Fixing the<i>F</i>Test for Equal Variances

Abstract

Two adjustments to the F test for variances are introduced that ameliorate the poor robustness properties of the test. A comparison of these adjustments to the classical F test, Levene/Brown-Forsythe's test, and squared ranks test demonstrates a substantial improvement for various sample sizes and distributions. One of the adjustments is extended to testing homogeneous variances in multiple samples and comparisons are made to Levene's, squared ranks, and Bartlett's tests for equal variability.