategies for the more difficult units to implement. Results of the second study were similar to the first in that phonological awaren ess tasks measured during ki ndergarten before and after the training program indicated a significant difference in favor of the training group ( p < .01). Mean differences between the training and control groups were larger in the second study. The findings in the second study also were different from the first with regard to the metalinguistic transfer assessment at the be ginning of first grade. Unlike the previous

PAGE 27

20 study, there were significant differences be tween the training gr oup and control group, with the training group performing better on most tasks ( p < .05). Subsequently, differences between the two groups also were found on later measures in first and second grade in reading and spelling. In addition to adding to the research base demonstrating the long-term effects of explicit phonological awareness traini ng, this study also verified the importance of consistency and integr ity when implementing a training program. Phonological Awareness Intervention Despite the research indicating the be nefits of including phonological awareness as a regular part of the reading curriculum, not all schoo ls have incorporated explicit phonological awareness training a nd anticipate that students will develop these skills without direct instruction. Although some students do acquire these skills, others do not and require intervention and remediation in this area. Some studies have explored the effect s of teacher-led phonological awareness training on students who are low in phonol ogical awareness skills. For example, Torgesen, Morgan, and Davis (1992) selected 51 students in their second semester of kindergarten from a pool of 143 students in seven kindergarten classes based upon their low scores of the Screening Test of Phonological Awareness (STOPA) and completed various training programs intended to im prove phonological awareness skills. The students in the study were administered pretests measuring phoneme segmentation, phoneme blending, alphabetic reading, and gene ral verbal ability. The students were randomly assigned to one of three groups (t wo experimental trai ning conditions and a language-experience training condition) that were matched by age and verbal ability. The students were placed into groups of thr ee to five and participated in 20-minute

PAGE 28

21 training sessions three times per week for se ven to eight weeks. One training group (AB) engaged in activities designed to teach segm enting and blending skills. A second training group (B) engaged in activities designed to teach blending sk ills only. The control group (C) engaged in activities that emphasized reading enjoyment. The segmenting and blending tests were administered again as posttests. Additionally, a reading analogue task was given to all of the students at the conclusion of the training sessions. This task taught the participants letter sounds and a letter-like symbol that represented each one. Once the associations were learned, the st udents were then asked to use the newlylearned symbols to read consona nt-vowel-consonant words. Torgesen, Morgan, and Davis (1992) f ound that on the segmenting task, the segmenting and blending group (AB) significan tly outperformed the control group, t (9) = 4.16, p < .05. There were no significant diffe rences between the blending group and control group on this task. Both tr aining groups (AB & B) significantly outperformed the control group on the blending task, t (9) = 4.11, p < .05 and t (9) = 6.05, p < .05, respectively. When analyzing the information from the reading analogue task, the authors looked at the data for errors and number of trials to reach criterion. There were no significant differences between the gr oups on the number of errors. There were, however, significant differences between th e AB training group and the control group on the number of trials, with the control group taking more than twice the number of trials to meet criterion than the AB training group (18.7 trials and 7.0 trials, respectively). Although this study indicated that phonologica l awareness training is an effective intervention for students with skill deficits in this area, many of these training programs are designed to be taught in small groups and require significant teacher time and

PAGE 29

22 resources that are not always available. In schools and classrooms where students have deficiencies in phonological awareness a nd resources do not allow for small group intervention, other alternatives need to be explored. Computer Instruction in Phonological Awareness Computer technology has grown tremendous ly in our education system and advances have lead to software development aimed at improving students reading skills. The number and types of computer interv ention programs designed to train phonological awareness have increased during the last d ecade. The following section will explore the various studies that have been completed to examine the effec tiveness of computer programs in improving phonological awareness skills. Computer programs in general have been shown to be an effective addition to enhancing phonological awareness skills in a comprehensive reading program. For example, Hecht and Close (2002) conducted a study with kindergarten students to examine the effectiveness of a computer-assi sted phonemic awareness training program. The students were pre-tested in the fall of their kindergarten year with assessments measuring phonemic segmenting, phonemic ble nding, letter name knowledge, letter sound knowledge, letter writing knowledge, word reading, invented spelling, vocabulary knowledge, and concepts about print. The students in the experimental group were exposed to the Waterford Early Reading Program (WERP-1) as a regular part of their classroom literacy curriculum. This pr ogram provides instruction in phonological awareness skills, letter knowledge print concepts, and oral la nguage skills. Students in the control group were selected from a di fferent school and did not receive specific phonological awareness training. The amount of time that the students were exposed to

PAGE 30

23 the WERP-1 was not controlled; however it was monitored through the teacher management system that includes daily usage in time for each student. Posttests were administered in the spring of the students kindergarten year. Approximately six months elapsed between the preand posttests. Th e results indicated that the students who received the WERP-1 computer-assisted inst ruction performed signi ficantly better than the control group on the phonemic awareness po sttests when all pre-tested emergent literacy skills were controlled ( p < .001). The students in the control group, however, made no significant improvements in their phon emic awareness skills. This study also indicated that the time that the students were exposed to the WERP-1 was associated with growth in phonemic awareness skills when controlling for in itial knowledge of emergent literacy skills. Computer-assisted instruction in phonological awareness has been shown to be effective across a variety of age levels. In a study by Foster et al. (1994), the effectiveness of a preliminar y version of DaisyQuest, a co mputer program designed to facilitate phonological awareness, was evaluate d with preschool students. Twenty-seven students were selected out of 100 based on scores on the Phonological Awareness Test (PAT) and the Peabody Picture Vocabulary Te st Revised (PPVT-R), a measure of general verbal ability. Students with scores above 20 (out of a possible 30) on the PAT or a standard score below 75 on the PPVT-R were not included in the study. The 27 students were randomly assigne d to an experimental ( n=12) or control ( n =15) group. There were no significant differe nces between the two groups w ith regard to age, scores on the phonological awareness test, or vocabulary test. The two groups also were administered an additional phonological awareness test, the Screening Test of

PAGE 31

24 Phonological Awareness Experimental Versio n (STOPA-E) as a pretest. The version of DaisyQuest evaluated in this study ta ught skills of recognizing rhyming words; recognizing words with the same beginni ng, middle, and ending sounds; recognizing words segmented into phonemes; and counti ng the number of sounds in words. The preschool students used the program for 2025 minutes across 20 sessions. The PAT and STOPA-E were administered again as postte sts. Approximately one month elapsed between preand posttests. Results indicated that the children in the experimental group obtained significantly higher posttest scores on both the PAT and the STOPA-E, when adjusted for pre-test PAT and STOPA-E scores, F (1,21) = 6.1, p < .02 and F (1,21) = 14.4, p < .001, respectively. A study completed by Lonigan and colleagues (Lonigan, Driscoll, Phillips, Cantor, Anthony, & Goldstein, 2003) also examin ed the effects of a computer-assisted phonological awareness training program with pr eschool students. This study expanded upon the previous study in that it targeted stude nts at risk for edu cational difficulties who were attending Head Start. Forty-five childre n participated in the study and were initially assessed on their oral language and genera l cognitive abilities, print knowledge, and phonological sensitivity skills. Students were randoml y assigned to either the experimental group ( n = 22) or the control group ( n = 23). The experimental group used DaisyQuest and Daisys Castle, two comp lementary phonological awareness computer programs, for eight weeks. Children worked on the programs for 15 to 20 minutes, four to five days per week. Following the training, 41 students were administered a vocabulary assessment, print knowledge task s, and phonological sensitivity tasks. Overall, children who participated in the computer-assisted instruction group displayed

PAGE 32

25 more growth in phonological se nsitivity skills than the control group. Specifically, these results were significant for the Rhyme O ddity, Rhyme Matching, Word Elision, and Syllable/Phoneme Elision tasks (all p s < .05). Foster et al. (1994) repeated their study with kindergart en students in their second semester. A sample was selected from four kindergarten classes based on a vocabulary measure. Students with the highest and lowest scores were taken out of the sample in order to create a more homogeneous group. The students were then matched according to their scores and randomly assigned to eith er the experimental group (n=34) or control group (n=35). The students were given a co mputerized test calle d Undersea Challenge (which assesses phonological awareness), th e Screening Test of Phonological Awareness (STOPA), the Production Test of Segmenti ng, and the Production Test of Blending as pretests. The students in the experiment al group used the DaisyQuest program, a computer software program designed to tr ain phonological awareness, for 16 20-minute sessions daily. The same tests that were ad ministered at the beginning of the training were administered again as posttests. At th e end of the training period, the experimental group obtained significantly higher posttest scores than the control group on the Undersea Challenge, F (1,66) = 13.1, p < .01, the Production Test of Segmenting, F (1,66) = 57.8, p < .01, and the Production Test of Blending, F (1,66) = 29.4, p < .001. No differences occurred between the groups on the STOPA. Th e experimenters felt that a ceiling effect (most students in both groups obtained a perf ect or near perfect score) may have attributed to this lack of difference on the STOPA. Finally, Barker and Torgesen (1995) exam ined the effects of computer-assisted phonological awareness inst ruction on a sample of first gr ade at-risk students. Students

PAGE 33

26 in this study were nominated by their teach ers and then administered assessments of phonological awareness skills and basic word recognition skills. Fifty-four students participated in the study based on the scr eening process and were administered the following pretests: Undersea Challenge, sound categorization, phoneme elision task, production test of segmenting, pr oduction test of blending, Wo rd Analysis subtest from the Woodcock-Johnson Reading Mastery Te st, non-word reading task, Word Identification subtests from the Wood cock-Johnson Reading Mastery Test, analog reading task, and the Vocabulary subtest from the Stanford-Binet IV Revised. The 54 students were randomly assigned to one of three conditions including a phonological awareness training group using the DaisyQuest and Daisys Castle programs, a phonological decoding training control group usi ng the Hint and Hunt program, and the attentional control group usi ng several math-oriented computer software programs. DaisyQuest and Daisys Castle are two co mplementary computer software programs which include the following activities: rec ognizing words that rhyme; recognizing words that have the same beginning, middle, and ending sounds; blending onsets and rimes; blending phonemes; and counting the number of sounds in each word. The Hint and Hunt program provides practi ce with basic short vowel sounds and the math-oriented attentional control group software programs were designed to provide practice in basic addition and subtraction skills. Once the training period began, the student s used the computer programs for 25minute sessions four days per week (Barker & Torgesen, 1995). All of the measures that were administered as pretests were administ ered again as posttests with the exception of the Stanford-Binet IV Vocabulary test. The experimental phonological awareness

PAGE 34

27 training group made signif icant gains on the measures of Undersea Challenge, F (1,46) = 14.02, p < .001, segmenting, F (1,46) = 25.90, p < .001, and elision, F (1,46) = 9.60, p < .003. When comparing group differen ces, the phonologica l awareness group significantly outperformed th e phonological decoding cont rol group on the Undersea Challenge ( t = 32.5, p < .002), segmenting ( t = 4.55, p < .001), and elision measures ( t = 2.71, p < .009). When analyzing reading outcomes, the phonological awareness group significantly outperformed the phonological decoding control group on the Word Identification subtest ( t = 3.41, p < .001). A more specific use of computer programs was assessed by Pokorni, Worthington, and Jamison (2004). These authors conducted a study to determine whether computer programs were as effective as teacher intervention when addressing students who have special needs. Two com puter programs (Fast ForWord and Earobics Step 2) were chosen for the study due to their focus on phonological awareness and their publishers claims that students will experi ence dramatic improvements in language and reading skills following their use. The study attempted to assess whether one or more of the three intervention programs results in greater gains than the other programs in phonemic awareness, language, or reading-relate d skills and if students in individual intervention groups made gains in phonemic awareness, language, or reading-related skills. Sixty students from a large school dist rict between the ages of seven years, six months to nine years of age were selected for the study. All th e participants were receiving school-based speech/language services outlined in an IEP, were reading more than one year below grade level, were from English speaking families, and were not known to have a hearing impairment.

PAGE 35

28 Pokorni, Worthington, and Jamison (2004) administered pretest measures four to six weeks before the intervention began and in cluded a hearing screen ing, two subtests of the Phonological Awareness Test (Phoneme Blending and Phoneme Segmentation raw scores), three subtests of the Clinical Evaluation of Language Fundamentals 3 (Concepts and Directions, Recalling Sentences, Li stening to Paragraphs standard scores), and four subtests of the Woodcock Language Proficiency Battery Revised (Letter-Word Identification, Passage Comprehension, Word Attack, and Spelling standard scores). Students were randomly assigned to one of three intervention groups. The first intervention group used the Fast ForWord (FFW ) program that was designed to target various language and reading skills. The second interventi on group utilized the Earobics Step 2 program, which provides instructi on in auditory and phonological processing skills. The third intervention group was instructed in the Lindamood Phonemic Sequencing Program (LiPS), a small group te acher-led intervention program focusing on awareness of sound, oral-motor features of sounds, sound discrimination for consonant and vowels, and beginning readi ng and spelling tasks. Each group received three hours of intervention daily for 20 days during a summer program with breaks provided including a mid-day lunch break. Computer in terventions were provided with individual computers and headphones. The LiPS program wa s provided in small groups of four (the CD-ROM exercises were not used during this study). Pokorni, Worthington, and Jamison (2004) ad ministered posttest measures (same measures as pretest) six to eight weeks after the intervention ended. Data were analyzed on 54 of the 60 students. In this study, the intervention groups did not differ substantially in their improvement on segmenting phonemes, language subtests, or reading-related

PAGE 36

29 subtests. The students in the LiPS intervention performed signifi cantly better than the other two groups on blending phonemes, F (2, 51) = 6.42, p < .01. Within group differences indicated that the Earobics group improved significantly in segmenting phonemes (Cohens d = 1.66) and the LiPS group improved significantly in segmenting (Cohens d = 1.33) and blending phonemes (Cohens d = 2.32). Significant improvements were not found on any of the measures with the Fast ForWord group. None of the programs were associated with significant tran sfer effects to language or reading. Mitchell and Fox (2001) also examined the comparison of computer-delivered and teacher-led instruction in their study. Unlik e the students in the Pokorni et al. (2004) study, these students were not re ceiving special education servi ces and were identified as at-risk for reading failure. In this study, 36 kindergarten a nd 36 first grade students from six classrooms at each grade level were selected from a middle class, suburban elementary school. Students were identified as at-risk based on teacher report of low reading ability, receptive voca bulary measured by the Peabod y Picture Vocabulary Test Third Edition, and performance on the Literacy Initiative for Everyone (LIFE), which was a district-designed informal reading i nventory. Students who were referred by their teachers and were below district expectati ons on the LIFE were administered the PPVTIII. Participants were then randomly sele cted from among the students who received a standard score of 85 or a bove on the PPVT-III. Mitchell and Fox (2001) administered pret est measures including the LIFE (a district designed informal reading inventory that assesses lett er knowledge, print concepts, phonological awareness, sight word recognition, writing vocabulary, text

PAGE 37

30 reading level, and comprehension of gene ral verbal ability) and the Phonological Awareness Test (a measure of phonological proc essing). The interven tions consisted of computer-administered instruction in phonol ogical awareness (DaisyQuest and Daisys Castle), a technology control group (drawing and mathematics software), and teacheradministered instruction in phonological awareness (Phonological Awareness Kit and Phonological Awareness Intermediate Kit). St udents were randomly assigned to one of three experimental conditions ( n=24 ) with equal numbers ( n=12 ) of kindergarten and first grade students in each group. Students received a practice session followed by five hours of instruction in 20-mi nute sessions spread over a f our-week period. No other direct phonological awareness inst ruction occurred as a regular part of the curriculum. All treatments occurred outside the classroom (A & C in computer lab; B in resource room). Teacher-directed instruction occurred in two groups of six students. Similarly, the computer-assisted instruction groups were divided into groups of six and were allowed to select the programs they wished to use and choose their own learning goals. Experimenter interaction with the students was limited to answering questions about the program operation and assisting with accessi ng and opening the software. Posttest measures included four subtests of th e PAT (Rhyme Discrimination and Production, Phoneme Isolation, Phoneme Segmentation, and Blending). Following the training period, Mitchell a nd Fox (2001) found th at the computerassisted instruction group showed a significant difference when compared to the technology control group on Phoneme Isolation ( p < .001) and Blending ( p = .02) and approached significance on Phoneme Segmentation ( p = .06). The teacher-delivered instruction group demonstrated significant differences compared to the technology

PAGE 38

31 control group on Rhyme Disc rimination and Production ( p = .04), Phoneme Isolation ( p < .001), Phoneme Segmentation ( p = .001), and Blending ( p = .02). There were no significant differences found between the teacher-delivered instructional group and the computer-administered instructional group. A factor analysis was conducted to determine if the four subtests measured the same construct. The correlation between the scores on each individual posttest and the cumulative score, ranging from .57 to .86, indicated that all four test s do measure the same construc t, operationally defined as phonological awareness. The re sult of the analysis of covariance of the combined adjusted posttest scores showed a signif icant difference among the three treatment groups. There was no significant interaction e ffect between grade level and treatment. Both the teacher-delivered in struction group and the comput er-administered instruction group had significantly higher to tal phonological awareness scor es than the instructional technology control group. In summary, the literature supports the strong connection between phonological awareness skills and reading achievement. Students who lack phonological awareness skills are more likely to have reading di fficulty. Additionally, phonological awareness is a skill that can be identified as early as kinde rgarten as an area to target for remediation to prevent reading failure. This early id entification and intervention is essential considering the relatively s hort window of time availabl e to implement meaningful interventions. Although the positive eff ects of phonological awareness training have been documented, this is not a sk ill area that is consistently included as an area of explicit instruction in all early readi ng curricula. Given that all students do not acquire these skills without direct instruct ion, intervention methods need to be explored. While

PAGE 39

32 research has shown that teacher-led, smallgroup interventions in phonological awareness are effective, the research on the eff ectiveness of computer-based phonological awareness programs is more limited. T echnology advances and benefits have substantially increased the use of computer instruction by many school systems; therefore, the utility of this technology needs to be examined. The Current Study The purpose of the current study was to i nvestigate the effect s of two computer software programs on the early literacy skills of students at risk for reading failure. This study also aimed to provide information about whether the effects of the two programs were influenced by a students grade in school. Specifically, this study addressed the following research questions: 1. Are there outcome differences in kinderga rten students early literacy skills among two computer-delivered phonological awarene ss training programs (Earobics Step 1 and Lexia Early Reading) and a control group? 2. Are there outcome differences in first grad e students early literacy skills among two computer-delivered phonological awareness training programs (Earobics Step 1 and Lexia Early Reading) and a control group? 3. What are the effects of two compute r-delivered phonological awareness training programs (Earobics Step 1 and Lexia Early Reading) on the development of early literacy skills in students at risk for reading failure? 4. Are there outcome differences in early lit eracy skills between kindergarten and first grade students among two phonological awareness training programs (Earobics Step 1 and Lexia Early Readi ng) and a control group?

