| The problem investigated in this research study relates to the referral rate for special education evaluations between students of teachers who were exposed to the professional development and resources associated with the federally funded Reading First program and students of teachers who were not exposed to the program. The purpose of this study was to investigate whether a statistically significant difference in student referral rates for special education evaluations existed at the selected schools. The sample selected for the study was drawn from two elementary schools located in a single Massachusetts school district. Participating students will have completed Grade 1 through Grade 3 at one of the two school sites by June 2008. Students from one school comprised the treatment group and students in the other school comprised the comparison group. Although the comparison school did not have access to the Reading First resources, both schools used common reading assessments instruments. The research design selected for this study was an ex post facto causal--comparative, retrospective design analyzing longitudinal data for the 2006-08 school years. The data were processed using the analysis of variance (ANOVA) statistical technique to test for differences in single-dependent variables across the two student groups and a multivariate analysis of variance (MANOVA) to simultaneously test for differences in multiple dependent variables across the two student groups. To assist with addressing the nonrandom selection of the two selected populations, an analysis of student Massachusetts Comprehensive Assessment Scores (MCAS), gender, and free and reduced-price lunch were analyzed using a chi-square test for significance. Several factors were identified highlighting potential causes for the difference between the groups. The factors that were identified were gender, ethnicity, free or reduced-price lunch program, and the Reading First training program.;Combinations of these factors were also explored to be a possible factor for such difference. From the statistical results, with the exception of gender, it can be concluded that all factors, including their combinations, are significant factors that can affect the assessment scores of student from different grade levels. |
