Read Microsoft Word - Frederick, Jessica Research Proposal Paper.doc text version

Saxon Phonics versus Harcourt Trophies 1 Running head: SAXON PHONICS VERSUS HARCOURT TROPHIES

Differences in Reading Fluency Between First-Grade Students Instructed With Saxon Phonics Versus Harcourt Trophies

Tennessee Technological University Instructor: Julie Baker Fall 2010

Saxon Phonics versus Harcourt Trophies 2 Abstract The purpose of this quantitative study was to examine the effects of a supplemental phonics program, Saxon Phonics, on the oral reading fluency of first-grade students at Jasper Elementary School located in Marion County, Tennessee. The design was descriptive, causal-comparative, and ex post facto. The independent variables were Harcourt Trophies and Saxon Phonics instructional programs, gender, and ability levels of low, middle, and high. The dependent variable was oral reading fluency as measured by AIMSweb, while the covariate was nonsense word fluency also measured by AIMSweb. The 2008-2009 and 2009-2010 AIMSweb data was collected with assistance from the literacy specialist upon approval from the Institutional Review Board in July 2010. SPSS ran a factorial analysis of covariance to analyze and interpret the data. The results of the analysis revealed a significant difference between the Harcourt Trophies and Saxon Phonics instructional programs. The significance level was set at p = .05.

Saxon Phonics versus Harcourt Trophies 3 Table of Contents Page Number Chapter One: Introduction Introduction Research Questions Significance of the Study Hypotheses Definitions Chapter Two: Review of Literature Introduction Harcourt Trophies and Saxon Phonics Other Instructional Programs and Reading Achievement Gender Ability Level Oral Reading Fluency Summary Chapter Three: Methodology Introduction Sample Design Measurement and Instrumentation Delimitations Materials 32 32 33 33 34 34 11 11 14 18 22 26 30 5 7 7 8 8

Saxon Phonics versus Harcourt Trophies 4 Data Collection Limitations Data Analysis/Statistical Procedures Chapter 4: Data Presentation and Analysis Introduction Presentation of Data Data Analysis Chapter 5: Conclusions, Discussion, and Recommendations Introduction Summary of Results Conclusions Discussion of Results Recommendations References 45 45 46 46 47 48 37 37 42 34 35 36

Saxon Phonics versus Harcourt Trophies 5 Chapter 1: Introduction Perhaps one of the most important things a child learns in school is how to read. Reading ability seems to affect every facet of life. It is no question of where students need to be when they leave first grade. Yet, the question of how to get them there still remains. Curriculum directors, principals, and teachers across the country continue to debate among a traditional, basal approach, an explicit, systematic phonics approach, or both to develop fluent readers. Establishing an approach to literacy, educators discover, is only the tip of the iceberg. Logical questions of which programs to incorporate and in what grade levels to incorporate them always seem to follow. These questions send teachers on a witch-hunt to find the programs that will yield the highest reading achievement. With the newly implemented Race to the Top legislation, the state of Tennessee struggles to keep up with the ever-increasing reading standards. The 2006-2007 Record of Achievement for all Tennessee fourth-grade students indicated only 27% proficient among the National Assessment of Educational Progress (NAEP) data (NCES, 2008). Jasper Elementary School, a rural school located in Marion County, Tennessee, faced the same problem. According to their Tennessee School Improvement Planning Process (TSIPP, 2009) manual, "TCAP scores indicate an area of need in language arts school-wide and AIMSweb scores in oral reading fluency and early literacy for first and second grade reading corroborate this conclusion" (p. 18). The data forced educators at Jasper Elementary to question whether the Harcourt Trophies instruction implemented during the 2008-2009 school year was adequate. The teachers debated if adding a supplemental phonics program would be appropriate at the first grade level. A report by the National Institute of Child Health and Human Development (NICHD, 2000) stated:

Saxon Phonics versus Harcourt Trophies 6 The most important grade for teaching phonics is thought to be first grade when formal instruction in reading typically begins in the United States. Children have foundational knowledge and are ready to put it to use in learning to read and write. In contrast, introducing phonics instruction in grades above first means that children who were taught to read in some other way may be required to switch gears in order to incorporate phonics procedures into their reading and writing. (p. 2-105) The purpose of this quantitative study was to examine the effects of the supplemental phonics program, Saxon Phonics, on the oral reading fluency of first-grade students at Jasper Elementary School. More specifically, this study determined if a difference existed between the 124 students instructed with only Harcourt Trophies basal reading program and the 172 students instructed with supplemental Saxon Phonics. Whether or not a significant difference existed would allow Jasper Elementary School to contemplate keeping or eliminating the program, as it is purchased at a considerable expense. Proponents of Saxon Phonics claim it is most beneficial when used in conjunction with a core reading program (FCRR, 2003). Strengths of the program as listed by the Florida Center for Reading Research (2003) included, "A multi-sensory approach and use of manipulatives, built-in assessments to monitor student progress and guide instruction, clearly stated learning objectives for every activity, and clearly linked instruction across components" (p. 4). However, Wicker found in his 2007 research that the group instructed with Saxon Phonics in combination with Harcourt Trophies did not have significant gains in reading achievement. Conversely, the research conducted by Leib (2001) resulted in the Saxon Phonics group making significantly higher gains in reading.

Saxon Phonics versus Harcourt Trophies 7 The National Reading Panel (NRP) stated, "Correlational studies have identified phonemic awareness and letter knowledge as the two best school-entry predictors of how well children will learn to read during the first two years of instruction" (NICHD, 2000, p. 2-9). Despite the countless research on various instructional approaches, a limited number of studies exist on specific instructional programs such as Saxon Phonics. The purpose of this study was to shed some light on the instructional programs of Saxon Phonics and Harcourt Trophies and their ability to increase reading achievement, one of the most difficult challenges facing Tennessee educators today. Research Questions · Will there be a difference in AIMSweb reading fluency scores between first- grade students who received supplemental Saxon Phonics instruction and students who received only Harcourt Trophies basal reading instruction, after adjusting for ability level? · Will this difference, if any, vary depending on gender and ability levels of low, middle, and high? Significance of the Study With reading being the ultimate goal of first grade, curriculum directors, principals, and teachers are constantly seeking instructional programs or methods that will result in reading achievement. Jones (2006) reiterated the need to choose research-based programs as mandated by No Child Left Behind. The problem lies in selecting effective programs that are within budget. This research studied the differences between students instructed with only a basal reading program and students instructed with a supplemental Saxon Phonics program. Of the literature reviewed, only two studies focused on Saxon Phonics. Consequently, the current study added to a body of research where few studies have been conducted.

Saxon Phonics versus Harcourt Trophies 8 Wicker (2007) stated the following, "Refill kits for Saxon Phonics K cost the county an additional $25 per kindergarten student per year, totaling approximately $12,000 per year" (p. 7). It can be seen from Wicker's research that supplemental programs come at a considerable expense. If research showed the traditional, basal reading approach is valuable without a supplemental phonics program, then school systems can find a better means of distributing funds. Additionally, this study is significant in that it addressed the theory of ability levels and examined the effects of Saxon Phonics across reading levels of low, middle, and high. As Agnew (2005), Sippola (1985), and Snell (2007) discovered, one program does not prove best for all ability levels. Therefore, this research allowed educators to determine if Saxon Phonics could be better utilized in small group instruction as opposed to whole group instruction. Hypotheses · There will be a difference in AIMSweb reading fluency scores between first- grade students who received supplemental Saxon Phonics instruction and students who received only Harcourt Trophies basal reading instruction, after adjusting for ability level. · This difference, if any, will vary depending on gender and ability levels of low, middle, and high. Definitions · Ability Level: The students were divided into thirds among ability levels of low, middle, and high as determined by oral reading fluency scores. The bottom portion comprised low ability, the middle portion comprised middle ability, and the top portion comprised high ability. Ability level was an independent variable in this study.

