New Research Validates Studies Showing Small Schools Reduce Power of Poverty on Achievement

A recent study that attempted to identify reasons other than school size to explain the apparent ability of small schools to reduce the impact of poverty on student achievement concluded that the findings of earlier studies on the subject are valid: smaller school size is indeed associated with a weakening of the relationship between poverty and student achievement.

Theodore Coladarci analyzed 8th grade reading and math scores in Maine to test the hypothesis that a statistical artifact, and not school size, is responsible for the differential impact of socio-economic status on achievement in small and large school settings. The report, “School Size, Student Achievement, and the ‘Power Rating’ of Poverty: Substantive Finding or Statistical Artifact?” was published by Education Policy Analysis Archives and is available at

At issue in Coladarci’s analysis are a series of studies concluding that poverty has less influence over achievement in smaller schools and districts. (For a summary of the studies, see RPM, April 2004.) Many of these studies were conducted by or in cooperation with the Rural Trust. The Rural Trust coined the term “poverty’s power rating” to describe the statistic that measures the strength of poverty’s influence over achievement.

Coladarci, a researcher at the University of Maine and editor of the Journal of Research in Rural Education, has previously called attention to the greater year-to-year variation of aggregate test scores in smaller schools. (In “Gallup Goes to School,”, he recommends caution in using such scores to evaluate schools for such things as the Adequate Yearly Progress requirement of No Child Left Behind.) The recent paper deploys several statistical procedures to determine whether “poverty’s power rating” might be the product of this volatility.

In the end, Coladarci rejects the study’s hypothesis and concludes: “In short, the celebrated interaction of poverty and school size has survived a sincere attempt to empirically cast doubt on it. Consequently, we can have greater confidence in this interaction than I believe was warranted before.”