Examination of achievement gaps among fourth grade students in a selected Louisiana school district.

摘要

A major concern for educators is that student achievement is disproportionate and not increasing for all groups at the same rate. This applied dissertation addressed the problem of unequal achievement for African American, economically disadvantaged, and academically at-risk students. This quantitative study utilized a correlational design to examine achievement relative to the percentage of students in classrooms who were (a) African American, (b) free or reduced priced meal (FARM) recipients, and (c) academically at-risk (AAR). The English Language Arts (ELA) and mathematics (math) test scores of 305 fourth-grade students from one Louisiana public school district were analyzed. ELA scores were negatively correlated with the percentage AAR-ELA (r = - 22, p .001), percentage AAR-math (r = -.21, p .001), and percentage FARM (r = - .23, p = .009). Math scores were negatively correlated with the percentage FARM (r = - .25, p .001), but not with any other independent variables. The regression model for African American students was not significant for ELA, F(4, 85) = 1.65, p = .169, or math, F(4, 85) = 1.47, p = .218. The regression model for Caucasian students was significant for ELA, F(4, 210) = 4.75, p = .002, and math, F(4, 210) = 5.11, p .001. None of the independent variables was predictive of African American students' ELA scores. Percentage FARM was predictive of African American students' math scores (beta = -.34, p = .047). Percentage African American (beta = .18, p = .035) and percentage FARM (beta = -.23, p = .015) were predictive of Caucasian students' ELA scores. Percentage African American (beta = .24, p = .007) and percentage FARM (beta = -.37, p .001) were predictive of Caucasian students' math scores. These findings indicate that grouping may affect achievement. To address achievement gaps, educational administrators must consider classroom composition. Further study should include (a) multi-level modeling to measure individual, class, and school level effects, (b) greater ethnic representation and various school types (combined or single level only) and settings (rural, suburban, urban), and (c) additional grade levels, and utilize qualitative analyses to evaluate peer interaction effects.