Class size and student achievement in the United States: A meta-analysis

The purpose of this meta­analysis is to review the effect of class size on student achievement. Class size reduction (CSR) is one of the most important and interesting educational issues in the world, especially since it is one of the most expensive choices in educational policy. In this meta­analysis, 17 studies (8 published and 9 unpublished) were analyzed using the random­effects model. The results of this review suggest that student achievement in small classes is better than that of large classes by .20 standard deviations. A fixed­effects categorical analysis was also conducted to find the sources of variance and moderator variables that predict the effects of CSR. First, effect sizes were higher in published studies than in unpublished studies. Second, in terms of school subjects, the results of CSR were generally positive. Third, the effect of CSR on student achievement was larger in elementary schools than in secondary schools. Fourth, the results of CSR were generally positive, except in 10th grade. Fifth, the results of CSR are mixed, but generally positive by location of states.