Describing School Effects with Residual Terms

Abstract

This study attempts to make more accessible an approach that can be used to rank the performance of individual schools using residual terms from a hierarchical linear model. Specifically, the study describes procedures that can be used to summarize the effect of schools when interactions between school practice and student background exist and applies them to a fairly realistic school effects data set. This article also serves as a reminder of the importance of considering differential school effectiveness by illustrating problems that can occur when the school effect is summarized by a single quantitative indicator.