The Power of Hypothesis Testing in Special Education Efficacy Research

Abstract

The efficacy of special class placement for mildly handicapped children was an early focus of special education research and has been the subject of at least two narrative reviews and one meta-analysis. In general, special class placement has proved to be no more effective than regular class placement on measures of achievement and social adjustment. However, the power with which the efficacy of special classes has been tested is unknown; thus inadequate statistical power represents an alternative explanation for the findings. The power of the statistical tests in 35 efficacy studies to detect small, medium, and large effect sizes was determined. The power to detect a medium effect size where one existed was no better than chance. However, 21 studies had adequate power to detect a large effect size. The effect sizes for these 21 studies on academic and social measures were highly variable and their averages quite small. Thus, conclusions drawn from previous reviews may be valid, in spite of the failure of earlier reviewers to consider statistical power. The implications of the present analysis for researchers, policy makers, and funding agencies are discussed in relation to the Regular Education Initiative and contemporary efficacy research.