PAGE 40

33 Chapter Three Method Participants The participants in the study were 39 ki ndergarten and 37 first grade students who were identified as at-risk for reading difficulties as measured by the Dynamic Indicators of Basic Early Literacy Skills. Of the 39 kindergarten participants, 26 (67%) were male and 13 (33%) were female. Kindergarten partic ipants were represented in the following ethnic groups: 15% Asian, 3% African Amer ican, 36% Hispanic, and 46% Caucasian. Of the 37 first grade participants, 18 (49%) we re male and 19 (51%) were female. First grade participants were represented in the following ethnic groups: 3% Asian, 9% African American, 39% Hisp anic, and 49% Caucasian. The elementary school where this study o ccurred is located in a large school district serving approximately 114,466 PreK 12 students located in the Southwest region of Florida. The elementary school has a total kindergarte n through fifth grade student population of 722. Students in the school represent the fo llowing ethnic groups: 60% Caucasian, 7% African American, 19% Hispanic, 8% Asian/Paci fic Islander, < 1% American Indian/Alaskan Native, and 5% Mu ltiracial. Additionally, approximately 73% of the students in this school are eligible for free or reduced lunch. Measures Early literacy skills were measured us ing the Dynamic Indicators of Basic Early Literacy Skills (DIBELS). Additionally, an acceptability survey was used with both

PAGE 41

34 teachers and students participating in the st udy to obtain their input on the utility and satisfaction of using the computer programs. DIBELS are individually administered standardized measures designed to measure phonemic awareness, alphabetic principle, and fluency with connected text (Good & Ka minski, 2002). The DIBELS measures were used as a screening to iden tify students who were target ed for this study. DIBELS measures also were administered at the be ginning and end of the intervention period as preand post-tests. In addition, DIBELS measures were administered once per week during the intervention pe riod to monitor individual studen t progress. In this study, three DIBELS measures, Initial Sounds Fluency (ISF), Letter Naming Fluency (LNF), and Phoneme Segmentation Fluency (PSF), were us ed for preand post-tests as well as to monitor the individual progre ss of the kindergarten student participants. Four DIBELS measures, Letter Naming Fluency (LNF), Phoneme Segmentation Fluency (PSF), Nonsense Word Fluency (NWF) and Oral Readin g Fluency (ORF), were used as preand post-tests as well as to monitor the indi vidual progress of the first grade student participants. The following section describes each DIBELS measure in detail, including evidence of reliability and va lidity. The teacher and student acceptability surveys are described following the DIBELS descriptions. Letter Naming Fluency (LNF) Letter Naming Fluency (LNF) is a measure of letter recognition. This task requires the student to oral ly identify upperand lowercase letters presented to them in random order on a sheet of paper. The student names as many letters (out of 120) as he or she can in one minute with the examiner providing the name if the student hesitates for three seconds The score is calculated by the number of correctly-named letters in one minute. LN F takes one minute to administer with an

PAGE 42

35 additional 15 seconds to deliver instructions. One-month, alternate-form reliability in kindergarten is .88. Criterion-related valid ity in kindergarten with the WoodcockJohnson Psycho-Educational Battery Revise d Readiness Cluster is .70. Predictive validity coefficients of LNF in kindergarten with Curriculum Based Measurement reading and the Woodcock-Johnson Psycho-E ducational Battery Revised Reading Cluster in first grade are .71 and .65, respectively (Kaminski & Good, 2002). Initial Sounds Fluency (ISF) Initial Sounds Fluency (ISF) is a measure of phonemic awareness. ISF evaluates a stude nts ability to recognize and produce the initial sound in a word given orally. Students are presented with four pictures that are named by the examiner. The examiner then asks the student to identify the picture that begins with a sound presented orally by the exam iner. The student is also asked to orally provide the initial sound in a word presented orally by the examiner. There are 16 items total on this measure and it can be administer ed in approximately three minutes. The score is calculated by totaling the amount of time that it takes the student to identify or produce the correct sounds a nd converting it into the number of correct onsets in a minute. There are over 20 alternate form s of this measure available for progress monitoring. Alternate-form reliability of I SF is .72 in January of kindergarten. The concurrent criterion-related validity of ISF with Phoneme Segmentation Fluency in January of kindergarten is .48. The concurrent criterion-related valid ity of ISF with the Woodcock-Johnson Psycho-Educational Batt ery Readiness Cluster score is .36. Additionally, predictive validity coefficients of ISF with spring first grade Oral Reading Fluency and the Woodcock-Johnson Psycho-Edu cational Battery Total Reading Cluster are .45 and .36, respectively (Good, Laimon, Kaminski, & Smith, 2002).

PAGE 43

36 Phoneme Segmentation Fluency (PSF) Phoneme Segmentation Fluency (PSF) is also a measure of phonemic awareness and a ssesses the students ability to segment threeand four-phoneme words into individua l phonemes or sounds. In this task, the student is given a word and asked to provide the individual phone mes, or sounds, that make up the word. Words are continuously presented for one minute. The score is calculated by how many phonemes the student correctly segments in one minute. Including instructions, this measure can be administered in approximately two minutes and over 20 alternate forms are available fo r progress monitoring. Two-week, alternateform and one-month, alternate-form reliab ility are .88 and .79, respectively in May of kindergarten. Concurrent criterion-related validity in spring of kindergarten with the Woodcock-Johnson Psycho-Educational Batt ery Readiness Cluster score is .54. Predictive validity of kindergarten spring PSF is .62 with both first grade winter Nonsense Word Fluency and first grade spring Oral Reading Fluency. Similarly, predictive validity of kindergarten sp ring PSF with the Woodcock-Johnson PsychoEducational Battery Tota l Reading Cluster is .68 (Good, Kaminsky, & Smith, 2002). Nonsense Word Fluency (NWF) Nonsense Word Fluency (NWF) is a measure of the alphabetic principle. In this meas ure, the student is pr esented with randomly ordered Vowel-Consonant and Consonant-V owel-Consonant nonsense words on a sheet of paper and asked to produce either the indi vidual sounds or the whole nonsense word. The child has one minute to produce as many le tter-sounds or words as he or she can. Including instructions, this m easure can be administered in approximately two minutes. Over 20 alternate forms are available for pr ogress monitoring. Concurrent, criterionvalidity of NWF with the Woodcock-Johns on Psycho-Educational Battery Readiness

PAGE 44

37 Cluster score is .36 in January of first grade a nd stronger in February of first grade at .59. Predictive validity of NWF in January of firs t grade with Oral Reading Fluency in May of first grade and May of second grade is .82 and .66, respectively. Predictive va lidity of NWF in January of first grade with the Woodcock-Johnson Psycho-Educational Battery Reading Cluster score is .66. One-month, alternate-form reliability in January of first grade is .83 (Good & Kaminski, 2002). Oral Reading Fluency (ORF) Oral Reading Fluency is a measure of accuracy and fluency with connected text. The DIBELS Oral Reading Fluency is based on Curriculum-Based Measurement (CBM) of r eading. On this measure, students are presented with a passage calibrated at their grade level and asked to read aloud for one minute. Words that are mispronounced, omitted or substituted are scored as errors. Additionally, a hesitation of more than three seconds is scored as an error. Words are scored as accurate if the stude nt self-corrects within three seconds. Students are asked to read three passages and the students oral r eading fluency rate is the median correct words per minute from the three passages. Test-retest reliability for CBM reading for elementary students ranged from .92 to .97. Alternate-form reliability of various passages at the same grade level ranged from .89 to .94. Criterion-related validity coefficients ranged from .52 to .91 (Good & Kaminski, 2002). The DIBELS measures of Initial Sounds Fluency, Phoneme Segmentation Fluency, Nonsense Word Fluency, and Oral Read ing Fluency are indicative of a students acquisition of the big ideas in early reading (i.e., phonol ogical awareness, alphabetic principle, and accuracy and fluency with connect ed text). These early reading skills are developmental and the acquisition of earlier skills such as phonological awareness

PAGE 45

38 facilitates the acquisition of later skills su ch as alphabetic principle. Based on this, benchmark goals for each DIBELS measure ha ve been established to indicate whether students are acquiring the early r eading skills necessary to lead to later reading success. A study by Good, Simmons, and Kameenui (2001) examined the utility of the DIBELS benchmark goals and the relationship of ear lier benchmark goals to later benchmark goals. Results of this study i ndicated that students who met earlier benchmark goals were likely to meet later benchmark goals. Table 1 displays the DIBELS measures benchmark goals and the correlation coefficients for th e strength between the subsequent skills. Table 1 DIBELS Benchmarks and Correlation Coefficients DIBELS Measure Early Reading Skill Assessed Timeline Benchmark Goal Correlation with Next Benchmark Initial Sounds Fluency (ISF) Phonemic Awareness Winter, Kindergarten 25 Initial Sounds Correct per Minute .34 with PSF Phoneme Segmentation Fluency (PSF) Phonemic Awareness Spring, Kindergarten 35 Phonemes Correct per Minute .38 with NWF Nonsense Word Fluency (NWF) Alphabetic Principle Winter, First Grade 50 Letter-Sounds Correct per Minute .78 with ORF Oral Reading Fluency (ORF) Fluency with Connected Text Spring, First Grade 40 Words Read Correct per Minute .82 with ORF Spring 2nd grade Student ratings of computer training All students participating in the study were asked three questions (i.e., How much did you like to do the computer activities? How much did you like leavi ng your class to do the comput er activities? How much would you like to have the computer activit ies at home?) at the end of the study to obtain the students rating of acceptability and enjoyment w ith regard to the computer software programs. The students provided responses by choosing a sad face for Not at all (scored as 0), a neutral face for A littl e (scored as 1), or a happy face for A Lot

PAGE 46

39 (scored as 2). These questions were simila r to those posed to students in a computer assisted instruction study conducted by Loni gan, Driscoll, Phillips, Cantor, Anthony, and Goldstein (2003). See Appendix A for this survey. Teacher ratings of computer training All kindergarten and first grade teachers were asked to complete an acceptability surv ey regarding the two programs used by the students. The acceptability survey used in this study was adapted from the Intervention Rating Profile 15 (Martens, Witt, Elliott, & Darveaux, 1985). Psychometric characteristics for the IRP 15 were examined in an initial study of 60 participants and cross-validated in a second independent sample. Results indicated split-half reliability coefficients ranging from .95 to .98 and inte rnal consistency correlation coefficients ranging from .88 to .98 (Martens, Witt, Elliott, & Darveaux, 1985). The IRP 15 consists of 15 questions regarding the acceptability of treatment for behavioral symptoms. The questions on the IRP 15 were modified to reflect the academic nature of the treatment used in this study. The modi fied survey consists of 10 items that are rated on a six-point Likert-type scale (1 = st rongly disagree, 6 = strongly agree). See Appendix B for this survey. Procedure The principal investigator in this study is the school psychologist assigned to the elementary school where the study was conducte d. The principal investigator met with the principal of the elementary school to explain the study and obtain permission to conduct the study at the school. Once permissi on was obtained from the principal of the school, the principal investigator met individually with the sc hool reading coach and with the kindergarten and first grade teachers at th eir grade level team meetings. The purpose

PAGE 47

40 of the grade level team meeting was to explain the study and enlist the teachers assistance. The teachers agreed to assist with the study and consequently provided information for logistics (e.g., schedules, pref erred times for the students to leave their class, additional supplies that may be needed). The elementary school where the study o ccurred regularly participates in an ongoing assessment process for students in kinde rgarten through third gr ade as a part of a Reading First grant. Four times per year, students in these grades are assessed with the Dynamic Indicators of Basic Early Literacy Ski lls (DIBELS). The stude nt participants in this study were recruited based on the Fa ll (administered in September) round of assessments. During this assessment period, ki ndergarten students were assessed on two DIBELS measures (i.e., Initial Sounds Flue ncy and Letter Naming Fluency). First grade students were assessed on f our DIBELS measures (i.e., Letter Naming Fluency, Phoneme Segmentation Fluency, Nonsense Word Fluenc y, and Oral Reading Fluency). Based on these assessments, classroom teachers were pr ovided with class reports that included each students score on each DIBELS measure and instructional recommendations based on those scores. These instructional recomme ndations are based on d ecision rules developed by Good, Simmons, Kameenui, Kaminski, and Wallin (2002). Instructional recommendations fall into three categories: Benchmark At Grade Level, Strategic Additional Intervention, and In tensive Needs Substantial Intervention. Kindergarten and first grade students whose instructi onal recommendation for the Fall assessment period was Intensive Needs Substantial in tervention were recruited for this study. The Institutional Review Board at the Univ ersity of South Florida approved the current study on 09/21/05. Subsequently, th e principal investigator met with the

PAGE 48

41 kindergarten and first grade teachers at a regularly scheduled grade level team meeting. The teachers were asked to bring their Fa ll DIBELS assessment results. Forty-four kindergarten students and 40 first grade st udents fell within the Intensive Needs Substantial Intervention category on the Fall DIBELS assessment. Each classroom teacher was given a cover letter and parent al consent form (see Appendix C) to send home to the students parents in an envelope with the stud ent within the preferred homeschool communication method for that class (e .g., agenda book, homework folder). The students were offered an incentive (e.g., stic ker, pencil) to bring back the permission form signed. This incentive was given to th e student regardless of whether or not the students parent indicated that he or she was able to participate in the study. Thirty-two permission forms from kindergarten students and 27 permission forms from first grade students were returned within three days of when the initial permission forms were sent home. Subsequently, a second series of perm ission forms were give to teachers to send home with the students who did not return forms. Within four days, an additional nine kindergarten and 11 first grade students returned forms. The 41 kindergarten (i.e., 93%) and 38 first grade (i.e., 95%) students who were given permission to participate in the study we re administered the pre-test assessments (i.e., DIBELS ISF, LNF & PSF for kindergar ten students and LNF, PSF, NWF, & ORF for first grade students) by the principal i nvestigator and a member of the school-based DIBELS assessment team in a quiet area near the students classroom. Subsequently, the students were randomly assigned into either one of the two experi mental groups or the control group after matching subjects on age and teacher. One experimental group used the computer-based phonological awarene ss program Earobics Step 1. A second

PAGE 49

42 experimental group used the computer-bas ed phonological awareness program Lexia Early Reading. The control group received no specific interventi on designated by the study. The participating school emphasizes te aching through technology in its general curriculum; therefore, students are repeatedly exposed to co mputer programs and general technology use beginning in ki ndergarten. Consequently, th e novelty effects of using computer software programs were considered to be minimal and a technology control group was not utilized. A rotation schedule was developed by the principal investigator based on input given by the teachers (see Appendix D). The software programs were loaded on 14 computers with headphones in the computer la b at the elementary school. The computers in the lab were numbered; and each student wa s assigned to a computer that he or she used throughout the intervention period. Before the intervention period began, the principal investigator trained the participan ts in small groups of five on the relevant software with regard to initiating and pr oceeding through the program and navigating the mouse (see Appendix E for training checklist). Students were requir ed to pass at least five out of six areas on the checklist before beginning the intervention. Both computer programs required extensive use of the mous e throughout the activitie s; therefore, the students were required to pass the task Use mouse to navigate activity before they began the intervention. The students were divided into four groups that alternated into the computer lab according to the rotation schedule. The prin cipal investigator and a teacher assistant monitored the students each day during their tr aining in the computer lab (see Appendix F for Proctor Schedule). The principal investig ator trained the teacher assistant who also

PAGE 50

43 was provided with all necessary materials, including the rotation schedule and proctor schedule. Students were picked up from class, ta ken to the computer la b, directed to their assigned computer, and assisted, if needed, in signing into the program. The proctors set a timer for 20 minutes, monitored the students wh ile they used the computer activities to assure that they were engaged in the program, and assisted with procedures as needed. At the end of the 20 minutes, the proctor in structed the students to sign off of their respective programs and returned them to class. This cycle was repeated three times until all four groups completed their time on the computer program. The students used the intervention 20 minutes daily for 25 days, resu lting in a total of eight hours, 20 minutes of exposure to the intervention. Student participation was monitored for each student on individual activity cards (see Appendix G). At the end of each training session, students received a sticker on their activity card and the date of the trai ning session was noted. Each Friday, the students were able to select a token of appreciation (e.g., eras er, pencils, stickers, small trinket) from a treasure box provided by the principal investigator Additionally, each Friday, after the regular traini ng sessions were complete fo r all groups, the proctor called in students who were absent during that week for makeup sessions. Additionally, two days were included at the end of the five -week intervention period for additional makeup sessions. Progress monitoring data were collected fo r all students in both experimental groups and the control group one time per w eek throughout the five-week intervention period. The data consisted of the same DIB ELS measures used for preand post-tests (i.e., ISF, LNF & PSF for kindergarten; LNF, PSF, NWF, & ORF for first grade). Data

PAGE 51

44 were collected by the principal investigator and a school-based member of the DIBELS assessment team. Members of the school-based DIBELS assessment team complete a six hour training conducted by distri ct level facilitators. The school-based reading coach assisted with ensuring the fidelity of the DIBELS assessment by observing both data collectors using the Assessment Integrity Checkli st for each DIBELS measure. Each data collector was observed on each measure tw ice with feedback given immediately following the observation. The school-based reading coach was trained in DIBELS assessment and to be a facilitator by a state le vel Master Trainer from the Florida Center for Reading Research. The post-test DIBELS measures were collected following the conclusion of the intervention period. At that time, the students also completed the student survey with the administrator of the pos t-tests. The teacher survey also was given to each teacher at the conclu sion of the intervention period. Computer-Administered Phonologica l Awareness Training Programs The two computer software programs that were used in this study were Earobics Step 1 and Lexia Early Reading. Numerous so ftware programs that target early reading instruction were explored for this study. Many of these programs address phonological awareness as only one component of early reading skills and typi cally included other activities to address language or phonics skills. The Ear obics and Lexia Early Reading programs were chosen because they both focus solely on phonological awareness and generally address the same skill areas within phonological awareness. Additional benefits of these two programs are that a basi c reading skill level is not needed to utilize these programs and both programs are availa ble for school and home use. The two programs are described in detail below.