Saxon Phonics versus Harcourt Trophies 9 · AIMSweb: For purposes in this study, AIMSweb was the dependent variable or the instrumentation that measured first-grade students' oral reading fluency from the 20082009 and 2009-2010 school years. The AIMSweb website (2008) defined the assessment as follows: AIMSweb is a benchmark and progress monitoring system based on direct, frequent and continuous student assessment. The results are reported to students, parents, teachers and administrators via a web-based data management and reporting system to determine response to intervention. (Homepage) · Gender: Students were identified as male or female from the AIMSweb report as an independent variable for this study. · Harcourt Trophies: In this study, Harcourt Trophies was the basal reading program and method of reading instruction for all first-grade students from 2008-2010. · Nonsense Word Fluency (NWF): NWF is an early literacy AIMSweb measurement that was used as a covariate in this study. The NWF measurement scored the students based on the number of non-real words read correctly in one minute (AIMSweb, 2008). Fall scores entered in August of their first grade year were gathered for both the 2008-2009 and 2009-2010 participants. · Oral Reading Fluency: Oral Reading Fluency also known as Reading-Curriculum-Based Measurement (R-CBM) is the AIMSweb measurement in which students are scored by the number of words read correctly in one minute on a reading passage minus any errors (AIMSweb, 2008). It was the dependent variable in this study. · Saxon Phonics: Saxon Phonics, in this study, was the supplemental instructional program the students received in 2009-2010. The program involved explicit, scripted phonics

Saxon Phonics versus Harcourt Trophies 10 lessons in such areas as high-frequency words, phonemic awareness, and alphabetizing etc. (Saxon, 2010).

Saxon Phonics versus Harcourt Trophies 11 Chapter 2: Review of Literature Between the years 1985 and 2010, many studies have examined the effects of various instructional programs on reading achievement of male and female students. The diversity among the participants ranged from their ability levels to the communities in which they resided. Contrastingly, few authors studied Saxon Phonics and Harcourt Trophies, while no literature existed on the newly implemented AIMSweb diagnostic assessment program. Of the fifteen quantitative dissertations reviewed here, six were descriptive and nine were experimental in nature. The chapter is categorized by the variables of the current study and ordered as follows: Harcourt Trophies and Saxon Phonics, Other Instructional Programs and Reading Achievement, Gender, Ability Level, and Oral Reading Fluency. Harcourt Trophies and Saxon Phonics The first dissertation outlined here is most equivalent to the current study because it included the same key variables. The independent variables were reading program, either Dr. Cupp or a combination of Saxon Phonics K and Harcourt Trophies, and gender (Wicker, 2007). The dependent variable was literacy scores on the Georgia Kindergarten Assessment ProgramRevised (GKAP-R) (Wicker, 2007). The objective of Wicker's (2007) research was to determine a difference between the two reading programs and their effects on reading readiness. Wicker (2007) performed quasi-experimental pretest-posttest control group design research at a rural elementary school in Georgia to answer the following research questions: 1. What is the difference in class gain scores on the literacy section of the GKAP-R for students utilizing Phonics K/Harcourt basal reading series when compared to students utilizing Dr. Cupp readers combining eight literacy components?

Saxon Phonics versus Harcourt Trophies 12 2. What is the difference in gain scores on the GKAP-R among students in the two opposing programs (Phonics K/Harcourt basal reading series and Dr. Cupp Readers) for students of gender? (p. 13) One delimitation of the study as identified by Wicker (2007) was the fact that those students who scored in the 20th percentile on the Bracken test were purposefully placed in the Early Intervention Program (EIP) using Dr. Cupp Readers. Of the 43 students in the sample, 24 were in the EIP classes and 19 were in the general education class (Wicker, 2007). Participating teachers administered and transferred scores of the August, January, and April testing periods into a spreadsheet provided by the researcher (Wicker, 2007). Statistical Package for the Social Sciences (SPSS) ran independent samples t-tests to analyze the data (Wicker, 2007). Wicker (2007) found, at the .05 level, that Group B, the Dr. Cupp Readers, scored significantly higher (t = 3.40, p = .00) than Group A (Phonics K/Harcourt). The findings for gender resulted in both males (t = 2.76) and females (t = 3.20) in Group B having statistically higher gains (p = .01) (Wicker, 2007). Wicker (2007) stated, "Because both boys and girls appeared to have benefited equally from being in Group B, the null hypothesis of no difference based on gender was retained" (p. 88). These results imparted information to the current research of ability levels, considering Group B participants were identified as low achievers by the Bracken test prior to the study. As in the preceding dissertation, the effects of a Harcourt basal reading program were studied here. More precisely, DiChiara's (2000) goal was to determine a difference between an at-risk group of students engaged in direct phonics instruction and an at-risk group of students engaged in traditional, basal instruction. The sample of 162 fourth-grade students from Lee County, Chambers County, and Opelika City School systems of Alabama were identified as at-

Saxon Phonics versus Harcourt Trophies 13 risk by low reading achievement (DiChiara, 2000). This focus on at-risk students only was a limitation of the study because the results could not be generalized to other populations (DiChiara, 2000). On the other hand, this information added to the current body of research on the interaction effect of instructional program and ability level. The one independent variable for this study was instructional method with the following two levels as confirmed by DiChiara (2000), "explicit, rule-based, phonics-based instruction (Direct Instruction) and traditional, activity-based, whole-language-based, basal reader instruction" (p. 71). The dependent variable was reading achievement measured by the Stanford Achievement Test (SAT) (DiChiara, 2000). The hypothesis most relevant to the current study stated there would be no significant difference between the direct instruction group and the basal reader group (DiChiara, 2000). For this experimental research design, students from the Lee County School system were randomly assigned to the direct instruction treatment group, while students from Opelika City and Chambers County School systems were randomly assigned to the basal instruction control group (DiChiara, 2000). Then, the researcher ran an analysis of variance (ANOVA) using the SPSS program to analyze the SAT pretest and posttest data at the .05 significance level (DiChiara, 2000). DiChiara (2000) reported, "The analysis revealed a significant difference between groups (t = 4.730, p = .000)" (p. 96). The supporting data, F = 22.371, was gathered from an ANOVA table in Appendix D. According to DiChiara (2000), "At-risk students who were taught using the phonics and rules-based approach significantly outperformed students taught through the traditional basal in two out of three individual school comparisons" (p. 106). The purposes of the next study were to compare Saxon Phonics and third grade reading achievement and to also determine the effects of gender (Leib, 2001). Leib (2001) hypothesized

Saxon Phonics versus Harcourt Trophies 14 in the null, "There will be no relationship between the students' reading achievement scores and the implementation of the Saxon Phonics Program" (p. 18). The independent variables applicable to the current study included instructional programs, Saxon Phonics and Silver Burdett Reading, and gender (Leib, 2001). The dependent variable was Iowa Tests of Basic Skills (ITBS) reading scores (Leib, 2001). While this study did not include Harcourt as in the two previous studies, Silver Burdett Reading was also a basal program. The sample consisted of 402 third-grade students from St. Clair County located in an Illinois suburban school district (Leib, 2001). The goal of the descriptive, causal-comparative research was to determine a difference between the 201 students who received only Silver Burdett basal reading instruction and the 201 students who received the additional Saxon Phonics program (Leib, 2001). With help from the school's curriculum director, ex post facto data was taken from the 1994-1995 and 1998-1999 school years and disaggregated by gender as indicated on the ITBS report sheets (Leib, 2001). Once the data was collected and coded, an independent t-test was conducted and indicated the following, "Students in phonics had significantly higher scores in reading [t (403) = 2.907, p = .00) < .05]" (Leib, 2001, p.76). Next, an ANOVA showed females (F = 4.24, p = .04) scored significantly higher with Saxon Phonics than males (Leib, 2001). Although Leib (2001) reported this study was limited to one school in a suburban area, it did provide information to the current study on Saxon Phonics specifically. Other Instructional Programs and Reading Achievement According to Jones (2006), "The selection and adoption of an effective, research-based core reading program in the primary grades is a critical step in the development of an effective school-wide reading initiative" (p. 3). The author's purpose was to examine the effect of a

