Heterogeneity in Student Academic Growth by Student Characteristics Across Schools

Using summative assessment data from a U.S. state, we measure student academic growth by school by student subgroup. We find that, within schools, growth is substantially correlated between students with and without disabilities. However, these correlations are not close to perfect, and imply substantive differences across schools in the growth of students with disabilities relative to growth among students without disabilities. We similarly find that growth is substantively but not nearly perfectly correlated within schools between students with low and high prior achievement. We find that differences in growth by prior achievement are substantively correlated within schools over time, while differences in growth by disability are modestly correlated over time in math and slightly correlated over time in ELA. The results suggest that there are some schools at which academic gaps by disability or prior achievement widen or narrow more so than others, and that measures of school growth that disaggregate by disability and prior achievement will often provide detail missed by overall growth measures. In contrast, we find that student growth is very highly correlated within schools across race/ethnicity and economic disadvantage subgroups.