PAGE 52

45 Earobics Step 1 Earobics Step 1 software is designed to develop auditory and phonological awareness skills in ch ildren ages four through seve n. It can also be used with older children who need a dditional practice in developing these skills. There are six interactive games that provide instructi on and practice in the following skills: rhyme identification; identification of beginning, middle, and ending sounds; segmenting words into phonemes; blending phonemes into words; auditory sequential memory; and sound discrimination. There are various levels of difficulty ranging from simple listening skills to activities with lette rs and sounds. The difficulty leve l is adjusted according to the performance of the child within each of the six ac tivities. The professional version of Earobics provides data collecti on reports used to identify part icular areas of difficulty for students using the program. Table 2 outlines the tasks, targeted skills, and number of levels associated with ea ch Earobics Step 1 game. Table 2 Earobics Step 1 Game Descriptions and Targeted Skills Game Title Task Targeted Skill(s) Levels Karloons Balloons Nine picture boxes are presented on the screen. The student is instructed to reca ll a series of sound effects, words, numbers, and speech sounds. The game advances in difficulty by increasing the number of sounds, concealing pictures until after sounds have been presented, and introducing two levels of background noise. Auditory Short-Term Memory, Auditory Sequential Memory, Auditory Performance with Competing Signals, Auditory Attention 38 C.C. Coal Car Activity 1: A target sound is introduced with a box on the screen displaying the letter that represents the sound. Another box on the screen displays the letter with a strike through it. The student is presented with a sound and instructed to click the letter if the sound corresponds with the letter. If the sound is not the target sound the student clicks the letter with the strike through it. The game advances in difficulty by presenting the sound within a word. Activity 2: A sound is presented and the student is instructed to click on the engine, coal car, or caboose to indicate if the sound is heard at the beginning, middle, or end. Phoneme Discrimination, Phoneme Identification, Phonological Sequencing, SoundSymbol Correspondence 74

PAGE 53

46 Table 2 (Continued) Earobics Step 1 Game Descriptions and Targeted Skills Game Title Task Targeted Skill(s) Levels Rap-ATap-Tap A series of drum beats or speech sounds is presented and the student is instructed to click the mouse one time for each sound heard. The game increases in difficulty by changing the amount of time between sounds presented and eliminating auditory feedback during the students response. The game advances to another level by instructing the student to count the number of syllables or speech sounds in a word. Auditory Short-Term Memory, Phonological Segmentation, Auditory Temporal Resolution 16 Caterpillar Connection Three pictures appear at the top of the screen. The student is presented with two words that make a compound word or syllables and sounds that blend into a word. The student is instructed to click on the picture that matches the word or compound word. The game increases in difficulty by using longer presentation intervals, increasing the number of syllables or sounds presented, and varying the similarity of the response choices. Phonological Blending, Auditory Attention, Auditory Short-Term Memory, Auditory Sequential Memory 56 Rhyme Time Activity 1: Three frogs present a different word, one which does not rhyme with the other two. The student is instructed to click on the frog whose word does not rhyme. The game increases in difficulty by increasing the number of frogs and introducing low and high level background noise. Activity 2: One frog speaks a word. Two other frogs each speak a different word, on of which rhymes with the first word spoken. The student is instructed to click on the frog whose word rhymes with the first. Rhyming, Auditory Attention, Auditory Sequential Memory, Auditory Short-Term Memory, Auditory Performance with Competing Signals 11 Basket Full of Eggs Activity 1: Two white hens are presented on the left side of the screen. A white and a brown hen are presented on the right side of the screen. Two vowel sounds are presented. The student is instructed to click on the two white hens if the vowel sounds are the same and on the white and brown hen if the vowel sounds are different. The game increases in difficulty by increasing the similarity of the two vowel sounds. Activity 2: Two consonant-vowel syllables are presented. The student is given the same instructions as in Activity 1 and the game increases in difficulty by increasing the similarity of the two consonant-vowel syllables. Auditory and Phoneme Discrimination, Auditory Attention, Auditory Short-Term Memory, Auditory Sequential Memory, Auditory Pattern Recognition, Auditory Temporal Ordering 114 Note. The information in this table was extracted from Ea robics Step 1 Clinic Software Users Guide by Cognitive Concepts, Inc., 2003, Evanston, IL.

PAGE 54

47 Lexia Early Reading Lexia Early Reading is a program designed to provide instruction and practice in r hyme identification, identification of beginning and ending sounds, segmenting words, and sound blendi ng. The program contains five highly interactive activities with co lorful graphics that are in troduced by a character named Lexie the Lion. Instructions are provided ora lly with no written text Each activity is designed with increasing levels of complexity to strengthen students skills within each area. Student progress is track ed through detailed reports ava ilable to teachers. Table 3 outlines the tasks, targeted skills, and number of levels associated with each Lexia Early Reading game. Table 3 Lexia Early Reading Game Desc riptions and Targeted Skills Game Title Task Targeted Skill(s) Levels Rhyme Time Three pictures are displayed, each representing a word. Each picture is highlighted and named by the computer. The student is then instructed to click on the two pictures representing the rhyming words. Patterns of Rhyme in Spoken Language 11 Sound Match Three pictures are displayed, each representing a word. Each picture is highlighted and named by the computer. The student is then instructed to click on the word (represented by the picture) that begins or ends with a certain sound. Identifying Beginning and Ending Sounds in Words 14 Word Snip One picture is displayed and named by the computer. Four balls are displayed beneath the picture and the student is instructed to drag down one ball for each syllable in the word. Segmenting Words into Syllables and Sounds 19 Sound Slide Three pictures are displayed, each representing a word. The student hears a word with a pause between syllables or sounds. The student is then instructed to click on the picture that represents the word presented. Blending Syllables and Sounds into Words 14 Consonant Tree 1 The student is presented with a letter name and corresponding sound. The computer associates the letter with a keyword pi cture and the student completes a puzzle of the keyword picture. The computer then demonstrates how the letter is formed or completes a letter search. The student is then instructed to identify pictures that begin or end with the targeted sound. Sound/Symbol Correspondence for t, b, m, s, k & f in Initial and Final Positions 14

PAGE 55

48 Table 3 (Continued) Lexia Early Reading Game Desc riptions and Targeted Skills Game Title Task Targeted Skill(s) Levels Consonant Tree 2 The student is presented with a letter name and corresponding sound. The computer associates the letter with a keyword pi cture and the student completes a puzzle of the keyword picture. The computer then demonstrates how the letter is formed or completes a letter search. The student is then instructed to identify pictures that begin or end with the targeted sound. Sound/Symbol Correspondence for p, g, r, n, d, z & l in Initial and Final Positions 16 Consonant Tree 3 The student is presented with a letter name and corresponding sound. The computer associates the letter with a keyword pi cture and the student completes a puzzle of the keyword picture. The computer then demonstrates how the letter is formed or completes a letter search. The student is then instructed to identify pictures that begin or end with the targeted sound. Sound/Symbol Correspondence for c, h, j, v, w, & y in Initial Position and Brother Pairs 12 Short Vowel Crate The student is presented with a letter name and corresponding sound. The computer associates the letter with a keyword pi cture and the student completes a puzzle of the keyword picture. The computer then demonstrates how the letter is formed or completes a letter search. The student is then instructed to identify pictures that begin or end with the targeted sound. Sound/Symbol Correspondence for Vowels in Initial Position 9 Consonant Pair Pond The student is presented with a letter name and corresponding sound. The computer associates the letter with a keyword pi cture and the student completes a puzzle of the keyword picture. The computer then demonstrates how the letter is formed or completes a letter search. The student is then instructed to identify pictures that begin or end with the targeted sound. Sound/Symbol Correspondence for sh, th, wh, & ch in Initial and Final Positions 9 Note. The information in this table was extracted from Lexia Early Reading Teacher Manual by Lexia Learning Systems, Inc., 2004, Lincoln, MA. Analysis This study used a multi-group pretest-treatment-posttest design. The dependent variables were the DIBELS measures of Initial Sounds Fluency (ISF), Letter Naming Fluency (LNF), and Phoneme Segmentation Fluency (PSF) for kindergarten and Letter Naming Fluency (LNF), Phoneme Segmentation Fluency (PSF), Nonsense Word Fluency

PAGE 56

49 (NWF), and Oral Reading Fluency (ORF) for fi rst grade. The independent variable was the specific computer-delivered phonol ogical awareness program. Demographic information was gathered for each of the th ree kindergarten and first grade groups. An analysis of variance (ANOVA) was completed to determine whether there were significant differences among pre-test gr oup mean scores on each of the pre-test measures for kindergarten and first grade. An ANOVA also was completed to determine whether there were significant differences in age among the three groups. Chi square analyses were conducted to determine whethe r differences occurred among the groups on gender, ethnicity, and teacher. Independent t-tests were co nducted to determine whether the groups differed significantly on the average amount of time that they spent utilizing each program. The DIBELS data were further analyzed by each research question. To answer the first research question (i.e., Are there outcome differences in kindergarten students early literacy sk ills among the computer-delivered phonological awareness training programs and the control group?), an analysis of covariance (ANCOVA) was conducted with the post-test score for each dependent variable, utilizing the pre-test scores as covariates. This anal ysis allowed for an examination of whether the three groups differed overall in their post-test scores on each of the dependent variables (i.e., ISF, LNF & PSF), while taking into ac count any differences that may have been present in pre-test scores. If a significan t overall difference was found for any of the dependent variables, post hoc procedures we re conducted to examine the differences in posttest scores between pairs of groups. The pr ocedures used to answer the first research question were repeated with the dependent variables for the first grade students (i.e., LNF, PSF, NWF, & ORF) to answer the s econd research question (i.e., Are there

PAGE 57

50 outcome differences in first grade student s early literacy skills among the computerdelivered phonological awareness traini ng programs and the control group?). Five sets of progress monitoring data in addition to the preand post-test scores were collected for each participant in the st udy. To answer the third research question (i.e., What are the effects of two compute r-delivered phonological awareness training programs on the development of early literacy skills in students at-risk for reading failure?), a hierarchical linear modeling (HLM) analysis was conducted. HLM enables the growth of individuals to be modeled a nd allows for the examination of differences among groups on initial status and gr owth on each dependent measure. Kindergarten and first grade students ha d two dependent vari ables in common in this study (i.e., LNF & PSF). To answer the fourth research ques tion (i.e., Are there outcome differences in early literacy skills between kindergar ten and first grade students among the phonological awareness training programs and the control group?), the analyses that were completed for the fi rst three questions were compared across kindergarten and first grade for the Letter Naming Fluency variable and the Phoneme Segmentation Fluency variable. This allowed for an examination of differences in program effects across grade levels. To provide information on the student and teacher ratings of acceptability for the intervention, the results of the student and teacher surveys were reviewed. The brief student survey consisted of th ree questions that were rated on a scale of zero to two, with two being the most favorable response. The t eacher survey consisted of 10 questions that were rated on a scale of one to six, with six being the most favorable response. The

PAGE 58

51 means, standard deviations, minimum a nd maximum scores are presented for each question for both the student and teacher surveys.

PAGE 59

52 Chapter Four Results The results of this study are discussed in three sections. First, the participants demographic data are presented. Second, the results regarding each specific research question are reported. Third, the results of th e teacher and student surveys are presented. Of the 84 students recruited for this study, 41 kindergarten (i.e., 93%) and 38 first grade (i.e., 95%) students were given permission to participate. Two kindergarten students and one first grade student moved during the cour se of the study; therefore, data from 39 kindergarten and 37 first grade students we re used for data analysis purposes. Descriptive Information Descriptive information is provided to portray demographic characteristics and preliminary baseline data of each of the thr ee groups. Table 4 displays the demographic data collected for kindergarten and first grad e students in the following categories: age (in months), gender, ethnicity, and teacher. The average age for kindergarten students across the three groups was five years, six months old (66 months, range = 13 months), while the average age for the first grade stude nts was seven years ol d (84 months, range = 22 months). An analysis of variance ( ANOVA) was conducted to determine whether significant differences occurred among the three groups on age. No statistically significant differences were found among the kindergarten groups, F (2, 36) = 0.145, p > .05, or among the first grade groups, F (2,34) = 1.05, p > .05, on age. Overall, twice as many kindergarten males participated in the study as kindergarten females although there

PAGE 60

53 were not statistically significant differen ces in the number of males and females represented among the groups for kindergarten, 2 (2, N = 39) = 1.62, p > .05. Among the first grade students, male and female students were more equally represented. Similarly, there were not statistically significant differences in the number of males and females represented among the groups for first grade, 2 (2, N = 37) = .38, p > .05. Hispanic and White students were represented far more often than Asian and Black students in both the kindergarten and first grade groups. There were not statistically significant differences in the number of stude nts represented from each ethnicity category among the groups in either kindergarten, 2 (6, N = 39) = 3.91, p > .05, or first grade, 2 (6, N = 37) = 6.75, p > .05. The majority of kindergarten students (i.e., 79%) participating in the study were from three classrooms. First grade participants were more evenly distributed across th e six first grade classes. There were not statistically significant differences in the number of students represented from each classroom among the groups for kindergarten, 2 (8, N = 39) = .76, p > .05, or first grade, 2 (10, N = 37) = 1.52, p > .05. There was not a statistically significant difference between the two experimental groups on mean amount of tim e spent on each program for kindergarten, t (24) = .59, p > .05. Similarly, a statistically significant difference was not found between the two experimental groups on the mean amount of time spent on each program for first grade, t (24) = 1.13, p >.05. Table 4 Demographic Information for Kindergarten and First Grade Kindergarten Total ( N =39) Control ( N =13) Earobics ( N =13) Lexia ( N =13) Age (in months) Mean 66.23 66.08 66.00 66.62 SD 3.10 3.64 3.22 2.57

PAGE 61

54 Table 4 (Continued) Demographic Information for Kindergarten and First Grade Kindergarten Total ( N =39) Control ( N =13) Earobics ( N =13) Lexia ( N =13) Gender Male 26 10 7 9 Female 13 3 6 4 Ethnicity Asian 6 3 1 2 Black 1 1 0 0 Hispanic 14 4 6 4 White 18 5 6 7 Teacher 1 5 2 1 2 2 3 1 1 1 3 11 3 4 4 4 9 3 3 3 5 11 4 4 3 Time on Program Mean 469.23 476.92 SD 38.83 26.89 First Grade Total ( N =37) Control ( N =12) Earobics ( N =13) Lexia ( N =12) Age (in months) Mean 84.41 86.50 83.54 83.25 SD 6.12 5.49 6.37 6.40 Gender Male 18 5 7 6 Female 19 7 6 6 Ethnicity Asian 1 0 1 0 Black 3 0 0 3 Hispanic 14 3 5 6 White 18 6 7 5 Teacher 6 4 2 1 1 7 6 2 2 2 8 7 2 3 2 9 9 3 3 3 10 7 2 3 2 11 4 1 1 2 Time on Program Mean 461.54 475.00 SD 34.12 24.31

PAGE 62

55 Research Question One: Kindergarten Outcomes To answer the first research question (i.e., Are there outcome differences in kindergarten students early literacy sk ills among the computer-delivered phonological awareness training programs and the contro l group?), an analysis of covariance (ANCOVA) was performed on each of the posttest dependent measures for kindergarten students. Analysis of covariance allows for the comparison of post-test group means while adjusting for differences that may have be en present in the initial baseline scores. Information on kindergarten pre-test measures will be presented first. Table 5 displays the mean, standard de viation, skewness, a nd kurtosis of the DIBELS pre-test scores for kindergarten stud ents. The distribution of the kindergarten control group ISF pre-test scores had a s light negative skew, a mean of 4.46, and a standard deviation of 3.23. The distribution of the kindergarten Earobics group ISF pretest scores had a moderate positive skew, a mean of 1.77, and a standard deviation of 1.88. The distribution of the kindergarten Lexia group ISF pre-test scores can be characterized as roughly symmetric with a m ean of 2.77 and a standard deviation of 2.35. The distribution of the kinderg arten control group LNF pre-te st scores had an extreme positive skew, a mean of 1.23, and a standard deviation of 1.83. The kindergarten Earobics group LNF pre-test scores produced a slight pos itive skewed distribution, a mean of 2.62, and a standard deviation of 2.10. The distribution of the Lexia group LNF pre-test scores had notable positive skew, a mean of 1.62, and a standard deviation of 2.14. All kindergarten students in all groups scored zero on the pre-test measure of PSF.