Saxon Phonics versus Harcourt Trophies 15 research-based basal reading program implemented in Tennessee schools after mandated by No Child Left Behind (NCLB) (Jones, 2006). In other words, the goal of the research was to determine a difference in reading achievement as measured by the Tennessee Comprehensive Assessment Program (TCAP), or the dependent variable, between students in 2000 before NCLB research-based instruction and students in 2004 after research-based instruction, or the key independent variable (Jones, 2006). While the researcher studied other demographic independent variables such as ethnicity, socioeconomic status, and special education status, the one pertinent to the current study was gender (Jones, 2006). The author of this research asked if reading achievement would increase due to the research-based curriculum and if gender would account for any differences (Jones, 2006). After approval from the Institutional Review Board (IRB) of Tennessee State University and after permission was granted by Fayettevile City Schools and Lincoln County Schools, the researcher collected ex post facto data for this descriptive, causal-comparative study (Jones, 2006). The sample of 1, 596, as stated by Jones (2006), "included third- and fifth-graders in both systems with a total of 845 students from the 2000 test administration and 751 students from the 2004 test administration" (p. v). To answer the research questions mentioned above, SPSS ran independent samples t-tests at the .01 significance level (Jones, 2006). The results (t = -.429, p = .668) indicated no significant difference, meaning the 2004 students did not make a significant gain as hypothesized in the null (Jones, 2006). Male students (t = -.236, p = .813) and female students (t = -.441, p = .659) also showed no significant difference between 2000 and 2004 test administrations (Jones, 2006). The limitations of not controlling for parental involvement, teacher equitability, and other supplemental programs utilized in the classrooms were considered (Jones, 2006).

Saxon Phonics versus Harcourt Trophies 16 Similar to the current study, the purpose of this dissertation was to determine if a supplemental phonics program called Orton Gillingham (OG) had an effect on reading achievement (Trepanier, 2009). Trepanier (2009) hypothesized that there would be a difference between the experimental group taught with the OG program in conjunction with the basal reading program and the control group taught with only the basal reading program. The dependent variable was the Standardized Test for Assessment of Reading (STAR) and the independent variable was the instructional programs of Orton Gillingham and Macmillan/McGraw-Hill (Trepanier, 2009). The study adhered to a quasi-experimental, nonequivalent control group design and took place in a suburban Georgia town (Trepanier, 2009). The 29 first-grade and 31 second-grade students, all conveniently sampled, were divided into an experimental group and a control group (Trepanier, 2009). The small sample sizes described here as well as a variety of teaching styles were limitations of the study (Trepanier, 2009). The researcher collected the data and assigned each subject a number to ensure confidentiality (Trepanier, 2009). Trepanier (2009) stated the data analysis procedures as follows: The SPSS program was used to analyze the pretest and posttest STAR scores data. A ttest for independent samples was conducted to determine if there was a statistical significant difference between the reading achievement scores of the classes where the Orton Gillingham program was used along with the basal program and the classes where only the basal was used. (p. 62) At the 95% confidence interval, both first-grade (t = 1.001, p = .325) and second-grade (t =.783, p = .530) data yielded results that were not significant and the hypothesis was rejected (Trepanier, 2009).

Saxon Phonics versus Harcourt Trophies 17 The purpose of the last dissertation in this section was to determine if a significant difference existed among the three instructional programs of (a) Macmillan/McGraw Hill (Control Group), (b) SRA Phonics (Experimental Group A), and (c) Word Building (Experimental Group B) and reading achievement (Sporleder, 1998). The study was large in scope and consisted of nineteen hypotheses, five independent variables, and four dependent variables (Sporleder, 1998). To align with the current study, this information was reduced to two hypotheses, two independent variables, and one dependent variable. The hypotheses most related to the current study as written by Sporleder (1998) were, "There is no significant interaction of gender and method on the adjusted means of reading achievement and there is no significant difference in the adjusted means of reading achievement among the treatment groups" (p. 82). The independent variables were the three levels of instructional method mentioned earlier and gender (Sporleder, 1998). The dependent variable was total reading achievement on the Woodcock Reading Mastery Tests-Revised (WRMT-R) (Sporleder, 1998). As in the previous study, the research design was quasi-experimental, nonequivalent control group (Sporleder, 1998). Sporleder (1998) recognized the group design as the main limitation of the study as stated: This quasi-experimental study was conducted in a nearby school district utilizing volunteer teachers with intact classrooms. There were three first-grade classrooms in each of the two experimental groups and three in the control group, making a total of nine first-grade classrooms participating in the study. Because neither the random selection of teachers nor the assignment of subjects to groups was possible, the equivalency in variables other than the treatment of the experimental and control groups cannot be assumed. (p. 76)

Saxon Phonics versus Harcourt Trophies 18 The population from southwestern Montana involved 141 first-grade students divided among the two experimental groups and the control group (Sporleder, 1998). Before the study, the researcher provided a workshop for the participating teachers and informed parents of the experiment through a personal letter (Sporleder, 1998). As reported by Sporleder (1998), "The statistical data was analyzed by the MSUSTAT Statistical Analysis Package Version 5.25 developed by Dr. Richard E. Lund at Montana State University" (p. 81). Sporleder (1998) ran a two-factor analysis of covariance (ANCOVA) at the .05 alpha and applied a Newman-Keuls to detect any differences among groups. With values of 107.7, 109.0, 108.7, 107.8, and 108.5, the group receiving word building instruction had consistently higher means, regardless of gender (Sporleder, 1998). Gender Four of the six preceding studies examined the effects of gender and various instructional programs and methods on reading achievement. The research reproduced here was significant to the current study because it took an exclusive look at the ability of gender to predict success in phonics. Fernandez (2008) stated in the introduction, "Children's ability to manipulate information, or their working memory, appears to vary according to their age and gender" (p. 3). The purpose of this correlational study was to not only establish a relationship between working memory and phonological awareness in males and females, but to also determine the degree to which these variables can predict phonological awareness (Fernandez, 2008). The purpose led the author to ask the corresponding research questions of: (a) Are working memory and phonological awareness related and (b) Are working memory and gender predictors of phonological awareness (Fernandez, 2008). Fernandez (2008) reported, "The independent variables were working memory and gender and the dependent variable was

Saxon Phonics versus Harcourt Trophies 19 phonological awareness" (p. 25). Both the independent variable of working memory and the dependent variable of phonological awareness were measured by The Woodcock-Johnson IIITests of Cognitive Abilities (WJ-III Cog) (Fernandez, 2008). The assumption that other variables such as socioeconomic status were equal among the participants was a limitation of this study (Fernandez, 2008). The sample consisted of 80 males and 78 females (N = 158) of ages eight to ten years in a Western Massachusetts school system (Fernandez, 2008). Only the students authorized by parent consent were given the WJ-III Cog subtests (Fernandez, 2008). According to Fernandez (2008), "Forms were numbered to increase confidentiality levels. A list with the students' name and number were kept separate and destroyed once forms with test protocols were matched with the numbers to ensure confidentiality" (p. 29). Testing at the .05 significance level, SPSS conducted a bivariate correlation for the first research question regarding relationship and a multiple regression for the second research question regarding prediction (Fernandez, 2008). A significant Pearson's correlation coefficient of r (156) = .524, p < .05 revealed a relationship between working memory and phonological awareness (Fernandez, 2008). This was examined further to reveal a significant correlation between working memory and phonological awareness for males (r (78) = .559, p < .05) and females (r (76) = .479, p < .05) (Fernandez, 2008). However, a linear regression analysis yielded p > .05, R = .53, R2 = .276, which indicated no significance on the extent of working memory and gender to predict phonological awareness (Fernandez, 2008). The purpose of the next study was to determine a difference in standardized reading and math scores between students in a departmentalized gender-inclusive classroom and students in a departmentalized coeducational classroom at East Marion Elementary School in Florida (Olson,

