Bootstrap Methods for Testing Homogeneity of Variances

This paper describes the use of bootstrap and permutation methods for lhe problem of testing homogeneity of variances when means are not assumed equal or known. The melhods are new in this context, and nontrivial, since lhe composite null hypothesis involves nuisance mean parameters. They allow the use of normal-:'theory test statistics such as F = sUs~ without the normality assumption which is crucial for validity of critical values obtained from the F distribution. In a Monte Carlo study the new resampling methods are seen to compare favorably with older methods, except in the case of heavily skewed distributions.