PAGE 63

56 Table 5 Kindergarten DIBELS Pre-test Scores by Group Control ( N =13) Earobics ( N =13) Lexia ( N =13) ISF Mean 4.46 1.77 2.77 SD 3.23 1.88 2.35 Skewness -0.21 0.93 0.19 Kurtosis -1.64 0.50 -1.72 LNF Mean 1.23 2.62 1.62 SD 1.83 2.10 2.14 Skewness 1.90 0.48 1.50 Kurtosis 3.29 -0.003 1.16 PSF Mean 0 0 0 SD .000 .000 .000 Skewness Kurtosis Note. DIBELS = Dynamic Indicators of Basic Early Literacy Skills. ISF = Initial Sounds Fluency. LNF = Letter Naming Fluency. PSF = Phoneme Segmentation Fluency. An analysis of variance (ANOVA) pr ocedure was conducted to determine whether significant differences occurred am ong kindergarten pre-test group mean scores on Initial Sounds Fluency a nd Letter Naming Fluency. Phoneme Segmentation Fluency was not included in this analysis because th e mean score for all groups was zero. The three groups in this study violate th e ANOVA assumption of independence. Additionally, not all groups met the assumption of normality. The ANOVA procedure is robust to violations of normality when sample sizes are large; however, given the sample size in this study (N = 13 per group), some caution shoul d be taken when interpreting results. Table 6 displays the results of the ANOVA. The F-ratio was statistically significant for ISF, F (2,36) = 3.707, p < .05, indicating that there were significant differences among the groups on the pre-test m ean scores for Initial Sounds Fluency. Post hoc tests were conducted to determin e which of the groups were different on the

PAGE 64

57 initial pre-test scores. The Games-Howell test results indicated that the control group and Earobics group di ffered significantly, p < .05, on the ISF pre-test measure. Additionally, a large effect size ( d = 1.02) was noted when comparing the Earobics group mean to the control group mean on the ISF pre-test meas ure. No other groups differed significantly from one another on the ISF pre-test measur e; however, a medium effect size was found when comparing the Lexia group m ean to the control group mean ( d = 0.60) and when comparing the Earobics group m ean to the Lexia group mean ( d = 0.50). The results of the ANOVA were not statistically significant for the pre-test measure of LNF, p > .05; however, a medium effect size was found when comparing the control group mean to th e Earobics group mean ( d = 0.71) and when comparing the means of the Earobics and Lexia groups ( d = 0.47). Table 6 Analysis of Variance for Kinde rgarten Pre-test Measures Type III Sum of Squares df Mean Square F Value Significance Pre-test Scores ISF 48.15 2 24.08 3.71 .034 LNF 13.28 2 6.64 1.61 .214 Table 7 displays the mean, standard de viation, skewness, a nd kurtosis of the DIBELS post-test scores for kindergarten students. The distribution of the kindergarten control group ISF post-test scores had a m oderate positive skew, a mean of 6.15, and a standard deviation of 3.00. The distribution of the kindergarten Earobics group ISF posttest scores had a slight negative skew, a m ean of 12.92, and a standard deviation of 4.61. The kindergarten Lexia group I SF post-test scores produced a slight negative skewed distribution with a mean of 9.92 and a standard deviation of 5.01. The distribution of the

PAGE 65

58 kindergarten control group LNF post-test scores had an extreme positive skew, a mean of 13.08, and a standard deviation of 10.00. The ki ndergarten Earobics group LNF post-test scores produced a slight positive skewed distribution, a mean of 21.08, and a standard deviation of 11.74. The distribution of the ki ndergarten Lexia group LNF post-test scores had moderate positive skew, a mean of 17.31, and a standard deviation of 12.91. All students in the kindergarten control group scor ed zero on the post-test PSF measure. The distribution of the Earobics group PSF post-test scores had a slight negative skew, a mean of 1.31, and a standard deviation of 0.75. Similarly, the kindergarten Lexia group posttest scores produced a slight negative sk ewed distribution with a mean of 1.31 and a standard deviation of 0.63. Table 7 Kindergarten DIBELS Post-test Scores by Group Control ( N =13) Earobics ( N =13) Lexia ( N =13) ISF Mean 6.15 12.92 9.92 SD 3.00 4.61 5.01 Skewness .70 -.35 .40 Kurtosis 1.083 -1.00 .21 LNF Mean 13.08 21.08 17.31 SD 10.00 11.74 12.91 Skewness 1.90 .66 .72 Kurtosis 3.29 .90 -.98 PSF Mean 0 1.31 1.31 SD .000 .75 .63 Skewness -.61 -.31 Kurtosis -.78 -.32 Note. DIBELS = Dynamic Indicators of Basic Early Literacy Skills. ISF = Initial Sounds Fluency. LNF = Letter Naming Fluency. PSF = Phoneme Segmentation Fluency.

PAGE 66

59 An analysis of covariance (ANCOVA) was conducted to examine differences among post-test measure mean scores on Init ial Sounds Fluency (ISF) and Letter Naming Fluency (LNF) while statistically controlling fo r pre-test scores. As with the pre-test scores, individuals within groups are not i ndependent and the assumption of normality is not tenable. The test of homogeneity of regression for ISF was not statistically significant, F (2,33) = 0.230, p > .05, indicating that the inter action between the covariate of pre-test ISF and the program factor wa s not significant. Similarly, the test of homogeneity of regression for LNF was not statistically significant, F (2,33) = 1.24, p > .05, indicating that the interaction between the covariate of pre-test LNF and the program factor was not significant. The test of homogeneity of regression for PSF was not conducted because the pre-test PSF mean scor e was zero for all groups; therefore, an analysis of variance was conducted on the pos t-test measure PSF. Table 8 displays the results of the ANCOVA for the ISF and LNF post -test scores, utilizing the pre-test scores for each as covariates. The results of th e ANOVA for the PSF post-test scores also are displayed in Table 8. Results of the ANCOVA indicated that there were statistically significant differences among the three adjusted means, F (2,35) = 12.19, p < .001, on the ISF posttest measure. Additionally, the partial eta squared of .411 suggested a strong relationship between post-test scores and program, controllin g for pre-test scores. The results of the ANCOVA indicated that there were no statistically signi ficant differences among the three adjusted means, F (2,35) = .72, p > .05, on the LNF post-test measure. The results of the ANOVA indicated that there were statistically si gnificant differences among the three means, F (2,35) = 23.12, p < .001, on the PSF post-test measure. Given the

PAGE 67

60 restriction of range in post-test PSF scores (i .e., 0 2), these results should be interpreted with caution. Table 8 Analysis of Variance and Analysis of Covariance for Kindergarten Post-test Measures Type III Sum of Squares df Mean Square F Value Significance Post-test Scores ISF (ANCOVA) 393.80 2 196.90 12.19 .000 LNF (ANCOVA) 176.79 2 88.40 .72 .494 PSF (ANOVA) 14.821 2 7.41 23.12 .000 Note. ANOVA was used for PSF because ther e was no variability on the pre-test. Further examination of the ANCOVA re sults for ISF was conducted using post hoc analysis procedures, which evaluate signi ficant differences in the adjusted means between groups. Table 9 di splays the adjusted means and post hoc analysis for differences in adjusted mean scores for Initial Sounds Fluency (ISF). The results indicate statistically significant differences between the control group and the two experimental groups, p < .05. The difference between the Ea robics group and the control group was nearly twice the difference between the Lexia group and the control group, indicating that the Earobics group performed higher on the ISF post-test measure than both the Lexia group and the control group. Additionally, a la rge effect size was noted when comparing the Earobics group mean to the control group mean ( d = 1.74) and when comparing the Lexia group mean to the control group mean ( d = 0.91) on the ISF post-test measure. The effect size was medium when comparing th e Earobics group mean to the Lexia group mean ( d = 0.62).

PAGE 68

61 Table 9 ANCOVA Post Hoc Analysis for Kindergarten Initial Sounds Fluency Control Group (Adjusted M = 5.21) Earobics Group (Adjusted M = 13.72) Lexia Group (Adjusted M = 10.07) Control Group (Adjusted M = 5.21) Earobics Group (Adjusted M = 13.72) 8.52* Lexia Group (Adjusted M = 10.07) 4.87* -3.65 p < .05 The results of the ANCOVA indicated that statistically significant differences were not found among the post-test ad justed mean scores for LNF, p > .05. When examining effect sizes for this variable, a small effect size was found when comparing the Lexia group mean to the control group mean and Earobics group mean, d = 0.37 and d = 0.31, respectively. However, a medium effect size was found when comparing the Earobics group mean to the control group mean, d = 0.73. Further examination of the ANOVA results for PSF was conducted using post hoc analysis procedures. Table 10 displays the means and post hoc analysis for differences in mean scores for Phoneme Segmentation Fl uency (PSF). Statistically significant differences were indicated between the c ontrol group and the two experimental groups, p < .05. Additionally, a large effect size was found when comparing the control group mean to the Earobics group mean ( d = 2.47) and when comparing the control group mean to the Lexia group mean ( d = 2.94).

PAGE 69

62 Table 10 ANOVA Post Hoc Analysis for Kindergarten Phoneme Segmentation Fluency Control Group ( M = 0.00) Earobics Group ( M = 1.31) Lexia Group ( M = 1.31) Control Group ( M = 0.00) Earobics Group ( M = 1.31) 1.31* Lexia Group ( M = 1.31) 1.31* 0.00 p < .05 Research Question Two: First Grade Outcomes To answer the second research question (i.e., Are there outcome differences in first grade students early literacy skills among the computer-delivered phonological awareness training programs and the contro l group?), an analysis of covariance (ANCOVA) was performed on each of the post-te st dependent measures for first grade, allowing for the comparison of post-test group means while adjusting for differences in pre-test scores. The pre-test scores for each variable are discussed first. Table 11 displays the mean, standard de viation, skewness, and kurtosis for each of the DIBELS pre-test measur es for first grade students. The control group LNF pre-test scores formed a moderate negative skewed distribution with a mean of 25.83 and a standard deviation of 4.49. The Earobics group LNF pre-test scores produced a notable negative skewed distribution, a mean of 28.23, and a standard deviation of 7.81. The distribution of the Lexia group LNF pre-test scores was roughly symmetric with a mean of 24.33 and a standard deviation of 14.90. The control group PSF pre-test scores produced a slightly positive skewed distribution with a mean of 23.17 and a standard deviation of 11.24. The Earobics group PSF pre-test scores formed a slightly positive

PAGE 70

63 skewed distribution with a mean of 19.54 a nd a standard deviation of 11.98. The Lexia group PSF pretest scores created a moderately positive skewed distribution with a mean of 18.83 and a standard deviation of 13.14. The distribution of the control group NWF pre-test scores had a slight negative skew with a mean of 16.08 and a standard deviation of 7.56. The Earobics group NWF pre-test sc ores produced a slight negative skewed distribution with a mean of 14.38 and a st andard deviation of 8.12. The Lexia group NWF pre-test scores formed a moderate pos itive skewed distribution with a mean of 13.42 and a standard deviation of 12.67. The control group ORF pre-test scores created a slight positive skewed distribution with a mean of 7.42 and a standard deviation of 5.62. The distribution of the Earobics group ORF pr e-test scores had a moderate positive skew with a mean of 9.08 and a standard deviation of 6.68. The Lexia group ORF pre-test scores produced a slight positive skewed distri bution with a mean of 7.00 and a standard deviation of 5.58. Table 11 First Grade DIBELS Pre-test Scores by Group Control ( N =12) Earobics ( N =13) Lexia ( N =12) LNF Mean 25.83 28.23 24.33 SD 4.49 7.81 14.90 Skewness -1 .03 -1.56 -0.15 Kurtosis 2.25 2.97 -1.34 PSF Mean 23.17 19.54 18.83 SD 11.24 11.98 13.14 Skewness 0.70 0.42 0.97 Kurtosis -0.06 -0.87 -0.10 NWF Mean 16.08 14.38 13.42 SD 7.56 8.12 12.67 Skewness -0.21 -0.28 0.96 Kurtosis -0.46 -0.74 1.03

PAGE 71

64 Table 11 (Continued) First Grade DIBELS Pre-test Scores by Group Control ( N =12) Earobics ( N =13) Lexia ( N =12) ORF Mean 7.42 9.08 7.00 SD 4.56 6.68 5.58 Skewness 0.57 1.35 0.33 Kurtosis 0.22 1.38 -0.50 Note. DIBELS = Dynamic Indicators of Basic Early Literacy Skills. LNF = Letter Naming Fluency. PSF = Phoneme Segmentation Fluency. NWF = Nonsense Word Fluency. ORF = Oral Reading Fluency. An analysis of variance (ANOVA) pro cedure also was conducted on the first grade pre-test group mean scores of Le tter Naming Fluency, Phoneme Segmentation Fluency, Nonsense Word Fluency, and Oral Re ading Fluency (see Table 12). Similar to the kindergarten groups, not all first grade groups met the ANOVA assumptions of independence and normality, which should be not ed when interpreting results. The test indicated that no sign ificant differences were found am ong the groups on any of the pretest mean scores. Table 12 Analysis of Variance for First Grade Pre-test Measures Type III Sum of Squares df Mean Square F Value Significance Pretest Scores LNF 97.04 2 48.52 .486 .619 PSF 130.68 2 65.34 .443 .646 NWF 43.79 2 21.90 .234 .793 ORF 30.49 2 15.24 .469 .630

PAGE 72

65 Table 13 displays the mean, standard de viation, skewness, and kurtosis for each of the DIBELS post-test meas ures for first grade students. The control group LNF posttest scores formed a roughly symmetrical distribution with a mean of 37.67 and a standard deviation of 8.97. The Earobics group LNF post-test scores produced a moderate negative skewed di stribution, a mean of 52.31, and a standard deviation of 13.83. The distribution of the Lexia group LNF post-test scores had a slight negative skew with a mean of 46.25 and a standard deviation of 14.33. The control group PSF post-test scores produced a slightly positive sk ewed distribution with a mean of 32.50 and a standard deviation of 10.57. The Earobics gr oup PSF post-test scores formed a slightly negative skewed distribution w ith a mean of 47.23 and a sta ndard deviation of 8.08. The Lexia group PSF post-test scores created a slightly negative skewed distribution with a mean of 36.75 and a standard deviation of 13.71. The distribution of the control group NWF post-test scores had a moderate positiv e skew with a mean of 27.67 and a standard deviation of 11.44. The Earobics group NW F post-test scores produced a moderate positive skewed distribution with a mean of 47.46 and a standard deviation of 19.65. The Lexia group NWF post-test scores formed a s light negative skewed distribution with a mean of 39.58 and a standard deviation of 15.12. The control group ORF post-test scores created a moderate positive skewed distribution with a mean of 13.08 and a standard deviation of 7.83. The distribution of the Earobics group ORF post-test scores had a slight positive skew with a mean of 29.31 and a standard deviation of 18.53. The Lexia group ORF pre-test scores produ ced a slight positive skewed distribution with a mean of 19.92 and a standard deviation of 9.65.

