Do Small Classes Lead to Larger Variability in Student Achievement? Evidence from a Randomized Experiment

Spyros Konstantopoulos, Northwestern University
Amelie Constant, Institute for the Study of Labor (IZA) and Georgetown University

Previous research from randomized experiments on class size indicates positive immediate and long-term effects of small classes on average achievement for all students. However, the effects of small classes on the variability on student achievement have not been examined. Differences in achievement variability between smaller and larger classes are of great importance, since they indicate differential effects of small classes across different types of students (e.g., lower or higher achieving students). Since previous findings on the social distribution of the small class effects are unclear, the present study uses data from a four-year large-scale randomized experiment to examine the effects of small classes on the achievement gap. The results indicate that higher achieving students benefit even more from being in small classes in early grades than other students. However, there is no indication that small class reduction is a mechanism that can reduce the achievement gap.