Evaluation of a Protocol for School Assignment Prioritizing Equity and Diversity

• In Event: Equity and Effectiveness: Evaluation of Policies and Programs

Abstract

Recent polling shows strong support for policies that desegregate schools along racial and socioeconomic lines. This study tested a protocol for school assignment and compared the resulting demographic and academic distributions against the current assignment system. The new model showed an average 55.3% reduction in the variability of student body composition between schools along all demographics. In the current model, schools explained 19.4% (p<.0001) of the variance in achievement compared to 0.002% in the new model (p=0.54). Pairwise-comparisons and effect sizes showed no differences between schools in the new model unlike in the current model. This protocol demonstrated potential as a policy solution for school assignment yielding undifferentiated schools that were demographically and academically representative of the overall school district.

Author

• David Stevens, University of California - Berkeley