PAGE 73

66 Table 13 First Grade DIBELS Post-test Scores by Group Control ( N =12) Earobics ( N =13) Lexia ( N =12) LNF Mean 37.67 52.31 46.25 SD 8.97 13.83 14.33 Skewness -0 .12 -0.95 -0.30 Kurtosis 0.88 -0.52 -1.95 PSF Mean 32.50 47.23 36.75 SD 10.57 8.08 13.71 Skewness 0.44 -0.56 -0.05 Kurtosis -1.02 -1.19 -2.12 NWF Mean 27.67 47.46 39.58 SD 11.44 19.65 15.12 Skewness 1.14 1.13 -0.18 Kurtosis 4.51 2.65 -0.15 ORF Mean 13.08 29.31 19.92 SD 7.83 18.53 9.65 Skewness 1.29 0.65 0.65 Kurtosis 0.48 0.08 -0.23 Note. DIBELS = Dynamic Indicators of Basic Early Literacy Skills. LNF = Letter Naming Fluency. PSF = Phoneme Segmentation Fluency. NWF = Nonsense Word Fluency. ORF = Oral Reading Fluency. An analysis of covariance (ANCOVA) was conducted to examine differences among post-test measure mean scores on Letter Naming Fluency (LNF), Phoneme Segmentation Fluency (PSF), Nonsense Word Fluency (NWF), and Oral Reading Fluency (ORF) while statistically controlling for pre-test scores. Similar to pre-test scores, not all groups met the assumptions of independence and normality with regard to post-test scores. The test of homogeneity of regression was not st atistically significant for LNF, F (2,31) = .97, p > .05; PSF, F (2,31) = 2.62, p > .05; NWF, F (2,31) = 1.90, p > .05; or ORF, F (2,31) = 1.48, p > .05. This indicates that the interaction between the covariate of each pre-test m easure and the program factor is not significant. Table 14

PAGE 74

67 displays the results of the ANCOVA for the LNF, PSF, NWF, and ORF post-test scores, utilizing the pre-test scores for each as covariates. The results of the ANCOVA indicated that there were significant differences among the three adjusted means, F (2,33) = 8.45, p = .001, on the LNF post-test measure. Additionally, the partia l eta squared of .34 suggested a st rong relationship between posttest scores and program, cont rolling for pre-test scores. The results of the ANCOVA for the PSF post-test measure indicated that th ere were significant differences among the three adjusted means, F (2,33) = 11.33, p < .001. The partial eta squared of .41 suggested a strong relationship between post-test scor es and program, controlling for pre-test scores. The results of the ANCOVA indicated that there we re significant differences among the three adjusted means, F (2,33) = 9.96, p < .001, on the NWF post-test measure. Additionally, the partia l eta squared of .38 suggested a st rong relationship between posttest scores and program, cont rolling for pre-test scores. The results of the ANCOVA for the ORF post-test measure indicated that th ere were significant differences among the three adjusted means, F (2,33) = 6.36, p = .005. The partial eta squared of .28 suggested a moderate relationship between post-test scor es and program, controlling for pre-test scores. Table 14 Analysis of Covariance for First Grade Post-test Measures First Grade Type III Sum of Squares df Mean Square F Value Significance Post-test Scores LNF 1044.37 2 522.18 8.45 .001 PSF 1758.52 2 879.26 11.33 .000 NWF 2990.34 2 1495.17 9.96 .000 ORF 1129.77 2 564.89 6.36 .005 Note. Pre-test scores for each variable were used as a covariate.

PAGE 75

68 Further examination of the ANCOVA re sults for LNF was conducted using post hoc analysis procedures, which evaluate signi ficant differences in the adjusted means between groups. Table 15 displays the ad justed means and post hoc analysis for differences in adjusted mean scores for Letter Naming Fluency (LNF). The results indicate statistically signifi cant differences between the control group and Lexia group, p < .05, as well as between the c ontrol group and Earobics group, p = .001. Statistically significant differences were not indicated between the Ea robics group and the Lexia group. Additionally, a large e ffect size was noted when co mparing the Earobics group mean to the control group mean ( d = 1.25). A medium effect size was noted when comparing the Lexia group mean with the control group mean (d = 0.72). The effect size was small when comparing the Earobics group mean to the Lexia group mean ( d = 0.33). Table 15 ANCOVA Post Hoc Analysis for First Grade Letter Naming Fluency Control Group (Adjusted M = 38.02) Earobics Group (Adjusted M = 50.26) Lexia Group (Adjusted M = 48.11) Control Group (Adjusted M = 38.02) Earobics Group (Adjusted M = 50.26) 12.24** Lexia Group (Adjusted M = 48.11) 10.09* -2.16 p < .05, ** p = .001 Further examination of the ANCOVA re sults for PSF was conducted using post hoc analysis procedures. Table 16 displays the adjusted means and post hoc analysis for differences in adjusted mean scores for Phoneme Segmentation Fluency (PSF). The results indicate statistically significant di fferences between the c ontrol group and Lexia group, p < .05, the control group and Earobics group, p < .001, as well as the Earobics

PAGE 76

69 group and the Lexia group, p < .05. Additionally, a large effect size was noted when comparing the Earobics group mean to the control group mean ( d = 1.57) and the Earobics group mean to the Lexia group mean ( d = 0.94). A small effect size was noted when comparing the Lexia group mean with the control group mean ( d = 0.35). Table 16 ANCOVA Post Hoc Analysis for First Grade Phoneme Segmentation Fluency Control Group (Adjusted M = 31.02) Earobics Group (Adjusted M = 47.75) Lexia Group (Adjusted M = 37.66) Control Group (Adjusted M = 31.02) Earobics Group (Adjusted M = 47.75) 16.73** Lexia Group (Adjusted M = 37.66) 6.64* -10.09* p < .05, ** p < .001 Further examination of the ANCOVA resu lts for NWF also was conducted using post hoc analysis procedures. Table 17 displays the adjusted means and post hoc analysis for differences in adjusted mean scores for Nonsense Word Fluency (NWF). The results indicate statistically signifi cant differences between the control group and Lexia group, p < .05, as well as between the c ontrol group and Earobics group, p = .001. Statistically significant differences were not indicated between the Ea robics group and the Lexia group, p > .05. Additionally, a large effect size was noted when comparing the Earobics group mean to the control group mean ( d = 1.22) and when comparing the Lexia group mean with the control group mean ( d = 0.89). A medium effect size was found when comparing the Earobics group m ean to the Lexia group mean ( d = 0.45).

PAGE 77

70 Table 17 ANCOVA Post Hoc Analysis for Firs t Grade Nonsense Word Fluency Control Group (Adjusted M = 26.11) Earobics Group (Adjusted M = 47.72) Lexia Group (Adjusted M = 40.87) Control Group (Adjusted M = 26.11) Earobics Group (Adjusted M = 47.72) 21.61** Lexia Group (Adjusted M = 40.87) 14.76* -6.85 p < .05, ** p = .001 Further examination of the ANCOVA re sults for ORF was conducted using post hoc analysis procedures. Table 18 displays the adjusted means and post hoc analysis for differences in adjusted mean scores for Oral Reading Fluency (ORF). The results indicate statistically significant differences between the control group and Earobics group, p < .05. Statistically significant differe nces were not indicated between the control group and the Lexia group or the Earobics group and the Lexia group, p > .05. Additionally, a large e ffect size was noted when compar ing the Earobics group mean to the control group mean ( d = 1.12). A medium effect size was noted when comparing the Lexia group mean with th e control group mean ( d = 0.78) and when comparing the Earobics group mean to the Lexia group mean ( d = 0.63). Table 18 ANCOVA Post Hoc Analysis for First Grade Oral Reading Fluency Control Group (Adjusted M = 13.81) Earobics Group (Adjusted M = 27.35) Lexia Group (Adjusted M = 21.31) Control Group (Adjusted M = 13.81) Earobics Group (Adjusted M = 27.35) 13.54* Lexia Group (Adjusted M = 21.31) 7.51 -6.03 p < .05

PAGE 78

71 Research Question Three: Kindergar ten and First Grade Development To answer the third research question (i.e ., What are the effects of two computerdelivered phonological awareness training progra ms on the development of early literacy skills in students at risk for reading fa ilure?), a hierarchical linear modeling (HLM) analysis was conducted. Data were gathered at seven timepoints (i.e., pre-test, 5 progress monitoring points, and post-test) for each of th e students in the two experimental groups and the control group to examine student pr ogress on each of the dependent variables over the course of the intervention period. Table 19 displays the coefficients for the first HLM analysis for all kindergarten students for ISF without level2 predictor variables. Incl uded in this table are the estimates of the coefficients for the aver age regression equation, the estimates of the variance components, and the reliability coefficient es timates of OLS regression parameters. The mean initial status fo r kindergarten ISF was 3.45, indicating that on average students had an initial pre-test scor e of 3.45. The average slope for kindergarten ISF was 1.08, indicating that their scores incr eased at each progre ss monitoring point by 1.08 initial sounds per minute. Both coeffi cients were statistically significant, p < .001. The estimates of the variation across the individually-estimated growth curves were statistically significant for both intercept and slope, p < .001. This indicates that there was considerable variation among individual students on both their initial pre-test scores and growth rate. Additionally, the average re liabilities for initial status and slope were .86 and .87, respectively. There was a negative correlation between initial status and slope ( r = -.18), suggesting that students who had higher pre-test scores made less growth over time.

PAGE 79

72 Table 19 Unconditional Model for Kindergarten ISF Coefficient Variance Component Standard Error df t-ratio Chisquare Fixed Effect Initial Status/Intercept 3.45 0.48 7.12* Slope/Time 1.08 0.14 7.66* Random Effect Initial Status 7.84 38 269.81* Slope 0.68 38 299.94* Reliability of OLS regression coefficient estimate Initial Status .86 Slope .87 *p < .001 Since the results indicated that the inte rcepts and slopes diffe r significantly across students, a second HLM analysis was conduc ted introducing the predictor variable, program (i.e., Control, Earobics, or Lexia). This analysis allows the examination of differences between the programs on the intercept and slope. The results of this analysis are displayed in Table 20. Table 20 Conditional Model for Kindergarten ISF Intercept/Initial Status Slope Coefficient Standard Error p-value Coefficient Standard Error p-value Earobics vs. Control -2.29 1.15 0.055 1.55 0.24 0.000 Lexia vs. Control -0.75 1.15 0.521 0.79 0.24 0.003 Earobics vs. Lexia -1.54 1.15 0.191 0.76 0.24 0.004 The results indicate that there were no statistically significant differences between the control group and the experimental groups or between the two experimental groups

PAGE 80

on initial status, p > .05. There were, however, sta tistically significant differences between the control group a nd Earobics group, the contro l group and Lexia group, and the Earobics group and Lexia group on slope, p <.01. Results show that the Lexia group gained 0.79 more initial sounds per minute than the control group. Additionally, the Earobics group gained 0.76 more initial sounds per minute than the Lexia group and gained 1.55 more initial sounds per minute than the control group. Figure 1 shows a comparison of the initial status and growth rate of the three groups on the dependent variable of Initial Sounds Fluency (ISF). Ov erall, both experimental groups showed a greater rate of growth when compared to th e control group; however, the growth rate of the Earobics group was nearly twice the growth rate of the Lexia group. 0 2 4 6 8 10 12 14 16 Pre-test12345Posttest Time (weeks)Initial Sounds Correct Per Minute Control Earobics Lexia Figure 1. Initial status and growth rate of the kindergarten Earobics, Lexia, and control groups on Initial Sounds Fluency. Table 21 displays the coefficients for the first HLM analysis for all kindergarten students for LNF without level-2 predictor va riables. The mean initial status for kindergarten LNF was 4.14, indicating that on average students had an initial pre-test 73

PAGE 81

74 score of 4.14. The average slope for kinde rgarten LNF was 2.39, i ndicating that their scores increased at each progress monitori ng point by 2.39. Both coefficients were statistically significant, p < .001. The estimates of the variation across the individuallyestimated growth curves we re statistically significant for both intercept and slope, p < .001, indicating that there was considerable variation among individual students on both their initial pre-test scores and growth rate. Additionally, the average reliabilities for initial status and slope were .76 and .91, respectively. There was a moderate positive correlation between initial status and slope ( r = .69), indicating that students who had higher pre-test scores made more growth over time. Table 21 Unconditional Model for Kindergarten LNF Coefficient Variance Component Standard Error df t-ratio Chisquare Fixed Effect Initial Status/Intercept 4.14 0.64 6.46* Slope/Time 2.39 0.29 8.33* Random Effect Initial Status 12.14 38 157.32* Slope 2.91 38 410.58* Reliability of OLS regression coefficient estimate Initial Status .76 Slope .91 *p < .001 The results indicated that the intercepts and slopes differ significantly across students; therefore, a second HLM analys is was conducted introducing the predictor variable (i.e., program) to examine the diffe rences between the programs on the intercept and slope. No statistically significant differe nces occurred between the control group and the experimental groups or between the two experimental groups on initial status or

PAGE 82

slope, p > .05. This indicates that the variance in the students initia l status and growth rate is not significantly account ed for by group membership. The results of this analysis are displayed in Table 21 and Figure 2. Table 22 Conditional Model for Kindergarten LNF Intercept/Initial Status Slope Coefficient Standard Error p-value Coefficient Standard Error p-value Earobics vs. Control 1.32 1.59 0.410 1.18 0.69 0.096 Lexia vs. Control -0.29 1.59 0.857 0.88 0.69 0.214 Earobics vs. Lexia 1.61 1.59 0.317 0.31 0.69 0.662 0 5 10 15 20 25 Pre-test12345Posttest Time (weeks)Letters Named Correctly Per Minute Control Earobics Lexia Figure 2. Initial status and growth rate of the kindergarten Earobics, Lexia, and control groups on Letter Naming Fluency. The coefficients for the first HLM anal ysis for all kindergarten students for PSF without any level-2 predictor va riables are displayed in Table 23. The mean initial status for kindergarten PSF was 0.14, indicating that on average students had an initial pre-test score of 0.14. The average slope for kinde rgarten PSF was 0.14, indicating that their 75

PAGE 83

76 scores increased at each progress monitori ng point by less than one phoneme segmented correctly per minute. Both coeffici ents were statistically significant, p = .01. The estimates of the variation across the indi vidually-estimated growth curves were statistically significant for slope, p < .001, but not for intercept, p > .50. All kindergarten students scored zero on the PSF pre-test m easure, therefore, there was no variation among individual students on initial status There was variation among individual students on their growth rate and the average reliability for the slope was .58. There was a negative correlation between initial status and slope ( r = -.962), which suggests that if students had higher pre-test scores th ey would make less growth over time. Table 23 Unconditional Model for Kindergarten PSF Coefficient Variance Component Standard Error df t-ratio Chisquare Fixed Effect Initial Status/Intercept 0.14 0.05 2.74* Slope/Time 0.14 0.02 6.31** Random Effect Initial Status 0.00 38 20.48 Slope 0.01 38 77.78** Reliability of OLS regression coefficient estimate Initial Status 0.01 Slope 0.58 *p < .01, **p < .001 Since the results indicated that the sl opes differ significantly across students, a second HLM analysis was conducted introducing the predictor variab le, program. This analysis allowed the examination of diffe rences between the programs on slope. The results of this analysis are displayed in Table 24.

PAGE 84

77 Table 24 Conditional Model for Kindergarten PSF Intercept/Initial Status Slope Coefficient Standard Error p-value Coefficient Standard Error p-value Earobics vs. Control -0.21 0.13 0.112 0.19 0.04 0.000 Lexia vs. Control -0.19 0.13 0.143 0.19 0.04 0.000 Earobics vs. Lexia -0.02 0.13 0.899 -0.01 0.04 0.901 The results indicate that there were si gnificant differences between control group and Earobics group and the contro l group and Lexia group, on slope, p <.01. No statistically significant diffe rences were found between the Earobics group and Lexia group on slope, p > .05. Results show that both the Earobics group and the Lexia group gained 0.19 more phonemes segmented correc tly per minute than the control group. Figure 3 shows a comparison of the initial stat us and growth rate of the three groups on the dependent variable of Phoneme Segmentation Fluency (PSF). Overall, both experimental groups showed a greater rate of growth when compared to the control group; however, the practical si gnificance of this should be taken into account given the restriction of range from the pre-te st to the post-test (i.e., 0 to 2). Table 25 displays the coefficients for the first HLM analysis for all first grade students for LNF without level-2 predictor variables. The mean initial status for first grade LNF was 26.40, indicating that on averag e students had an initial LNF pre-test score of 26.40. The average slope for kinde rgarten LNF was 3.17, indicating that on average their scores increased at each progress monitoring point by 3.17. Both coefficients were statistically significant, p < .001. The estimates of the variation across the individually-estimated growth curves we re statistically significant for both intercept

PAGE 85

and slope, p < .001, indicating that th ere was considerable va riation among individual students on both their initial pretest scores and growth rate. The average reliabilities for initial status and slope were .96 and .88, respectively. There was a weak negative correlation between initial status and slope ( r = -.002), suggesting th at students who had higher pre-test scores made sl ightly less growth over time. 0 0.2 0.4 0.6 0.8 1 1.2 1.4 PreTest 12345PostTest Time (weeks)Phonemes Segmented Correctly Per Minute Control Earobics Lexia Figure 3. Initial status and growth rate of the kindergarten Earobics, Lexia, and control groups on Phoneme Segmentation Fluency. Table 25 Unconditional Model for First Grade LNF Coefficient Variance Component Standard Error df t-ratio Chisquare Fixed Effect Initial Status/Intercept 26.40 1.61 16.42* Slope/Time 3.17 0.27 11.96* Random Effect Initial Status 91.65 36 853.57* Slope 2.29 36 301.39* Reliability of OLS regression coefficient est Initial Status .96 Slope .88 *p < .001 78

PAGE 86

79 Since the results indicated that the inte rcepts and slopes diffe r significantly across students, a second HLM analysis was conducte d introducing the predictor variable of program allowing the examination of differe nces between the programs on the intercept and slope. The results of this an alysis are displayed in Table 26. Table 26 Conditional Model for First Grade LNF Intercept/Initial Status Slope Coefficient Standard Error p-value Coefficient Standard Error p-value Earobics vs. Control 1.31 4.02 0.746 2.10 0.55 0.001 Lexia vs. Control -0.42 4.10 0.919 1.57 0.56 0.009 Earobics vs. Lexia 1.73 4.02 0.669 0.53 0.55 0.339 The results indicate that there were no statistically significant differences between the control group and the experimental groups or between the two experimental groups on initial status, p > .05. Similarly, there were no were no statistically significant differences between the two expe rimental groups on the slope, p > .05. There were, however, statistically significant differen ces between the control group and Earobics group, p = .001 as well as between the control group and the Lexia group, p < .01 on the slope. Figure 4 shows a comparison of the init ial status and growth rate of the three groups on the dependent variable of Letter Naming Fluency (LNF). Results show that the Earobics group gained 2.10 more letters na med correctly per minute than the control group at each progress monitoring point.

