An empirical investigation of the power of the F-test when the underlying distribution is nonnormal and the within-cell variances are not homogeneous. It is shown that even for small samples the analysis-of-variance F-test is valid, although the usual assumptions of normal distribution and homogeneous error variances are seriously violated. The test is shown to be valid for both Type I and Type II error. An analysis of the correlation between the numerator and denominator of the null hypothesis F-ratio demonstrates that the distribution of F is little affected by the shape of the underlying distribution.
This report is part of the RAND Corporation Research memorandum series. The Research Memorandum was a product of the RAND Corporation from 1948 to 1973 that represented working papers meant to report current results of RAND research to appropriate audiences.
This document and trademark(s) contained herein are protected by law. This representation of RAND intellectual property is provided for noncommercial use only. Unauthorized posting of this publication online is prohibited; linking directly to this product page is encouraged. Permission is required from RAND to reproduce, or reuse in another form, any of its research documents for commercial purposes. For information on reprint and reuse permissions, please visit www.rand.org/pubs/permissions.
The RAND Corporation is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.