Saxon Phonics versus Harcourt Trophies 20 2010). While the departmentalized classroom setting differed from a self-contained classroom setting in the current study, it did provide insight about the effects of gender on reading achievement. In addition, this study followed the same descriptive, causal-comparative, factorial design (Olson, 2010). Of the six hypotheses tested by Olson (2010), the two most appropriate for the current study included: "(a) There is a significant difference between male and female students in their reading standardized test scores in departmentalized classrooms and (b) There is a significant interaction between gender and class type regarding reading scores" (pp. 8-9). The independent variables were gender and class type with the two categories of departmentalized genderinclusive and departmentalized coeducational (2 x 2) (Olson, 2010). The dependent variable was reading achievement measured by Florida Comprehensive Assessment Test (FCAT) (Olson, 2010). Olson (2010) collected ex post facto data from the 2008 school year, which included a sample of 202 fourth- and fifth-grade students enrolled in the departmentalized gender-inclusive and coeducational classrooms. This small sample size restricted to one school was a limitation of the study (Olson, 2010). The data collection procedures, as reported by Olson (2010), went as follows: (a) scores were coded for confidentiality and stored on a disk, (b) data was entered into an Excel spreadsheet, and (c) a two-way ANOVA was run for statistical significance. The results for both class types combined together proved to be insignificant (F [1,202] = 0.4, p = 0.5120 > 0.05), which suggested no difference in reading between male and female students (Olson, 2010). The interaction of gender and class type also had no significant effect (F [1,202] = 0.2, p = 0.6939 > 0.05) (Olson, 2010). Olson (2010) acknowledged:

Saxon Phonics versus Harcourt Trophies 21 The results suggest that gender differences in reading scores may not depend on the type of learning environments, or vise versa. In other words, based on the evidence, there is no interaction among the four subgroups (female gender-inclusive, male genderinclusive, female coeducational, and male coeducational) on the reading portion of the standardized test. (p. 92) The author of the next study quoted, "Although quality reading instruction may be offered during the regular school year, some students do not achieve at a rate that allows them to keep pace with other students their age" (Reed, 2001, p. 3). The purpose of the research was to determine a difference in reading achievement between a group of low-achieving first-grade students who participated in a summer school program and a group of low-achieving first-grade students who did not participate in the summer school program (Reed, 2001). Reed (2001) hypothesized in the null that there would be no difference between the students who did and did not participate based on gender. Like the current study, the research design was descriptive, causal-comparative and used ex post facto data from May 1999, August 1999, and February 2000 (Reed, 2001). The two independent variables in this study were participation in summer school and gender and the dependent variables were the three administrations of Terra Nova previously mentioned (Reed, 2001). As reported by Reed (2001), "The population in this study was first grade students in southwest Missouri who scored below the 50th percentile on the reading portion of CTB/McGraw-Hill Terra Nova achievement test" (p. 37). Group A included the 30 students who went to summer school, while Group B included the 44 students who did not go to summer school (Reed, 2001).

Saxon Phonics versus Harcourt Trophies 22 To test the effects of the independent variables on the dependent variables, SPSS ran a multivariate analysis of covariance (MANCOVA) (Reed, 2001). A significant interaction was found between gender and treatment (F (1, 73) = 4.20, p = .044 < .05) (Reed, 2001). Therefore, Reed (2001) applied a post-hoc and described the following, "The data presented in Table 4 indicates that the female students who enrolled in the summer reading program in June 1999 outperformed the male students" (p. 49). The possibility that students could have regressed between the conclusion of summer school and the posttest administration was a limitation taken into consideration by the researcher (Reed, 2001). Ability Level Agnew's (2005) research, comparable to the current, was descriptive, causalcomparative, and ex post facto. Additionally, all variables related to the current study. The dependent variable was reading achievement measured by the Stanford Achievement Test (SAT) (Agnew, 2005). The independent variables were instructional method and ability level (Agnew, 2005). The ability level of the students was determined by the scaled score they received on the 2002 SAT administration (Agnew, 2005). Agnew (2005) identified the levels as follows: "(a) Level 1- Below Basic, (b) Level 2- Basic, (c) Level 3- Proficient, and (d) Level 4- Advanced" (p. 58). The purpose of the study was to determine if teaching using a phonics approach or balanced approach was more effective (Agnew, 2005). Agnew (2005) said in the hypothesis that there would be no difference in the reading scaled score of the 225 students instructed with a phonics approach and the 146 students instructed with a balanced approach. One limitation of the study was that the fourth-grade students in the sample were only from Christian schools accredited by the International Christian Accrediting Association (Agnew, 2005).

Saxon Phonics versus Harcourt Trophies 23 Before achievement data was gathered, the researcher interviewed the school administrators to verify the frequency and type of instruction used in the classroom (Agnew, 2005). SPSS analyzed the data that was collected using a factorial (4 x 2) ANOVA and employing a .05 significance level (Agnew, 2005). Though there was no significant effect of instructional method, there was a main effect for level (F = 5.652, p = .001 < .05) (Agnew, 2005). Because of this, Agnew (2005) conducted the Tukey Post Hoc Multiple Comparison at the levels. Level 1 with p-values of .038, .011, and .001 showed significance. Agnew (2005) reported, "The balanced approach (Group B) of Level 1 showed the largest increase, with 83% of students moving to Level 2 on their standardized test scaled score after a year of balanced language arts instruction" (pp. 69 & 71). The purpose of the next research conducted in 1985 was to assess the effectiveness of three reading techniques on the reading achievement of low, middle, and high ability first-grade students (Sippola, 1985). The independent variables were ability level with the three levels declared in the purpose and reading technique with the three treatments of listen-alone, listenread, and read-alone (Sippola, 1985). The two dependent variables were reading vocabulary and reading comprehension measured by ITBS, the test Leib (2001) used in a previous study (Sippola, 1985). Sippola (1985) hypothesized there would be a difference in reading vocabulary and reading comprehension scores among the three treatments of listen-alone, listen-read, and readalone. Sippola (1985) stated specifically, "First grade students completing the listen-read treatment condition will score significantly higher" (p. 61). The 90 first-grade participants in the sample were from three elementary schools in the Anchorage, Alaska School District (Sippola, 1985). Similar to the current study, the researcher did not control for students who had been

Saxon Phonics versus Harcourt Trophies 24 retained (Sippola, 1985). In addition, the generalization of this research was limited to an urban setting in Alaska (Sippola, 1985). This study utilized an experimental randomized block design (Sippola, 1985). Once ability levels were determined by the Holt Basic Reading Series Pupil Placement Test instrumentation as noted by Sippola (1985), "Each student from a given ability level was randomly assigned to one of three treatment groups" (p. 40). The teachers continued with daily instructional routines and applied the treatments for 15-minute sessions three times per week (Sippola, 1985). To analyze the effects of reading technique and ability level on reading achievement, Sippola (1985) ran two 3 x 3 analyses of variance with a .01 significance level. The large scope of these analyses yielded many results (Sippola, 1985). Ability level, an independent variable of both this study and the current study, had a significant main effect in vocabulary, F (2, 81) = 377.90, p = .0001 < .01, and a significant main effect in reading comprehension, F (2, 81) = 230.81, p = .0001 < .01 (Sippola, 1985). With F = 1.87, high ability was the only group to not yield significant simple main effects for treatment in reading comprehension (Sippola, 1985). A Tukey's Honestly Significant Difference Test revealed that both low and middle ability readers scored significantly higher on reading vocabulary and reading comprehension subtests with the listen-read treatment as hypothesized (Sippola, 1985). According to the author of the next study, "Numerous research studies exist on the importance of phonemic awareness in reading acquisition, yet a deficit exists in the literature base concerning daily writing and the phonemic awareness development of kindergarten students representing different literacy levels" (Snell, 2007, p. 4). Therefore, the purpose of Snell's (2007) study was to determine if daily writing makes a difference in the phonemic awareness of students with varying abilities or not.