PAGE 87

0 10 20 30 40 50 60 Pre-test12345Posttest Time (weeks)Letters Named Correctly Per Minute Control Earobics Lexia Figure 4. Initial status and growth rate of the first grade Ea robics, Lexia, and control groups on Letter Naming Fluency. Table 27 displays the coefficients for the first HLM analysis for all first grade students for PSF without level-2 predictor variables. The me an initial status for first grade PSF was 20.62, indicating that on average st udents had an initial pre-test score of 20.62. The average slope for first grade PSF was 3.02, indicating that their scores increased at each progress monitoring point by 3.02. Both coefficients were statistically significant, p < .001. The estimates of the variation across the individually-estimated growth curves were statistically si gnificant for both intercept and slope, p < .001. This indicates that there was cons iderable variation among indivi dual students on both their initial pre-test scores and growth rate. Additionally, the average reliabilities for initial status and slope were .96 and .90, respect ively. There was a moderate negative correlation between initial status and slope ( r = -.43), suggesting that students who had higher pre-test scores made less growth over time. 80

PAGE 88

81 Table 27 Unconditional Model for First Grade PSF Coefficient Variance Component Standard Error df t-ratio Chisquare Fixed Effect Initial Status/Intercept 20.62 1.88 10.95* Slope/Time 3.02 0.33 9.12* Random Effect Initial Status 126.17 36 927.77* Slope 3.68 36 373.71* Reliability of OLS regression coefficient estimate Initial Status .96 Slope .90 *p < .001 A second HLM analysis was conducted in troducing the predic tor variable of program, allowing the examination of differe nces between the programs on the intercept and slope. The results of this an alysis are displayed in Table 28. Table 28 Conditional Model for First Grade PSF Intercept/Initial Status Slope Coefficient Standard Error p-value Coefficient Standard Error p-value Earobics vs. Control -5.04 4.61 0.283 3.09 0.64 0.000 Lexia vs. Control -5.38 4.70 0.261 1.33 0.65 0.048 Earobics vs. Lexia 0.35 4.61 0.941 1.76 0.64 0.010 The results indicate that there were no statistically significan t differences between the control group and the experimental groups or between the two experimental groups on initial status, p > .05. There were, however, sta tistically significant differences between the control gr oup and Earobics group, p < .001, the control group and Lexia

PAGE 89

group, p < 05, and the Earobics gro up and Lexia group on slope, p = .01. Results show that the Lexia group gained 1.33 more phonemes segmented correctly per minute at each progress monitoring point than the contro l group. Additionally, the Earobics group gained 1.76 more phonemes segmented correctly per minute at each progress monitoring point than the Lexia grou p and gained 3.09 more phonemes segmented correctly per minute at each progress mon itoring point than the contro l group. Figure 5 shows a comparison of the initial status and growth rate of the three groups on the dependent variable of Phoneme Segmentation Fluency (PSF). Overall, both experimental groups showed a greater rate of growth when co mpared to the control group; however, the growth rate of the Earobics group was more than twice the growth rate of the Lexia group. 0 10 20 30 40 50 60 Pre-test12345Posttest Time (weeks)Phonemes Segmented Correctly Per Minute Control Earobics Lexia Figure 5. Initial status and growth rate of the first grade Ea robics, Lexia, and control groups on Phoneme Segmentation Fluency. Table 29 displays the coefficients for the first HLM analysis for all first grade students for NWF without level-2 predictor variables. The m ean initial status for first 82

PAGE 90

83 grade NWF was 15.59, indicating that on average students had an initial pre-test score of 15.59. The average slope for first grade NW F was 3.96, indicating that their scores increased at each progress monitoring point by 3.96. Both coefficients were statistically significant, p < .001. The estimates of the variation across the individually-estimated growth curves were statistically si gnificant for both intercept and slope, p < .001, indicating that there was cons iderable variation among indi vidual students on both their initial pre-test scores and growth rate. Additionally, the average reliabilities for initial status and slope were .93 and .92, respectively. There was a slight positive correlation between initial status and slope ( r = .27), suggesting that studen ts who had higher pre-test scores made more growth over time. Table 29 Unconditional Model for First Grade NWF Coefficient Variance Component Standard Error df t-ratio Chisquare Fixed Effect Initial Status/Intercept 15.59 1.58 9.85* Slope/Time 3.96 0.40 9.84* Random Effect Initial Status 86.03 36 506.97* Slope 5.49 36 426.46* Reliability of OLS regression coefficient estimate Initial Status .93 Slope .92 *p < .001 The results indicated that the intercep ts and slopes differ significantly across students; therefore, a second HLM analys is was conducted introducing the predictor variable, program. The results of th is analysis are displayed in Table 30.

PAGE 91

84 Table 30 Conditional Model for First Grade NWF Intercept/Initial Status Slope Coefficient Standard Error p-value Coefficient Standard Error p-value Earobics vs. Control -1.34 3.93 0.735 3.61 0.79 0.000 Lexia vs. Control -3.18 4.01 0.432 2.21 0.81 0.010 Earobics vs. Lexia 1.85 3.93 0.641 1.41 0.79 0.085 The results indicate that there were no statistically significan t differences between the control group and the experimental groups or between the two experimental groups on initial status, p > .05. There were, however, sta tistically significant differences between the control gr oup and Earobics group, p < .001 and the control group and Lexia group, p = .01 on slope. Results show that the Lexia group scores on NWF increased by 2.21 more units than the control group at each progress monitoring point. Additionally, the Earobics groups scores on NWF increased by 3.61 more units than the control group at each progress monitoring point. Figure 6 shows a comparison of the initial status and growth rate of the three groups on the depende nt variable of Nonsense Word Fluency. Table 31 displays the coefficients for the first HLM analysis for all first grade students for ORF without level-2 predictor variables. The mean initial status for first grade ORF was 9.15, indicating that on average students had an initial pre-test score of 9.15. The average slope for first grade OR F was 2.01, indicating that their scores increased at each progress monitoring point by 2.01. Both coefficients were statistically significant, p < .001. The estimates of the variation across the individually-estimated growth curves were statistically si gnificant for both intercept and slope, p < .001. This indicates that there was cons iderable variation among indivi dual students on both their

PAGE 92

initial pre-test scores and growth rate. The av erage reliabilities for in itial status and slope were .92 and .90, respectively. There was a posi tive correlation between initial status and slope ( r = .56), suggesting that students who ha d higher pre-test scores made more growth over time. 0 10 20 30 40 50 60 Pre-test12345Posttest Time (weeks)Letter-sounds Correct Per Minute Control Earobics Lexia Figure 6. Initial status and growth rate of the first grade Ea robics, Lexia, and control groups on Nonsense Word Fluency. Table 31 Unconditional Model for First Grade ORF Coefficient Variance Component Standard Error df t-ratio Chisquare Fixed Effect Initial Status/Intercept 9.15 1.17 7.81* Slope/Time 2.01 0.29 6.83* Random Effect Initial Status 46.63 36 443.01* Slope 2.88 36 363.01* Reliability of OLS regression coefficient estimate Initial Status .92 Slope .90 *p < .001 85

PAGE 93

86 A second HLM analysis was conducted introducing the pred ictor variable, program, allowing the examination of differe nces between the programs on the intercept and slope. The results of this an alysis are displayed in Table 32. Table 32 Conditional Model for First Grade ORF Intercept/Initial Status Slope Coefficient Standard Error p-value Coefficient Standard Error p-value Earobics vs. Control 2.39 2.89 0.415 2.39 0.61 0.001 Lexia vs. Control -0.21 2.95 0.943 1.15 0.62 0.073 Earobics vs. Lexia 2.60 2.89 0.415 1.23 0.61 0.051 The results indicate that there were no statistically significant differences between the control group and the experimental groups or between the two experimental groups on initial status, p > .05. There were statistically significant differences between the control group and Earobics group, p = .001, and the Earobics group and Lexia group, p = .05, on slope. Results show that the Earobi cs group gained 1.23 more words read correctly per minute at each progress monitoring point than the Lexia group and gained 2.39 more words read correctly per minute at each progres s monitoring point than the control group. Figure 7 shows a comparison of the initial status and growth rate of the three groups on the dependent variable of Or al Reading Fluency (ORF). Overall, the Earobics group showed a greater rate of growth when compar ed to both the control group and the Lexia group.

PAGE 94

0 5 10 15 20 25 30 Pretest 12345Posttest Time (weeks)Words Read Correctly Per Minute Control Earobics Lexia Figure 7. Initial status and growth rate of the first grade Ea robics, Lexia, and control groups on Oral Reading Fluency. Research Question Four: Kindergar ten and First Grade Comparison To answer the fourth research question (i.e., Are there outcome differences in early literacy skills between kindergarten and first grade students among the phonological awareness training programs and the control group?), the analyses that were completed for the first three questions were compared ac ross kindergarten and first grade. The two variables that were common across kindergarte n and first grade (i.e., LNF and PSF) were used in the comparison. Letter Naming Fluency was the first variab le in common across both grades. On the LNF variable for the kindergarten student s, the results of th e ANCOVA indicated no statistically significant di fferences among the adjusted post-test means of the experimental and control groups, p > .05. Additionally, a small effect size was noted when comparing the Lexia group post-test m ean to the control group post-test mean, d = 0.33, and the Earobics group post-test mean to the Lexia group post-test mean, d = 0.31. 87

PAGE 95

88 The HLM analysis indicated that the Earobi cs group gained 1.18 letters named correctly per minute more than the control group at each progress monitoring point. The Lexia group gained 0.88 letters named correctly per minute more than the control group. These differences, however, were not consider ed to be statistically significant, p > .05. The results of the ANCOVA indicated statistically significant differences among the adjusted post-test means of the firs t grade experimental and control groups, p = .001. Specifically, statistically signi ficant differences were found between the Earobics group and control group, p = .001, as well as between the Le xia group and the control group, p < .05. Additionally, a large effect size wa s found when comparing the Earobics group post-test mean to the control group post-test mean, d = 1.25. A medium effect size was found when comparing the Lexia group post-te st mean to the control group post-test mean, d = 0.72, and a small effect size was found when comparing the Earobics group to the Lexia group, d = 0.33. The results of the HLM analysis suggested similar findings. Specifically, the Earobics group gained 2.10 le tters named correctly per minute than the control group and the Lexia group gained 1.57 le tter names correctly per minute than the control group. Both of these were consid ered to be statistically significant, p =.001 and p < .01, respectively. Overall, the same patterns emerged when comparing the results of the kindergarten and first grade students on the LNF variable. For both grade levels, the Earobics group made more gains and showed better outcomes than both the Lexia and control groups. Additionally, the Lexia group demonstrated mo re gains and better outcomes than the control group. Although si milar results occurred across both grades, these results were only considered to be statistically significant for first grade students.

PAGE 96

89 The second variable in common betw een the grades was the Phoneme Segmentation Fluency variable. The kinderg arten students demonstrated a very small range between the PSF pre-test scores and post-test scores. Specifically, all groups scored zero on the pre-test measures and the maximum score on the post-test measure was a two. Due to this restriction of range, a meaningful comparison cannot be made across grade levels for the PSF variable. Teacher and Student Survey Results To gather information regarding student enjoyment of the computer programs and also the teacher rating of acceptability of the programs, surveys were administered at the conclusion of the intervention period. This section describes the surveys used with the teachers and students as well as th e results of the surveys. The student surveys consisted of three questions. Students responded to the three questions posed in the survey by pointing to a sad face representing Not at All (scored as 0), a neutral face representing A Little (s cored as 1), or a happy face representing A Lot (scored as 2). Total score ranges from 0 to 6, with higher scores indicating high acceptability. In this study, Cronbachs al pha for this scale was 0.78. Nunnally (1978) has indicated 0.7 to be an acceptable reliability coefficient. Student survey results including mean, standard deviation, minimum score, and maximum score are presented in Table 33. In response to the first question (i.e., How much did you like to do the computer activities?), all first grade students in both the Earobics and Lexia groups responded with the most positive score. Similarly, the st udents in the kindergarten Lexia group all responded with the most positive score to the first question. The mean score of the

PAGE 97

90 kindergarten Earobics group was 1.92 with a minimum score of one and a maximum score of two. The high mean score would indicate that the majority of students in this group reported that they liked doing the computer activities A Lot. In response to the second question (i.e., How much did you like leaving your class to do the computer activities ?), the first grade Earobics first grade Lexia, and kindergarten Lexia groups responded si milarly with means of 1.77, 1.75, and 1.85, respectively. The kindergarten Earobics gr oup responded overall less positively to this question, with a mean of 1.38, a minimum score of zero and a maximum score of two. One student in this group responded Not at A ll to this question, while seven students in this group responded A Little. The response to the third question (i.e ., How much would you like to have the computer activities at home?) were sim ilar across the kindergarten Lexia group, first grade Earobics group, and first grade Le xia group, with means of 1.92, 1.92, and 2.00, respectively. All groups had minimum scores of one and maximum scores of two. The mean for the kindergarten Earobics group was 1.69. Four students in this group indicated that they would like to have the programs at home A Little. Table 33 Student Survey Results Kindergarten First Grade Earobics (N=13) Lexia (N=13) Earobics (N=13) Lexia (N=12) Question One Mean 1.92 2.00 2.00 2.00 SD .28 .00 .00 .00 Min 1.00 2.00 2.00 2.00 Max 2.00 2.00 2.00 2.00

PAGE 98

91 Table 33 (Continued) Student Survey Results Kindergarten First Grade Earobics (N=13) Lexia (N=13) Earobics (N=13) Lexia (N=12) Q u e s t i o n T w o Mean 1.38 1.85 1.77 1.75 SD .65 .38 .44 .45 Min .00 1.00 1.00 1.00 Max 2.00 2.00 2.00 2.00 Question Three Mean 1.69 1.92 1.92 2.00 SD .48 .28 .28 .00 Min 1.00 1.00 1.00 2.00 Max 2.00 2.00 2.00 2.00 Total Mean 5.00 5.77 5.69 5.75 SD 1.08 .44 .63 .45 Min 3.00 5.00 4.00 5.00 Max 6.00 6.00 6.00 6.00 Note Question One = How much did you like to do th e computer activities? Question Two = How much did you like leaving your class to do the computer activities? Question Three = How much would you like to have the computer activities at home? Intervention acceptability was assessed fr om the teachers perspective utilizing a survey adapted from the Intervention Rating Profile 15 (IRP-15; Martens, Witt, Elliott, & Darveaux, 1985). The survey contains ten it ems rated on a 6-point Likert-type scale (1 = Strongly Disagree, 6 = Strongl y Agree). Total score ranges from 10 to 60, with higher scores indicating high acceptability. In this study, Cronbachs alpha for this scale was 0.79. All five kindergarten teach ers and all six first grade teachers had at least one student in their class on each of the programs; therefore, each kindergarten and first grade teacher completed the survey for both the Earobics program and the Lexia program. Teacher survey results including mean, st andard deviation, minimum score, and maximum score are presented in Table 34.

