On power and sample size determinations for the Wilcoxon-Mann-Whitney test

Abstract

This article considers power and sample size calculations for the Wilcoxon-Mann-Whitney test. An exact variance large-sample method is introduced and explicit formulas are derived for the uniform, normal, double exponential and exponential shift models. A Monte Carlo Simulation Study is conducted to evaluate the exact variance procedure and compare its performance with some other large sample methods. The results show the remarkable accuracy of the suggested formula and the fundamental limitations of the other approximations. Practical considerations in determination of sample size for the two-sample location problem are also discussed.