Saxon Phonics versus Harcourt Trophies 25 The following research questions were asked: (a) Does daily writing have an increasing effect on phonemic awareness and (b) Does daily writing have an increasing effect on phonemic awareness of different ability levels (Snell, 2007). The independent variables, all significant to the current study, were instructional method (writing program or no writing program), ability level (at-risk, some-risk, or low-risk), and gender (Snell, 2007). The dependent variable was phonemic awareness evaluated by the Phonological Awareness Test (PAT) (Snell, 2007). The research used a quasi-experimental pretest-posttest design (Snell, 2007). The sample consisted of 77 kindergarten students in northeastern Pennsylvania (Snell, 2007). As identified by the Initial Sound Fluency (ISF) portion of Dynamic Indicators of Basic Early Literacy Skills (DIBELS), 53 were at low-risk, 17 were at some-risk, and seven were atrisk (Snell, 2007). Because of the low number of students in both the some-risk and at-risk ability groups, these two were combined for statistical analysis (Snell, 2007). Other than this sample not being representative of other kindergarten students, a weakness of the study was the researcher's inability to control phonemic awareness development from other instructional programs (Snell, 2007). The procedures were executed as follows: (a) Forty students in the experimental group wrote daily, while 37 students in the control group did not; (b) pretest and posttest data was collected during 2006-2007; and (c) SPSS 13.0, based on a .05 level, ran a two-sample t-test (Snell, 2007). The results of the two-sample t-test concluded that the experimental group (M = 42.6) scored significantly higher than the control group (M = 33.5), t = -3.06 (75), p = 0.003 (Snell, 2007). The low-risk (on grade level) experimental group (M = 40.6) scored significantly higher than the low-risk control group (M = 29.6), t = -3.03 (51), p = 0.004 (Snell, 2007). The some/at-

Saxon Phonics versus Harcourt Trophies 26 risk experimental group (M = 48.7), on the other hand, did not score significantly higher than the some-risk/at-risk control group (M = 39.9), t = -1.94 (22), p = 0.066 (Snell, 2007). Snell (2007) observed: These results suggest that the exposure to daily writing did not have as strong an influence on the phonemic awareness development of the some/at-risk students in the experimental group as it had on the low-risk students in the same group. (p. 75) Oral Reading Fluency From the words of Walker (2008), "Reading fluency is critical to proficiency in reading" (p. iv). The purpose of this study was to determine the effects of read-aloud, choral reading, and repeated reading strategies on first grade oral reading achievement (Walker, 2008). The goal, and hypothesis, of the research was to find a significant difference in first grade oral reading rates between students who did and did not receive instruction with fluency-building strategies (Walker, 2008). The independent variable was instructional method, either additional fluency-building strategies or traditional reading instruction (Walker, 2008). The dependent variable, as in the current study, was oral reading fluency measured by DIBELS (Walker, 2008). The research employed a quasi-experimental pretest-posttest control group design (Walker, 2008). The sample for the study consisted of 56 first-grade students in a suburban-rural Pennsylvania school district (Walker, 2008). Walker (2008) reported that 26 students were in the experimental group who received explicit fluency-building strategies, while the remaining 30 students were in the control group who received traditional reading instruction from Houghton-Mifflin basal program.

Saxon Phonics versus Harcourt Trophies 27 Prior to the study, the participating teachers in the experimental classrooms attended a one-day conference titled "Creating Fluent Readers from Phonics to Fluency: Strategies for Achieving Reading Proficiency" (Walker, 2008). The study was conducted from January to May with treatments being administered at least two hours per week (Walker, 2008). Walker (2008) gathered the DIBELS pretest and posttest data and ran multiple t-tests at a significance level of .05 for statistical analysis. The one-tailed t-test yielded p = 0.905, indicating that the fluency-building strategies did not have a significant effect on the oral reading fluency of the experimental group as hypothesized by the researcher (Walker, 2008). Experimental mortality or loss of subjects due to relocation was a limitation realized by the researcher of this study (Walker, 2008). The purpose of the subsequent study, like the previous, was to identify the effects of fluency-building strategies on oral reading fluency (Czepull, 2007). The goal of this research, however, was to determine a difference in reading scores between students taught with the fluency-building strategies and students taught with a phonics program, Sonday (Czepull, 2007). Of the six research questions asked by Czepull (2007), the following were most relevant to the current study: 1. What improvements did students make in words per minute scores with Sonday, a method of phonics instruction, and methods of fluency instruction including repeated reading and modeling reading? 2. What differences exist in the improvement of words per minute scores between students taught with Sonday, a method of phonics instruction, and methods of fluency instruction including repeated reading and modeling fluency? (p. 4)

Saxon Phonics versus Harcourt Trophies 28 The dependent variable was the DIBELS measure of oral reading fluency, while the independent variable was the instructional method of either Sonday or fluency activities (Czepull, 2007). Because the researcher used intact groups as opposed to randomization to assign groups, the design was quasi-experimental pretest-posttest nonequivalent control group (Czepull, 2007). The sample (N = 29), from a low-middle class school in South Dakota, included 14 third-grade students who received Sonday instruction (control group) and 15 third-grade students who received explicit fluency instruction (experimental group) (Czepull, 2007). A delimitation of the study, as reported by Czepull (2007), was the purposeful sampling of only 29 students limited to one elementary school. After receiving permission from the superintendent, Czepull (2007) collected pretest and posttest DIBELS scores. SPSS ran paired t-tests to analyze research question one and an ANCOVA for research question two, both of which were conducted at the .05 significance level (Czepull, 2007). As inquired in question one, both the control group (t (13) = -13.353, p = .000) and experimental group (t (14) = -6.081), p = .000) made significant improvement in oral reading fluency (Czepull, 2007). However, in comparison to one another, ANCOVA results of F =3.733, p = .064 showed no significant difference between the group who received Sonday phonics instruction and the group who received fluency-building strategies (Czepull, 2007). The author of the last dissertation suggested, "To ensure that the diagnostic testing process is going to positively impact student reading achievement, educators need to administer only the assessments that will provide the most accurate early identification of students with reading difficulties" (Hollinger, 2009, p. 2). The purposes of this descriptive, correlational study were to establish a relationship between diagnostic assessment performance and reading

Saxon Phonics versus Harcourt Trophies 29 achievement and to identify the assessment that best predicted reading achievement (Hollinger, 2009). For this study, the researcher asked: (a) What is the relationship between diagnostic assessment performance and reading achievement and (b) Which assessments best predict reading achievement for male and female students (Hollinger, 2009). The following six diagnostic assessments, or independent variables were: (a) Kindergarten Readiness AssessmentLiteracy (KRA-L), (b) first grade Ohio Diagnostic Reading Test, (c) first grade DIBELS Oral Reading Fluency (ORF), (d) second grade Diagnostic Reading Test, (e) second grade DIBELS ORF, and (f) third grade DIBELS ORF (Hollinger, 2009). Another independent variable was gender, while the dependent variable was the third grade Ohio Reading Achievement Test (Hollinger, 2009). The sample, as reported by Hollinger (2009), was "Forty-nine third-grade students who had completed the six different diagnostic reading assessments" (p. 41). These participants, like the participants in the current study, were from a small, rural community (Hollinger, 2009). Some limitations of the study included small sample size, no analysis of the Ohio Reading Achievement Test subtests, and errors made when scores were entered into the database (Hollinger, 2009). The procedure of Hollinger's research went as follows: The relevant assessment data was housed in a database maintained by the participating school district. The student scores were collected and submitted by the classroom teachers beginning in kindergarten and continuing to the release of the third grade Ohio Reading Achievement Test scores in July, which were mailed to the school by the Ohio Department of Education. A copy of the database was given to the researcher with

