Patterns in Illinois educational school data

Abstract

We examine Illinois educational data from standardized exams and analyze primary factors affecting the achievement of public school students. We focus on the simplest possible models: representation of data through visualizations and regressions on single variables. Exam scores are shown to depend on school type, location, and poverty concentration. For most schools in Illinois, student test scores decline linearly with poverty concentration. However, Chicago must be treated separately. Selective schools in Chicago, as well as some traditional and charter schools, deviate from this pattern based on poverty. For any poverty level, Chicago schools perform better than those in the rest of Illinois. Selective programs for gifted students show high performance at each grade level, most notably at the high school level, when compared to other Illinois school types. The case of Chicago charter schools is more complex. Up to 2008, Chicago charter and neighborhood schools had similar performance scores. In the last few years, charter students' scores overtook those of students in traditional schools as the number of charter school locations increased.

Beyond linear regression: A reference for analyzing common data types in discipline based education research

“…Physics education researchers have used linear regression models to test many hypotheses related to educational outcomes such as student performance, e.g., Refs. [3][4][5][6][7][8] and student evaluation scores [9].…”