Abstracts Mathematics

DISTRICT IMPLEMENTATION OF INSTRUCTIONAL COACHING AND ITS IMPACT ON ACADEMIC ACHIEVEMENT OF STUDENTS WITH DISABILITIES IN READING AND MATHEMATICS

by Julia Renberg

The purpose of this retrospective comparative study was to evaluate the differences in the academic achievement of students with disabilities (SWD) on Virginia standardized assessments in mathematics (N = 96) and reading (N = 69) across three years of tests (2016–18), when the children were in Grades 3, 4, and 5. Two cohorts were drawn from 14 low-performing Title I schools from a large public school division; the treatment cohort attended seven schools where instructional coaching had been implemented by the district, while the control sample was drawn ex post facto from seven schools without the intervention. The data from the 2016 (Grade 3) assessment were used as a pretest, with postintervention data gathered from the following two years. Results were interpreted through effect sizes and confidence intervals. In reading, the treatment group had lower baseline scores (M = 300) and pass rates (M = 7%) than the control group (M = 364 and M = 26% respectively), but saw larger gains over time, closing the achievement gap with respect to pass rates (M = 41%, Cohen’s h = 0.78); however, the 95% CI strongly overlapped, indicating that there were no differences at the .05 level of statistical significance. The treatment group also had larger improvements in student mean scores than the control group did (41 vs. 27 points), but the between-subjects effects measured with repeated-measures ANOVA indicated this difference in scores was not statistically significant (p > .05); the wide range in values of the effect size (Cohen's drm = 0.22) with negative and positive 95% confidence intervals capturing zero for reading [-.27, .51] reflected both the positive and negative effects of instructional coaching. No significant differences and negligible effect sizes were observed in mathematics. The internal and external validity of the results were affected by selection and sampling bias, sample size, regression toward the mean, and the ecological fallacy. More research is recommended to explain why the scores of some students in both the treatment and control groups increased between Grades 3 and 5, while other students’ scores declined or remained constant.