Saxon Phonics versus Harcourt Trophies 30 student names removed. The student scores were entered into a Microsoft Excel spreadsheet and were statistically analyzed by SPSS. (pp. 52-53) The results of a Pearson Correlation at the .01 significance level, for question one, revealed that the strongest relationship (r = .709) existed between the second grade Ohio Reading Diagnostic Test and the Ohio Reading Achievement Test (Hollinger, 2009). First grade DIBELS ORF (r = .169) and third grade DIBELS ORF (r = .114) correlation coefficients were not significant (Hollinger, 2009). For male participants, a logistic regression of (-2 Log Likelihood= 20.77, 2 (1) = 16.60, p<.0001) indicated that second grade DIBELS ORF was the best predictor (Hollinger, 2009). A logistic regression of (-2 Log Likelihood= 12.89, 2 (1) = 7.12, p<.0001) yielded the second grade Ohio Reading Diagnostic Test as the best predictor for female participants in reading achievement (Hollinger, 2009). Summary From Wicker's (2007) study of 43 students in a rural setting to Jones' (2006) study of 1,596 students in an urban setting, the goal of all research reviewed was to determine the methods that resulted in reading achievement. Aside from the studies of correlation conducted by Fernandez (2008) and Hollinger (2009), the authors of the literature sought to find differences among groups in regards to instructional program, gender, and ability level. Although no causal relationships were inferred due to post hoc fallacy, some related significant findings occurred. Both DiChiara (2000) and Leib (2001) learned students instructed with a phonics program scored significantly higher (p = .00) than students instructed with only a basal reading program. Leib (2001), in addition, discovered females with F = 4.24, p = .04 values did significantly better than males with the Saxon Phonics program. Similarly, Reed's (2001) results (F (1, 73) = 4.20, p = .044 < .05) yielded female students outperformed male students in a

Saxon Phonics versus Harcourt Trophies 31 developmental summer reading program. On the other hand, Olson (2010) determined from F (1,202) = 0.4, p = 0.5120 that there was no significant difference on reading scores between male and female students. Fernandez (2008) uncovered gender to be a weak predictor of phonological awareness through a linear regression. For ability level, Snell (2007) established students on grade level had significantly higher reading scores when engaged in a daily writing program from t = -3.03 (51), p = 0.004. Sippola (1985) verified a listen-read technique to be most beneficial for readers of low ability in the 1985 research. Finally, Hollinger (2009), through a logistic regression (-2 Log Likelihood= 20.77, 2 (1) = 16.60, p<.0001), found DIBELS ORF, a diagnostic assessment similar to AIMSweb presented in the current study, to be the strongest predictor of reading achievement in male participants. In should be concluded, however, that additional research must be conducted, especially in the areas of Harcourt Trophies and Saxon Phonics.

Saxon Phonics versus Harcourt Trophies 32 Chapter 3: Methodology The purpose of this quantitative study was to examine the effects of a supplemental phonics program, Saxon Phonics, on the oral reading fluency of first-grade students at Jasper Elementary School located in Marion County, Tennessee. Specifically, this study determined if a difference existed between the 124 students instructed with only Harcourt Trophies basal reading program and the 172 students instructed with supplemental Saxon Phonics. The research design was descriptive, causal-comparative, ex post facto. Upon approval from the Institutional Review Board in July 2010, AIMSweb data was collected and entered into SPSS in October 2010. An analysis of covariance was run for statistical analysis. This chapter outlines the methods and procedures involved in the study. Sample The participants in this study included first-grade students who completed the AIMSweb assessment at Jasper Elementary School during the 2008-2009 and 2009-2010 school years. All students who possessed both nonsense word fluency and oral reading fluency scores were chosen for inclusion in the study. The 296 approximate participants were grouped by ability as determined by fluency scores. Of the 296 participants, 124 were from the 2008-2009 school year and received only Harcourt Trophies instruction, while 172 were from the 2009-2010 school year and received supplemental Saxon Phonics instruction. The participants were representative of Jasper Elementary School located in Marion County, Tennessee. Jasper Elementary School is a pre-kindergarten through fourth grade school with an enrollment of 674 students as of 2009 (TDOE, 2010). The 674 students encompassed the following ethnicities: 94.8 % White, 2.5 % African American, 2.1 % Hispanic, and .6 % Asian/Pacific Islander (TDOE, 2010). Seventy-six percent of the school population was

Saxon Phonics versus Harcourt Trophies 33 considered economically disadvantaged (TDOE, 2010). Male students (N = 373) comprised 52.3 % of the population, while females (N = 340) comprised 47.7 % (TDOE, 2010). Design The research design was categorized as descriptive, causal-comparative and employed ex post facto procedures. Because existing data from the 2008-2009 and 2009-2010 school years was used, the independent variables were not manipulated. Measurement and Instrumentation AIMSweb is a standardized, diagnostic assessment based on frequent monitoring (AIMSweb, 2008). The early literacy measures for first grade included letter naming fluency, letter sound fluency, phoneme segmentation fluency, nonsense word fluency, and oral reading fluency (AIMSweb, 2008). Oral reading fluency, the dependent variable, consisted of students reading a passage aloud for one minute, while the administrator calculated the number of words read correctly minus errors (AIMSweb, 2008). The website stated the following concerning reliability and validity: The Assessment and Information group of Pearson announced the National Center on Response to Intervention (NCRTI) has given its AIMSweb Reading Curriculum-Based Measurement the highest possible rating for validity and reliability. NCRTI reviewed AIMSweb, Pearson's leading web-based progress monitoring and response to intervention (RTI) system for grades 1-8, as a part of the organization's first annual evaluation of Reading Screening Tools. (AIMSweb, 2008, para.1) The main reliability established with the oral reading fluency measure was alternate-form reliability (Shinn, 2002). After the passages are written, practiced with Fry readability formula, and edited by a group of teachers and paraprofessionals, the reliability correlations are calculated

Saxon Phonics versus Harcourt Trophies 34 for each passage (Shinn, 2002). The alternate-form reliability correlation coefficient for grade one passages was r = 0.89 (Shinn, 2002). The validity of the curriculum-based oral reading fluency measurement, according to Shinn (2002), had a correlation coefficient ranging from .60 to .80. Delimitations This study did not include demographic variables such as ethnicity and socioeconomic status. In addition, this study did not control for students who were retained in first grade or students who received additional reading intervention. Though not addressed, these variables could have had an impact on first-grade reading fluency scores. Materials No materials were needed to perform this study. Data Collection Permission to access student data was granted by Timothy P. Bible, principal of Jasper Elementary School. The data was collected in July 2010 upon approval from Tennessee Technological University's Institutional Review Board. The data included first grade AIMSweb nonsense word fluency and oral reading fluency scores, gender, and ability level from the 20082009 and 2009-2010 school years. With assistance from the literacy specialist, data was obtained from the Jasper Elementary School AIMSweb web-based file. Once the data was collected, it was organized into a Microsoft Excel spreadsheet. Student identification numbers comprised column one of the spreadsheet. Student names were replaced with a student identification code of 1-296 for the approximate number of participants. Column two contained the dependent variable, AIMSweb oral reading fluency score. Column three, the covariate, was the students' AIMSweb nonsense word fluency score. Column four,

Saxon Phonics versus Harcourt Trophies 35 independent variable one, held instructional program indicated by number one (1) for Harcourt Trophies or number two (2) for Saxon Phonics. Gender, independent variable two, was housed in column five and was coded with number one (1) for male or number (2) for female. Column six, the last independent variable of ability level, was coded with number one (1) for low ability, number two (2) for middle ability, or number (3) for high ability. The names of the column labels were simplified for SPSS recognition. Limitations The descriptive, causal-comparative, ex post facto design of this research lend itself to the following three limitations identified by Best and Kahn (2006): · · · "The independent variables cannot be manipulated" (p. 145). "Subjects cannot be randomly, or otherwise, assigned to treatment groups" (p. 145). "Causes are often multiple and complex rather than single and simple" (p. 145).

