An adaptive allocation for continuous response using Wilcoxon-Mann-Whitney score

In clinical trials, adaptive designs are sometimes used to force imbalance in allocation between two or more competitive treatments. Existing works are mostly based on dichotomous or polychotomous responses. In the present work, assuming that the responses are continuous variables, we investigate a two-treatment allocation scheme based on Wilcoxon-Mann-Whitney scores. The proposed scheme is used to provide two nonparametric test procedures for identity of two univariate continuous populations against location alternatives. Again, we give an exactly distribution free solution for generalized Behrens-Fisher problem. (C) 2003 Elsevier B.V. All rights reserved.