Power considerations for Kolmogorov–Smirnov and Anderson–Darling two-sample tests

The two-sample Kolmogorov–Smirnov and Anderson–Darling tests assess the hypothesis that two given samples come from the same population. In this paper, the power of each test was measured using a variety of alternative distributions and varying the sample size. Recommendations for the more powerful test for each common distribution are given, depending on whether the distribution has heavy or light tails.