The limitations unique to this particular study included: · Maturation of the students may have occurred over the course of the school years between the measures of nonsense word fluency in the fall to oral reading fluency in the spring. · · The research was performed in a short, condensed time frame. The research was limited to one elementary school in Marion County, Tennessee and, therefore, could not be generalized to other populations. · The sample size was relatively small and those students who did not have both nonsense word fluency and oral fluency scores were eliminated. · AIMSweb is a formative, diagnostic assessment rather than an achievement test such as the Tennessee Comprehensive Assessment Program (TCAP). However, because first-

Saxon Phonics versus Harcourt Trophies 36 grade students at Jasper Elementary School were not administered an achievement test, AIMSweb scores were the only available data. · Classroom teachers at Jasper Elementary School had not been trained in administering AIMSweb measures. Therefore, the lack of stability over testers, from reading interventionists to reading assistants, may have resulted in decreased reliability of the assessment. Data Analysis/Statistical Procedures Data was entered into SPSS and screened for missing cases, normality, and possible outliers. A statistical adjustment of initial ability level was made based on the covariate of nonsense word fluency scores. A factorial analysis of covariance (ANCOVA) was conducted to determine significant main effects of the independent variables on the dependent variable. Significance was set at the .05 level.

Saxon Phonics versus Harcourt Trophies 37 Chapter 4: Data Presentation and Analysis The purpose of this quantitative study was to determine if a difference in reading fluency existed among the 124 first-grade students instructed with only Harcourt Trophies basal reading program and the 172 first-grade students instructed with supplemental Saxon Phonics at Jasper Elementary School located in Marion County, Tennessee. Upon Institutional Review Board approval in July 2010, data was collected via the AIMSweb web-based file with the assistance of the school's literacy specialist. The data, which included students' oral reading fluency score, nonsense word fluency score, and gender, was coded and organized into a Microsoft Excel spreadsheet. SPSS conducted a factorial analysis of covariance (ANCOVA) set at the .05 significance level. This chapter presents the results of the data analysis and includes tables/figures to support these findings. Presentation of Data The collected data of 296 first-grade students organized in a Microsoft Excel spreadsheet was imported into SPSS for data analysis. Prior to running an ANCOVA, SPSS screened the data for missing cases, outliers, normality of the variables, and homogeneity of variance. The first step in this cleaning process included eliminating the 34 missing cases as shown in Table 4.1. Table 4.1 Summary of Missing Cases Case Processing Summary Cases Valid Missing N Percent N Percent 262 88.5% 34 11.5% 262 88.5% 34 11.5%

DV-ORF CovariateNWF

Total N Percent 296 100.0% 296 100.0%

Saxon Phonics versus Harcourt Trophies 38 Of the 262 remaining participants, five outliers with an oral reading fluency score greater than 145 were removed. Figure 4.1 illustrates the extreme scores taken out of the study to prevent skewed data. Figure 4.1 Box-and-Whisker Plot of Extreme Cases

Further screening revealed one additional extreme case with an oral reading fluency score greater than 141. This outlier is shown in Figure 4.2.

Saxon Phonics versus Harcourt Trophies 39 Figure 4.2 Box-and-Whisker Plot of Extreme Case

The screening and removal of 34 missing cases and six total outliers resulted in 256 normally distributed oral reading fluency scores represented in Figure 4.3, Table 4.2, and Figure 4.4.

Saxon Phonics versus Harcourt Trophies 40 Figure 4.3 Histogram of Normally Distributed ORF

Table 4.2 Stem-and-Leaf Plot of Normally Distributed ORF DV-ORF Stem-and-Leaf Plot Frequency Stem & Leaf 2.00 0 . 99 12.00 1 . 334455667799 24.00 2 . 000023344555566778899999 38.00 3 . 00001111111222233344447788888889999999 35.00 4 . 00111223345556666666778888888899999 31.00 5 . 0012222222222233556666678889999 26.00 6 . 00011222222345556668899999 28.00 7 . 0000111112333444455556677779 14.00 8 . 01111122333556 18.00 9 . 122222223444445579 11.00 10 . 00003455789 9.00 11 . 024556889 6.00 12 . 018999 2.00 13 . 00 Stem width: 10 Each leaf: 1 case(s)

Saxon Phonics versus Harcourt Trophies 41 Figure 4.4 Box-and-Whisker Plot of Normally Distributed ORF

As mentioned, 256 participants with normally distributed oral reading fluency scores remained after the screening process was completed. The mean score was 59.27 and the standard deviation was 28.708. The descriptive statistics for the participants by variables of instructional program, gender, and ability level are listed in Table 4.3. Table 4.3 Descriptive Statistics of Variables Between-Subjects Factors Instructional Program Gender Ability 1 2 1 2 1 2 3 Value Label Harcourt Saxon male female low middle high N 107 149 128 128 87 88 81

Saxon Phonics versus Harcourt Trophies 42 Data Analysis After data was collected, it was entered into a Microsoft Excel spreadsheet. The first column contained student identification, which replaced student names with numbers 1-296. The dependent variable, AIMSweb oral reading fluency scores, was in column two. Column three housed the covariate, or the students' nonsense word fluency scores. Columns four through six were the independent variables of instructional program, gender, and ability level. Instructional programs were coded with one (1) for Harcourt Trophies and two (2) for Saxon Phonics. Gender was coded with one (1) for male and two (2) for female. Ability level was coded with one (1) for low ability, two (2) for middle ability, and three (3) for high ability. The data was then imported into SPSS and screened for missing cases, outliers, normality of the variables, and homogeneity of variance. Once the data was clean and normally distributed, SPSS ran a factorial analysis of covariance (ANCOVA). The significance level was set at .05 to determine if any significant differences existed as asked in the following research questions. Research Question 1 Will there be a difference in AIMSweb reading fluency scores between first- grade students who received supplemental Saxon Phonics instruction and students who received only Harcourt Trophies basal reading instruction, after adjusting for ability level? The results of the ANCOVA revealed a significant main effect of instructional program [F (1, 243) = 8.882, p = .003]. The students instructed with Saxon Phonics (M = 60.40) scored significantly higher than those students instructed with only Harcourt Trophies (M = 57.69) as shown in Figure 4.5.

Saxon Phonics versus Harcourt Trophies 43 Figure 4.5 Estimated Marginal Means of DV-ORF

Research Question 2 Will this difference, if any, vary depending on gender and ability levels of low, middle, and high? The results of the ANCOVA showed no significant interaction effects between instructional program and gender [F (1, 243) = .036, p = .850] or between instructional program and ability level [F (2, 243) = .182, p = .833]. These findings can be viewed in Table 4.4.

Saxon Phonics versus Harcourt Trophies 44 Table 4.4 ANCOVA Table of Between-Subject Effects Tests of Between-Subjects Effects Dependent Variable:DV-ORF Source Type III Sum of Squares df Corrected Model 1.833E5 12 Intercept 82071.798 1 CovariateNWF 8933.319 1 IVProgram 980.888 1 IVGender 118.685 1 IVAbility 69575.378 2 IVProgram * IVGender 3.973 1 IVProgram * IVAbility 40.262 2 IVGender * IVAbility 107.753 2 IVProgram * IVGender * 158.685 2 IVAbility Error 26836.113 243 Total 1.109E6 256 Corrected Total 210158.402 255 a. R Squared = .872 (Adjusted R Squared = .866)

Mean Square 15276.857 82071.798 8933.319 980.888 118.685 34787.689 3.973 20.131 53.877 79.343 110.437

F 138.331 743.157 80.891 8.882 1.075 315.001 .036 .182 .488 .718

Sig. .000 .000 .000 .003 .301 .000 .850 .833 .615 .489

Saxon Phonics versus Harcourt Trophies 45 Chapter 5: Conclusions, Discussion, and Recommendations This quantitative study examined the differences in oral reading fluency between firstgrade students instructed with supplemental Saxon Phonics and first-grade students instructed with only Harcourt Trophies basal reading program. The sample consisted of 296 first-grade students at Jasper Elementary, a school located in Marion County, Tennessee. The dependent variable was oral reading fluency as measured by AIMSweb, while the covariate was nonsense word fluency. The independent variables were Harcourt Trophies and Saxon Phonics instructional programs, gender, and ability levels of low, middle, and high. A factorial analysis of covariance was conducted to determine if significant differences existed among the variables. This chapter looks closely at the results of the study and makes recommendations for future research. Summary of Results The first research question asked if there would be a difference in AIMSweb reading fluency scores between first-grade students who received supplemental Saxon Phonics instruction and students who received only Harcourt Trophies basal reading instruction, after adjusting for ability level. The results of the ANCOVA revealed a significant main effect of instructional program [F (1, 243) = 8.882, p = .003], with those students instructed with Saxon Phonics (M = 60.40) scoring significantly higher than those students instructed with only Harcourt Trophies (M = 57.69). The second research question asked if this difference varied depending on gender and ability levels of low, middle, and high. The results of the ANCOVA showed no interaction effects between instructional program and gender [F (1, 243) = .036, p = .850] or between instructional program and ability level [F (2, 243) = .182, p = .833].