PAGE 99

92 The total mean scores for both kindergarten groups (i.e., 46.00) and both first grade groups (i.e., 46.33) were similar and indicated high overall accep tability among the teachers for both kindergarten and first grad e students. The two questions rated the lowest by the kindergarten teachers were T his intervention should prove effective in changing the students academic skill deficit and I liked the procedures used in this intervention, each earning mean scores of 4.00. The two questions rated the lowest by the first grade teachers were I would sugge st the use of this intervention to other teachers and I liked the procedures used in this intervention, earning mean scores of 4.00 and 4.17, respectively. An interesting findin g is that all teachers rated all questions the same across the two programs, indicating that they did not distinguish between two programs when rating their acceptability. Table 34 Teacher Survey Results Kindergarten First Grade Earobics (N=5) Lexia (N=5) Earobics (N=6) Lexia (N=6) 1. This would be an acceptable intervention for the students academic skill deficit. Mean 4.80 4.80 5.00 5.00 SD 1.10 1.10 .63 .63 Min 3 3 4 4 Max 6 6 6 6 2. This intervention should prove effective in changing the students academic skill deficit. Mean 4.00 4.00 4.50 4.50 SD .71 .71 .84 .84 Min 3 3 3 3 Max 5 5 5 5 3. I would suggest the use of this intervention to other teachers. Mean 4.20 4.20 4.00 4.00 SD .84 .84 .63 .63 Min 3 3 3 3 Max 5 5 5 5

PAGE 100

93 Table 34 (Continued) Teacher Survey Results Kindergarten First Grade Earobics (N=5) Lexia (N=5) Earobics (N=6) Lexia (N=6) 4. The students academic skill deficit is severe enough to warrant use of this intervention. Mean 4.80 4.80 5.00 5.00 SD .45 .45 .63 .63 Min 4 4 4 4 Max 5 5 6 6 5. I would be willing to use this intervention in the classroom setting. Mean 5.00 5.00 5.00 5.00 SD .71 .71 .63 .63 Min 4 4 4 4 Max 6 6 6 6 6. This intervention would not result in negative side-effects for the student. Mean 4.80 4.80 5.00 5.00 SD .45 .45 .63 .63 Min 4 4 4 4 Max 5 5 6 6 7. This intervention would be appropriate for a variety of students. Mean 4.80 4.80 4.50 4.50 SD .45 .45 .84 .84 Min 4 4 3 3 Max 5 5 5 5 8. This intervention is consistent with those that I have used in the classroom setting. Mean 4.80 4.80 4.50 4.50 SD .45 .45 .84 .84 Min 4 4 3 3 Max 5 5 5 5 9. I liked the procedures used in this intervention. Mean 4.00 4.00 4.17 4.17 SD .00 .00 .75 .75 Min 4 4 3 3 Max 4 4 5 5

PAGE 101

94 Table 34 (Continued) Teacher Survey Results Kindergarten First Grade Earobics (N=5) Lexia (N=5) Earobics (N=6) Lexia (N=6) 10. Overall, this intervention would be beneficial for the student. Mean 4.80 4.80 4.67 4.67 SD .45 .45 .82 .82 Min 4 4 3 3 Max 5 5 5 5 Total Score Mean 46.00 46.00 46.33 46.33 SD 5.29 5.29 6.59 6.59 Min 37 37 34 34 Max 51 51 54 54

PAGE 102

95 Chapter Five Discussion The purpose of this study was to examine the effects of two computer-delivered phonological awareness training programs, Earobi cs Step 1 and Lexia Early Reading, on the early literacy skills of st udents at risk for reading failu re. This study also aimed to provide information about whether or not th e effects of the computer programs were different across kindergarten and first grade. Data were collected before, during, and following a five-week interventi on period in which students us ed the computer programs for 20 minutes daily. The purpose of this chapter is to discuss the findings and implications of the study, which are presented in four sections. In the first section, the results of the study are presented at each grad e level. Comparisons across grade levels and the results of the teacher and student su rveys also are discusse d in this section. Second, practical implications of this study are discussed. Limitations of the current study are presented in the third section. Finall y, the contributions made to the literature by the current study as well as directions for future research are pres ented in the fourth section. Major Findings of This Study Data in this study were analyzed in three ways for each dependent measure at each grade level. First, comparisons among the adjusted mean post-test scores were made across the three groups (i.e., Earobics, Lexia, and control) to determine whether differences existed in outcomes, while cont rolling for any differences that may have

PAGE 103

96 occurred in the pre-test scores. Second, effect sizes were computed for the post-test scores as a descriptive measure to assess th e magnitude of differences between groups. Third, comparisons of intercepts and slopes we re made to determine whether differences existed in initial status or rate of grow th among the three groups. Assumptions were violated for some analyses conducted in this study; therefore it is important to examine all the information for each variable to dr aw conclusions about the effects of the computer programs across groups. Kindergarten findings This study found that significant differences existed among the three groups on both outcomes and gr owth rates for the first kindergarten variable examined, Initial Sounds Fluency. Specifically, the Earobics group adjusted mean post-test score was significantly higher than the control group adjusted mean posttest score. Additionally, the gains made by the Earobics group were significantly higher than the control group over the course of th e study. Similarly, the Lexia group adjusted mean post-test score was significantly higher than the control group adjusted mean posttest score and the gains made by the Lexia group were significantly higher than the control group over the course of the study. The outcomes between the Earobics group and Lexia group were not significantly differen t; however, the gains made over time were significantly different, with th e Earobics group making higher gains than the Lexia group. Additionally, while a large effect size wa s found between the Lexia and control group post-test scores, the largest effect size be tween post-test scores was found between the Earobics group and the control group. This means that although bot h computer programs appear to improve phonemic awareness skil ls as measured by DIBELS Initial Sounds Fluency, the Earobics program had a great er impact than the Lexia program.

PAGE 104

97 The results of this study did not indicat e such significant differences among the three groups on the second vari able examined for kindergar ten, Letter Naming Fluency. No significant differences were found am ong the groups on post-test adjusted mean scores or slope. A medium effect size wa s found when comparing the Earobics group to the control group on post-test scores, while th e effect sizes compar ing the Lexia group to the control group and the two experimental groups were both small. When examining growth rate, the Earobics group gained more letters named correctly per minute over the course of the study than both the Lexia group and the control group. Therefore, while the Earobics group appeared to have a greater impact than the Lexia group and the control group, this effect was not as significant on the kindergarten Letter Naming Fluency variable as was demonstrated on the Init ial Sounds Fluency variable. A possible explanation for this is that the kindergar ten core curriculum largely emphasizes the acquisition of letter names dur ing the beginning of the school year when the study was conducted. The increased instruction in this sk ill for all students may have contributed to the lack of significant findings among groups for this variable. The third variable examined for kindergarten students was Phoneme Segmentation Fluency. The results of the anal yses conducted for this variable indicated that both the Earobics and Lexia groups demonstrated more positive outcomes and better gains over time than the control group. No differences were noted between the two experimental groups. These analyses, however, need to be interpreted with caution and with regard for practical signi ficance due to the floor effect demonstrated on this variable for kindergarten students. Specifically, all gr oups scored zero on the pre-test measure for this variable. Additionall y, throughout the data collected for progress monitoring

PAGE 105

98 purposes as well as post-test da ta, the maximum score was two. This restriction of range cautions making strong conclusions about th e impact these two programs had on the Phoneme Segmentation Fluency variab le for kindergarten students. First grade findings The first variable examined for first grade students was Letter Naming Fluency. The results of the anal yses conducted for this variable indicated that the post-test adjusted mean score fo r the Earobics group was significantly higher than the control group. The post-test adjust ed mean score for the Lexia group also was significantly higher than the c ontrol group. Similarly, gain scores for both the Earobics and Lexia groups were significantly highe r than the control group. While both experimental groups outperformed the c ontrol group, the differences between the Earobics group and the control group were gr eater than the differen ces between the Lexia and the control group on both post-test scores and rate of growth. This finding is additionally supported by the large effect size found when comparing the Earobics group to the control group versus the medium ef fect size found when comparing the Lexia group to the control group. Phoneme Segmentation Fluency was the sec ond variable examined for first grade students. The results of the analyses conduc ted for this variable indicated that the Earobics program was superior to both the Lexia program and the control group when examining outcomes and rate of gains. Significant differences were found among posttest adjusted mean scores when comparing the Earobics group to the Lexia group and the control group. Significant di fferences also were found when comparing the Lexia group to the control group; however these differen ces were not as great. Effect sizes demonstrated support for this finding, with a large effect size found between the Earobics

PAGE 106

99 group and the control group as well as betw een the Earobics group and Lexia group. A small effect size was found when compari ng the Lexia group to the control group. Comparisons of slopes also confirmed the superior performance of the Earobics group when compared to the co ntrol and Lexia groups. The third variable examined for first grade was Nonsense Word Fluency. The results of the analyses for this variable s uggested that the Earobi cs group and the Lexia group performed better than the control group on both outcomes and rate of growth. While both experimental groups outperformed the control group, the differences between the Earobics group and control group were grea ter than the differences between the Lexia group and control group on both adjusted mean pos t-test scores and slope. Effect sizes computed for comparing each experimental group to the control group were both considered to be large; however the effect size found when comparing the Earobics group to the control group was larger than the e ffect size found when comparing the Lexia group to the control group. The final variable examined for first grade was Oral Reading Fluency. The results of the analyses for this variable in dicated that the Earobics program again made more gains over time and demonstrated better adjusted mean post-test scores than the control group. The differences between th e Lexia group and control group were not significant for either outcomes or rate of gr owth. When comparing the Earobics group to the Lexia group, significant differences were not noted in adjusted mean post-test scores; however, the difference in slope approached significance. The la rge effect size found when comparing the Earobics to the control group supports the finding that the Earobics

PAGE 107

100 group was superior to both the Lexia group and the control group as measured by the Oral Reading Fluency measure. Kindergarten and first grade comparison Letter Naming Fluency (LNF) and Phoneme Segmentation Fluency (PSF) were the common variables across the two grade levels. On the LNF variable, the Earobics group made more gains and showed better outcomes than both the Lexia and control group s across both grade levels. Additionally, the Lexia group demonstrated more gains a nd better outcomes than the control group. Although similar patterns in performance occurre d across both grades, these results were only considered to be statistically si gnificant for first grade students. On the PSF variable, the first grade Earobi cs group performed significantly better than both the first grade Lexi a and control groups on the adju sted mean post-test scores and slope. The results of the analyses for the PSF variable for the kindergarten students indicated better performance by both the Earo bics group and Lexia group when compared to the control group. However, as stated pr eviously, the conclusions that can be made when comparing the kindergarten and fi rst grade students on Phoneme Segmentation Fluency are limited due to the restriction of range demonstrated by kindergarten students on this variable. On the whole, Earobics has demonstrated a positive impact on the development of early literacy skills for students at-risk for reading failure. Student s utilizing the Lexia program also demonstrated better outcomes than the control group; however, the gains made by the students using the Earobics prog ram were stronger and more consistent across variables. Although th ere is more overall data to support this finding for first grade students than for kindergarten stude nts, the results for kindergarten students

PAGE 108

101 support this conclusion as well. The Earobi cs Step 1 and Lexia Early Reading programs both focus on training students in various ar eas of phonological awareness. When the two programs are compared regarding skill in struction, both programs offer training in rhyming, blending, segmenting, and identificatio n of beginning and ending sounds. In addition to this, the Earobics Step 1 program includes training in phoneme discrimination, auditory atten tion and memory, and the identi fication of medial sounds. The multifaceted instruction provided by the Earobics Step 1 program may help explain the superior performance of the students w ho utilized the program in this study. Consumer satisfaction survey Student and teacher su rveys were completed to gather information about the acceptability of the computer intervention. Results of the student survey indicated that overall, the majority of student s enjoyed using the computer programs A Lot and would like to have th e programs at home. Three out of four groups expressed that they did not mind leav ing their classroom to use the computer programs, while a few students in one group (i .e., kindergarten Earobics) reported that they did not like leaving the classroom at all or only liked leaving the classroom A Little. These students expressed that they would prefer to have the computer activities in their classrooms. During the study, many of the kindergarten students returned to class following their designated time in the com puter lab for only a few moments before leaving the classroom again to attend related art classes. These multiple transitions at the very beginning of the school day may have contributed to the lower ratings by these students on this item. The results of the teacher survey indicat ed that overall teachers felt that the computer programs were an acceptable intervention for the students academic skill

PAGE 109

102 deficit. Responses were similar across ki ndergarten and first grade teachers. An interesting finding was that the teachers did not appear to distinguish between the two programs when rating their acceptability. Th ese results mean that on the whole, the computer programs were considered to be an acceptable intervention by both the student participants and their teachers. Practical Implications Early identification of students at-risk for reading difficulties is critical to improving the students opportunity to become a successful reader. Equally important is differentiating effective interv entions to put into place once students at-risk for reading failure are identified. The students who util ized the computer programs in this study increased their early literacy skills more th an the students who did not use the computer programs. Additionally, none of the student s who utilized the Earobics or Lexia programs during the study continued to fall within the high risk category on DIBELS benchmarks at the conclusion of the study. This is important because the use of technology and computer software continues to grow in both home s and schools and yet the systematic evaluation of these program s is limited compared to the number of programs available for use. Although both the Earobics and Lexia pr ograms in this study had positive effects on the devel opment of early literacy skills in students at risk for reading failure, the Earobics program dem onstrated superior results across DIBELS measures and across grade levels. What th is means for schools is that students who utilize the Earobics program as an interven tion have a greater chance of improving early literacy skills and decreased ri sk for future reading failure.

PAGE 110

103 Limitations There were several limitations of the current study. The first is related to generalizability. To increase the implementation integrity of the study, it was conducted at one school site. Due to this constraint, the results of this study have limited external validity. Conducting the study at one school site also impacted the sample size, which was small (i.e., 12 to 13 students per group). This study also targeted a specific group of students (i.e., kindergarten a nd first grade students identified as at-risk for reading failure), which increased the likelihood of a small sample size and also impacted the generalizability of the findings to other groups of students. The small sample size also decreased the power of the analyses conducted in the st udy. Methodological limitations include the limited usable measures for th e kindergarten students and therefore limited useable measures available to compare resu lts across grade levels. Specifically, the DIBELS Phoneme Segmentation Fluency (PSF) measure was found to have a restriction of range for the kindergarten students, allowi ng for meaningful conclusions to be made on only two dependent variables for kindergar ten students. Additionally, due to the restriction of range for PSF, only one variable was able to be utili zed in the comparison of kindergarten and fi rst grade students. Contribution to the Literature a nd Directions for Future Research Research in the area of early literacy computer software is recent in history and limited in scope. The current study contri butes to the body of current research by exploring a comparison between two progr ams that assert to improve phonological awareness and early literacy skills. While many studies have been conducted to determine the effects of util izing phonological awareness co mputer programs compared

PAGE 111

104 to no use of computer programs (e.g., Hecht & Close, 2002; Foster et al., 1994; Lonigan et al., 2003) or compared technology trai ning to teacher-training (e.g., Pokorni, Worthington & Jamison, 2004; Mitchell & F ox, 2001), few studies have evaluated the impact of two computer programs against one an other. The results of this study indicated that the Earobics program was superior to the Lexia program in increasing the early literacy skills of students who are at-risk for reading failure. Given the limited window of opportunity available to intervene with students who are deficient in early literacy skills, it is important to identify programs with the greatest impact. Future research should continue to systematically evaluate specific computer software to aid in establishing effective interventions for students at-risk for reading failure. In addition to the many directions that fu ture research could take in this area, there are other questions rega rding phonological awareness training programs, out of the scope of this study, that could be addressed in future research on this subject. Future studies should include a larger sample size across several schools to increase external validity. Future studies should also extend the length of the intervention period to determine whether longer exposure to the computer programs would impact the outcomes. Additionally, longitudinal data would be beneficial to expl ore. Specifically, it would useful to know whether the differences between groups in this study held up over time. Future research designed to collect data months after the in tervention period ended would allow these questions to be answered. Future research should also include other measures, in addition to the DIBELS, to assess early literacy skills. For example, The Phonological Awarene ss Test (Robertson &

PAGE 112

105 Salter, 1997) is appropriate for use with kindergarten students and measures rhyming, segmentation, isolation, deletion, substitu tion, and blending. Utilizing additional measures would allow for a broader picture of a students early literacy skills, especially with regard to kindergarten students.

PAGE 113

106 References Adams, M.J. (1990) Beginning to read: Thinking and learning about print Cambridge, MA: MIT Press. Al Otaiba, S., & Fuchs, D. (2002). Characteristics of children who are unresponsive to early liter acy intervention: A review of the literature. Remedial and Special Education, 23(5), 300-316. Barker, T.A., & Torgesen, J.K. (1995). An evaluation of computer-assisted instruction in phonological awareness with below average readers. Journal of Educational Computing Research 13, 89-103. Brennan, F., & Ireson, J. (1997). Training phonological awareness: A study to evaluate the effects of a program of metalinguistic games in kindergarten. Reading and Writing: An Interdisciplinary Journal 9, 241-263. Byrne, B., & Fielding-Barnsley, R. ( 1995). Evaluation of a program to teach phonemic awareness to young children: A 2an d 3-year follow-up and a new preschool trial. Journal of Educational Psychology, 87(3), 488-503. Cognitive Concepts, Inc. (2003). Earobics Step 1 Clinic Software Users Guide Evanston, IL. Erickson, G. C., Foster, K.C., Foster, D. F., Torgesen, J.K., & Packer, S. (1992). DaisyQuest [Computer Software]. Sco tts Valley, CA: Great Wave Software.

PAGE 114

107 Fletcher, J.M., Shaywitz, S.E., Shankweiler, D.P., Katz, L., Liberman, I.Y., Stuebing, K.K., Francis, D.J., Fowler, A.E., & Shaywitz, B.A. (1994). Cognitive profiles of reading disability: Comparisons of discrepancy and low achievement definitions. Journal of Educational Psychology 86, 6-23. Foster, K.C., Erickson, G.C., Foster, D. F., Brinkman, D., & Torgesen, J.K. (1994). Computer administered instruction in phonological awareness: Evaluation of the DaisyQuest program. The Journal of Research and Development in Education 27, 126137. Good, R.H., Gruba, J., & Kaminski, R.A. ( 2002). Best practices in using Dynamic Indicators of Basic Early Literacy Skills (D IBELS) in an outcomes-driven model. In A. Thomas & J. Grimes (Eds.) Best practices in school psychology IV. Bethesda, MD. National Association of School Psychologists. Good, R. H., & Kaminski, R. A. (2002). Nonsense Word Fluency. In R. H. Good & R.A. Kaminski (Eds.) Dynamic Indicators of Basi c Early Literacy Skills (6th ed.). Eugene, OR: Institute for the Development of Educational Achievement. Available: http://dibels.uoregon.edu/ Good, R. H., Kaminski, R. A., & Dill, S. (2002). DIBELS Oral Reading Fluency. In R. H. Good & R.A. Kaminski (Eds.) Dynamic Indicators of Basic Early Literacy Skills (6th ed.). Eugene, OR: Institute for the Development of Educational Achievement. Available: http://dibels.uoregon.edu/

PAGE 115

108 Good, R. H., Kaminski, R. A., & Smit h, S. (2002). Phoneme Segmentation Fluency. In R. H. Good & R.A. Kaminski (Eds.) Dynamic Indicators of Basic Early Literacy Skills (6th ed.). Eugene, OR: Institute for the Development of Educational Achievement. Available: http://dibels.uoregon.edu/ Good, R. H., Laimon, D., Kaminski, R. A., & Smith, S. (2002). Initial Sounds Fluency. In R. H. Good & R.A. Kaminski (Eds.) Dynamic Indicators of Basic Early Literacy Skills (6th ed.). Eugene, OR: Institute for the Development of Educational Achievement. Available: http://dibels.uoregon.edu/ Good, R.H., Simmons, D.C., & Kameenui E.J. (2001). The importance and decision-making utility of a continuum of fluency-base d indicators of foundational reading skills for third-grade high-stakes outcomes. Scientific Studies of Reading 5(3), 257-288. Good, R.H., Simmons, D.C. Kameenui, E.J., Kaminski, R.A., & Wallin, J. (2002). Summary of decision rules for intensive, strategic, and benchmark instructional recommendations in kindergarten through third grade (Technical report #11). Eugene, OR: University of Oregon. Good, R.H., Simmons, D.C., & Smith, S.B. (1998). Effective academic interventions in the United States: Evalua ting and enhancing the acquisition of early reading skills. School Psychology Review 27(1), 755-766. Good, R. H., & Kaminski, R. A. (Eds.). (2002). Dynamic Indicators of Basic Early Literacy Skills (6th ed.). Eugene, OR: Institute for the Development of Educational Achievement.

