Comparing the Powers of the Wald‐Wolfowitz and Kolmogorov‐Smirnov Tests

Abstract The Wald‐Wolfowitz test and the Kolmogorov‐Smirnov test are two well‐known tests which can be used to test for differences between two population distributions, where these distributions could differ in means, variances or shapes. This paper compares the simulated power of the Wald‐Wolfowitz test to that of the Kolmogorov‐Smirnov test in several situations. Differences in location only, differences in variance only, and differences in both location and variance are considered. In most cases, equal sample sizes of 10 and 20 are used.