Saxon Phonics versus Harcourt Trophies 46

Conclusions The purpose of this study was to examine the effects of supplemental Saxon Phonics on the reading fluency of first-grade students. More specifically, the study looked at differences between students who received Saxon Phonics instruction and students who only received Harcourt Trophies basal reading instruction. A factorial analysis of covariance revealed those students instructed with Saxon Phonics scored significantly higher than those students only instructed with Harcourt Trophies basal reading program. Therefore, it can be concluded from these results that explicit, systematic phonics instruction used in conjunction with the basal reading program yields higher reading fluency among first-grade students. In addition, this study analyzed whether gender and ability levels of low, middle, and high had an interaction effect with the instructional programs. Neither gender nor ability level had a significant effect. The conclusion that all students, regardless of gender or ability level, benefit from explicit, systematic phonics instruction is drawn from these results. Discussion of Results The ability of Saxon Phonics to yield significantly higher reading fluency scores among first-grade students was a major finding of this study. DiChiara (2000) similarly found that students who received phonics instruction, though not Saxon Phonics, outperformed students who only received a traditional approach to reading. Although third-grade students were studied, Leib (2001) also discovered from his results that students in Saxon Phonics had higher scores in reading than those instructed with the Silver Burdett basal reading program. On the other hand, the results of Wicker's (2007) study showed that students in another reading program had significantly higher gains than those in Saxon Phonics/Harcourt.

Saxon Phonics versus Harcourt Trophies 47 The second major finding of this study was that all students, regardless of gender or ability level, had higher oral reading fluency scores when instructed with Saxon Phonics. Likewise, in a correlation study, Fernandez (2008) did not find gender to be a predictor of phonological awareness. Contrastingly, both Reed (2001) and Leib (2001) conducted studies that showed females had an advantage over male students. As for the effects of ability level, Agnew (2001) dissimilarly determined that students of the lowest ability level made the largest increase with a balanced approach. Recommendations · Administrators should adopt an explicit, systematic phonics program in the primary grade levels. · Administrators should continue to review research on instructional programs, as these studies yield unpredictable results when different variables are considered. · Administrators should provide teachers with professional development on current phonics practices. · Teachers should devote time and planning to incorporating phonics into reading instruction. · · Students should learn phonics patterns in addition to sight word reading. Future research should study the correlation of instructional method and amount of time invested in phonics instruction on reading achievement.

Saxon Phonics versus Harcourt Trophies 48

References Agnew, W. T. (2005). The effects of instructional method on fourth-grade reading achievement: phonics versus a balanced approach in language arts instruction. (Doctoral dissertation). Retrieved from Dissertations & Theses: Full Text database. (Publication No. AAT 3220441). AIMSweb. (2008). Assessment and data management for RTI. Retrieved from http://www.aimsweb.com Best, J. W., & Kahn, J. V. (2006). Research in education (10th ed.). Boston: Pearson. Czepull, T. K. (2007). Analysis of a fluency method and a phonics method of reading instruction in third-grade students. (Doctoral dissertation). Retrieved from Dissertations & Theses: Full Text database. (Publication No. AAT 3272948). DiChiara, L. E. (2000). The effectiveness of direct instruction versus traditional basal reading instruction as it pertains to at-risk youth. (Doctoral dissertation). Retrieved from Dissertations & Theses: Full Text database. (Publication No. AAT 9978316). Fernandez, D. L. (2008). Gender differences in working memory and phonological awareness. (Doctoral dissertation). Retrieved from Dissertations & Theses: Full Text database. (Publication No. AAT 3310710). Florida Center for Reading Research. (2003). What is Saxon Phonics and spelling? Retrieved from http://www.fcrr.org/FCRRReports/PDF/SaxonPhonicsSpelling.pdf Hollinger, J. L. (2009). The relationship between students' reading performance on diagnostic assessments and the third grade reading achievement test in Ohio. (Doctoral dissertation). Retrieved from Dissertations & Theses: Full Text database. (Publication No. AAT 3393075).

Saxon Phonics versus Harcourt Trophies 49 Jones, B. C. (2006). The effects of a basal reading program on reading achievement in selected Tennessee schools. (Doctoral dissertation). Retrieved from Dissertations & Theses: Full Text database. (Publication No. AAT 3222584). Leib, T. (2001). The relationship between Saxon phonics, kindergarten through second grade, and student achievement scores in reading. (Doctoral dissertation). Retrieved from Dissertations & Theses: Full Text database. (Publication No. AAT 3014216). National Center for Education Statistics (NCES). (2008). Mapping Tennessee's educational progress 2008. Retrieved from http://www2.ed.gov/nclb/accountability/results/progress/tennessee.pdf Olson, C. A. (2010). The gender divide: the effectiveness of departmentalized gender-inclusive classrooms. (Doctoral dissertation). Retrieved from Dissertations & Theses: Full Text database. (Publication No. AAT 3402233). Reed, G. W. (2001). The relationship between participation in a developmental reading summer school program and reading achievement among low-achieving first grade students. (Doctoral dissertation). Retrieved from Dissertations & Theses: Full Text database. (Publication No. AAT 3014295). Saxon Publishers. (2010). Saxon Phonics and Spelling K-3. Retrieved from http://saxonpublishers.hmhco.com/en/saxonphonics.htm Shinn, M. M. (2002). Standard reading assessment passages for use in general outcome measurement. Retrieved from http://www.aimsweb.com/uploads/pdfs/passagestechnicalmanual.pdf

Saxon Phonics versus Harcourt Trophies 50 Sippola, A. E. (1985). The effects of three reading instruction techniques on the comprehension and vocabulary of first graders grouped by ability. (Doctoral dissertation). Retrieved from Dissertations & Theses: Full Text database. (Publication No. AAT 8521660). Snell, C. A. (2007). The impact of daily writing on kindergarten students' phonemic awareness. (Doctoral dissertation). Retrieved from Dissertations & Theses: Full Text database. (Publication No. AAT 3261224). Sporleder, R. L. (1998). A comparison of three approaches to literacy acquisition: traditional phonics, whole language, and spelling before reading. (Doctoral dissertation). Retrieved from Dissertations & Theses: Full Text database. (Publication No. AAT 9830111). Tennessee Department of Education. (2010). Tennessee Department of Education report card. Retrieved from http://www.tennessee.gov/education/reportcard Trepanier, K. (2009). The effectiveness of the Orton Gillingham instructional program when used in conjunction with a basal reading program. (Doctoral dissertation). Retrieved from Dissertations & Theses: Full Text database. (Publication No. AAT 3355062). TSIPP. (2009). Tennessee school improvement planning process. Jasper. U.S. Department of Health and Human Services, National Institute of Child Health and Human Development. (2000). Report of the National Reading Panel: Teaching children to read (NIH Publication No. 00-4754). Walker, H. E. (2008). The effects of fluency-building strategies on the oral reading rates of firstgrade students. (Doctoral dissertation). Retrieved from Dissertations & Theses: Full Text database. (Publication No. AAT 3338623).

Saxon Phonics versus Harcourt Trophies 51 Wicker, K. (2007). The effect of two reading programs on kindergarten students' reading readiness. (Doctoral dissertation). Retrieved from Dissertations & Theses: Full Text database. (Publication No. AAT 3274995).

Information

Microsoft Word - Frederick, Jessica Research Proposal Paper.doc

51 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

43198


You might also be interested in

BETA
HarcourtTrophies_Mar08
Microsoft Word - Frederick, Jessica Research Proposal Paper.doc