PAGE 116

109 Hecht, S.A., & Close, L. (2002). Emerge nt literacy skills and training time uniquely predict variability in responses to phonemic awareness training in disadvantaged kindergarteners. Journal of Experimental Child Psychology 82, 93-115. Juel, C. (1988). Learning to read and write: A longitudinal study of fifty-four children from first through fourth grade. Journal of Educational Psychology 80, 437447. Kaminski, R. A., & Good, R. H. (2002). Letter Naming Fluency. In R. H. Good & R.A. Kaminski (Eds.) Dynamic Indicators of Basic Early Literacy Skills (6th ed.). Eugene, OR: Institute for the Development of Educational Achievement. Available: http://dibels.uoregon.edu/ Kozminsky, L., & Kozminsky, B. G. (1995). The effects of early phonological awareness training on reading success. Learning and Instruction, 5, 187-201. Lexia Learning Systems, Inc. (2004). Lexia Early Reading Teachers Manual Lincoln, MA. Lindamood, C. H., & Lindamood, P.C. (1998). Lindamood Phoneme Sequencing Program (LIPS). Austin, TX: PRO-ED. Lonigan, C., Driscoll, K. Phillips, B., Cantor, B., Anthony, J., & Goldstein, H. (2003). A computer-assisted instruction phonol ogical sensitivity program for preschool children at-risk for reading problems. Journal of Early Intervention 25(4), 248-262. Lyon, G. R., & Chhabra, V. (1996). The curre nt state of science and the future of specific reading disability. Mental Retardation and Developmental Disabilities Research Reviews 2, 2-9.

PAGE 117

110 Martens, B. K., Witt, J. C., Elliott, S. N., & Darveaux, D. (1985). Teacher judgments concerning the acceptability of school based interventions. Professional Psychology: Research and Practice 16, 191-198. Mitchell, M.J., & Fox, B. J. (2001). Th e effects of computer software for developing phonological awareness in low-progress readers. Reading Research and Instruction 40 (4), 315-332. National Center for Edu cation Statistics. (2003). Report card for the nation and the states. Washington, DC: U.S. Department of Education, Office of Educational Research and Improvement. National Reading Panel. (2000). Report of the National Reading Panel. Teaching children to read: An evidence-based assessment of the scientific research literature on reading and its implications for reading instruction: Reports of the subgroups. Bethesda, MD: National Institute of Child Health and Human Development. Nunnally, J.C. (1978). Psychometric theory (2nd ed.). New York: McGraw-Hill. Pokorni, J.L., Worthington, C.K., & Jami son, P.J. (2004). Phonological awareness intervention: Comparison of Fast ForWord, Earobics, and LiPS. The Journal of Educational Research 97(3), 147-157. Robertson, C., & Salter, W. (1997). The phonological awareness test. East Moline, IL: LinguiSystems. Schneider, W., Kuspert, P., Roth, E., & Vise, M. (1997). Shortand long-term effects of training phonological awareness in kindergarten: Evidence from two German studies. Journal of Experiment al Child Psychology 66, 311-340.

PAGE 118

111 Snider, V.E. (1997). The relationship between phonological aw areness and later reading achievement. The Journal of Educational Research 90, 203-210. Snow, C., Burns, M., & Griffin, P. (Eds.), (1998). Preventing reading difficulties in young children Washington DC: National Academy Press. Torgesen, J.K., & Barker, T. (1995). Com puters as aids in the prevention and remediation of reading disabilities. Learning Disability Quarterly 18, 76-88. Torgesen, J. K., & Bryant, B. R. (1994). Phonological Awareness Training for Reading Austin: Pro-Ed. Torgesen, J. K., & Mathis, P.G. (1999). Assessment and Instruction in Phonological Awareness Greenwood, C. (Ed.) Florid a Department of Education. Torgesen, J.K, Morgan, S.T., & Davis, C. (1992). The effects of two types of phonological awareness training on word leve l reading in kindergarten children. Journal of Educational Psychology 84, 364-370. Torgesen, J.K., Wagner, R.K., & Rashotte C.A. (1994). Longitudinal studies of phonological processi ng and reading. Journal of Learning Disabilities 27, 276-286. Wagner, R.K., Torgesen, J.K., Rashotte, C.A., Hecht, S.A., Barker, T.A., Burgess, S.R., Donahue, J., & Garon, T. (1997). Ch anging causal relations between phonological processing abilities and word-level reading as children develop from beginning to fluent readers: A five-year longitudinal study. Developmental Psychology 33, 468-479.

PAGE 119

112 Appendices

PAGE 120

Appendix A: Student Survey Student Survey Group _______________ Program _____________ How much did you like to do the computer activities? Not at All (0) A Little (1) A Lot (2) How much did you like leaving your cla ss to do the computer activities? Not at All (0) A Little (1) A Lot (2) How much would you like to have the computer activities at home? Not at All (0) A Little (1) A Lot (2) Comments: ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ 113

PAGE 121

Appendix B: Teacher Survey Teacher Survey Adapted from the Intervention Rating Profile 15 (IRP-15) 1 Please circle the grade you teach: K 1st The purpose of this questionnaire is to obtain information that will aid in the selection of classroom interventions. Please circle the number which best describes your agreement or disagreement with each statement. Strongly Disagree Disagree Slightly Disagree Slightly Agree Agree Strongly Agree 1. This would be an acceptable intervention for the students academic skill deficit. 1 2 3 4 5 6 2. This intervention should prove effective in changing the students academic skill deficit. 1 2 3 4 5 6 3. I would suggest the use of this intervention to other teachers. 1 2 3 4 5 6 4. The students academic skill deficit is severe enough to warrant use of this intervention. 1 2 3 4 5 6 5. I would be willing to use this intervention in the classroom setting. 1 2 3 4 5 6 6. This intervention would not result in negative side-effects for the student. 1 2 3 4 5 6 7. This intervention would be appropriate for a variety of students. 1 2 3 4 5 6 8. This intervention is consistent with those that I have used in the classroom setting. 1 2 3 4 5 6 9. I liked the procedures used in this intervention. 1 2 3 4 5 6 10. Overall, this intervention would be beneficial for the student. 1 2 3 4 5 6 T o be comple ted by PI: E a r Lex 114 1 Mar t ens, B.K., W i tt, J.C., Ellio tt, S.N., & Darv eaux D. (1 9 8 5 ) Te acher j u dgments co ncer ning the acceptability o f scho o l based inter v e n tio n s. Pr o f es sio n al Psy c holo gy : Resear c h and P r actice, 16, 1 9119 8.

PAGE 122

Appendix C: Informed Consent Parental Permission (Parental Consent) Social and Behavioral Sciences University of South Florida Information for Parents who are being asked to allow their child to take part in a research study The following information is being presented to help you decide whether or not you want to allow your child to be a part of a research study. Please read this carefully. If you do not understand anything, ask the person in charge of the study or the person obtaining your consent. Title of research study: The Effects of Comput er-Delivered Phonological Awareness Training on the Early Literacy Skills of Students At-Risk for Reading Failure Person in charge of study: Deanne Gale Where the study will be done: General Information about the Research Stud y The purpose of this research study is to investigate whether two computer programs (Earobics Step 1 and Lexia Early Reading) are effective in improving the early literacy skills of young children. Your child is bei ng asked to participate because according to recent district assessment information, your child is having difficulty with early literacy skills. Plan of Study If you give permission for your child to participate in the study, he or she will be assigned to one of three groups: Earobics Step 1, Lexia Early Reading, or teacher-designed interventions. The two computer program s (Earobics Step 1 and Lexia Early Reading) will be on computers located at Pinellas Park Elementary. If your child is assigned to either the Earobics Step 1 or Lexia Early Reading group, they will use the program for 20 minutes daily during school hours. This 20-minute period will not take place during the required 90-minute reading block. The intervention period will last 5 weeks and your childs progress will be monitored throughout the 5-week period. Progress will be monitored by administering the Dynamic Indi cators of Basic Early Literacy Skills (DIBELS) to your child one time per week throughout the study. Each administration of the DIBELS will take approximately five minutes. At the end of the 5 weeks, a determination will be made regarding the effectiveness of the programs and whether or not your child should continue to use the intervention. At the end of the study, your child 115

PAGE 123

116 Appendix C: Informed Consent (Continued) also will be asked to complete a brief survey indicating how much they enjoyed using the computer programs. Additionally, as a part of this study, demographic information including your childs name, student number, gender, race, and age will be collected by the principal investigator through a review of his or her school records. Payment for Participation You or your child will not be paid for your ch ilds participation in this study. Your child will be offered small tokens of appreciation (e.g., pencils, erasers, stickers) throughout the study for their completion of sessions on the computer programs. Potential Benefits of Taking Part in this Research Study By taking part in this research study, your child will receive additional intervention in an area that is difficult for him or her. Additionally, he or she will help increase our knowledge of computer programs that are effective in increasing early literacy skills. Risks of Being a Part of this Research Study As a part of this research study, your child will miss a portion of his or her time in the classroom. In order to minimize the impact of this, your childs teacher will be consulted to determine the time of day in which missing instruction will have the least impact. Confidentiality of Your Childs Records We will keep the records of this study private by insuring that only the principal investigator in the study and your childs teacher will have access to the records of your childs progress with each of the programs. Such records will be kept in a locked filing cabinet on school property. However, certain people may need to see your childs study records. By law, anyone who looks at your childs records must keep them confidential. The only people who will be allowed to see these records are: The study staff. People who make sure that we are doing the study in the right way. They also make sure that we protect your childs rights and safety: A. USF Institutional Review Board (IRB) and their staff B. Others may include: People at USF who oversee research; Florida Department of Health; and the United States Department of Health and Human Services (DHHS) The results of this study may be published. However, the data obtained from your child will be combined with data from other children in the publication. The published results

PAGE 124

117 Appendix C: Informed C onsent (Continued) will not include your childs name or any other information that would personally identif y your child in any way. Volunteering to Take Part in this Research Study Your decision to allow your child to participate in this research study must be completely voluntary. You are free to allow your child to participate in this research study or to withdraw him/her at any time. If you choose not to allow your child to participate or if you remove your child from the study, there will be no penalty or loss of benefits that you or your child are entitled to receive. Questions and Contacts If you have any questions about this research study, contact Deanne Gale at If you have questions about your childs rights as a person taking part in a research stu dy, you may contact the Division of Research Compliance of the University of South Florida at (813) 974-9343. Consent for Child to Take Pa rt in this Research Study I freely give my permission to let my child take part in this study. I understand that this is research. I have received a copy of this consent form. ________________________ ________________________ ___________ Signature of Parent Printed Name of Parent Date of child ta king part in study ________________________ ________________________ ___________ Signature of person Printed Name of person Date obtaining consent obtaining consent Statement of Person Obtaining Informed Consent: I certify that participants have been provided with an informed consent form that has been approved by the University of South Flor id as Institutional Review Board and that explains the nature, demands, risks, and benefits involved in participating in this study. I further certify that a phone number has been provided in the event of additional questions. ________________________ ________________________ ___________ Signature of person Printed Name of person Date obtaining consent obtaining consent

PAGE 125

E 1 E 2 E 3 E 4 E 5 E 6 E 7 L 1 L 2 L 3 L 4 L 5 L 6 L 7 Pick up from: Class Class Class Class Class Class Class Class Class Class Class Class Class Class Student Student Student Stude nt Student Student Student Student Student Student Stude nt Student Student Student 8:00 8:20 Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teach er Teacher Teacher Teacher Teacher Drop off to: Rm. 24 T Rm. 24 T Rm. 24 T Rm. 24 F Rm 24 W Rm. 24 F Rm 24 W Rm. 24 T Rm. 24 T Rm. 24 T Rm. 24 F Rm 24 W Rm. 24 T Rm. 24 T Pick up from: Class Class Class Class Class Class Class Class Class Class Class Class Class Class Student Student Student Stude nt Student Student Student Student Student Student Stude nt Student Student Student 8:25 8:45 Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teach er Teacher Teacher Teacher Teacher Drop off to: Mus M PE W/F Art T Mus M PE W/F Art T Mus M PE W/F Art T Mus M PE W/F Art T Mus M Art W PE T/F Mus M PE W/F Art T Mus M Art W PE T/F Mus M PE W/F Art T Mus M PE W/F Art T Mus M PE W/F Art T Mus M PE W/F Art T Mus M PE W/F Art T Mus M Art W PE T/F Mus M Art W PE T/F Pick up from: Class Class Class Class Class Class Class Class Class Class Class Class Class Class Student Student Student Stude nt Student Student Student Student Student Student Stude nt Student Student Student 8:50 9:10 Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teacher Drop off to: Rm. 24 W Class Class Rm. 24 T Rm. 24 T Rm. 24 T Rm. 24 T Rm. 24 W Rm. 24 T Rm. 24 T Rm. 24 W Rm. 24 W Class Rm. 24 T Pick up from: Class Class Class Class Class Class Class Class Class Class Rm. 24 F Rm. 24 F Class Class Student Student Student Stude nt Student Student Student Student Student Student Stude nt Student Student Student 9:15 9:35 Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teacher Teacher Drop off to: Class Class Class Class Mus M Art W PE T/F Class Mus M Art W PE T/F Class PE M/W/F Art T Class Class Class Class PE M/W/F Art T Appendix D: Rotation Schedule 118

PAGE 126

Appendix E: Computer Training Checklist Student Teacher P u t O n H e a d p h o n e s S i g n O f f P r o g r a m S ig n O n P r o g r a m Se l e c t A c t i vi t y U s e M o u s e t o N a v i g a t e A c t i v i t y T r a n s i t i on B e t w e e n A c t i vi t i e s 119

PAGE 127

120 Appendix F: Proctor Schedule 7:55 to 8:00 Pick up group 1 on Schedule a nd escort to computer lab Direct students to assigned computer & assist with setup, if needed Begin timer at 8:00 for 20 minutes 8:00 to 8:20 Walk around to assure that stud ents are using programs & answer procedural questions (Do not he lp with answers to program) At 8:20 inform students to stop activity and sign off program 8:20 to 8:25 Return Group 1 Pick up Group 2 on Schedule and escort to computer lab Direct students to assigned computer & assist with setup, if needed Begin timer at 8:25 for 20 minutes 8:25 to 8:45 Walk around to assure that stud ents are using programs & answer procedural questions (Do not he lp with answers to program) At 8:45 inform students to stop activity and sign off program 8:45 to 8:50 Return Group 2 Pick up Group 3 on Schedule and escort to computer lab Direct students to assigned computer & assist with setup, if needed Begin timer at 8:50 for 20 minutes 8:50 to 9:10 Walk around to assure that stud ents are using programs & answer procedural questions (Do not he lp with answers to program) At 9:10 inform students to stop activity and sign off program 9:10 to 9:15 Return Group 3 back to class Pick up Group 4 on Schedule and escort to computer lab Direct students to assigned computer & assist with setup, if needed Begin timer at 9:15 for 20 minutes 9:15 to 9:35 Walk around to assure that stud ents are using programs & answer procedural questions (Do not he lp with answers to program) At 9:35 inform students to stop activity and sign off program Return Group 4

PAGE 128

121 Appendix G: Student Activity Card Name: ______________________ Teacher: _______________________ Program/Group: _____________ Computer Number: _____________ Date: _______ Time: 20 __ Date: _______ Time: 20 __ Date: _______ Time: 20 __ Date: _______ Time: 20 __ Date: _______ Time: 20 __ Treasure Box Date: _______ Time: 20 __ Date: _______ Time: 20 __ Date: _______ Time: 20 __ Date: _______ Time: 20 __ Date: _______ Time: 20 __ Treasure Box Date: _______ Time: 20 __ Date: _______ Time: 20 __ Date: _______ Time: 20 __ Date: _______ Time: 20 __ Date: _______ Time: 20 __ Treasure Box Date: _______ Time: 20 __ Date: _______ Time: 20 __ Date: _______ Time: 20 __ Date: _______ Time: 20 __ Date: _______ Time: 20 __ Treasure Box Date: _______ Time: 20 __ Date: _______ Time: 20 __ Date: _______ Time: 20 __ Date: _______ Time: 20 __ Date: _______ Time: 20 __ Treasure Box