Abstract:
|
In this paper, we propose a flexible multiple testing procedure that controls the false discovery rate (FDR) under positive dependence condition in a group sequential design. This procedure is constructed using the error spending function approach and it is a natural extension of the original BH method to a multi-stage procedure in a group sequential set- ting. An adaptive version of this group sequential BH procedure is also introduced, which is proved to make the FDR control under independence. Simulation studies were performed to investigate the performance of the proposed procedure in terms of power, FNR and the sample size saving, compared to its relevant competitor, relative to the fixed sample BH method applied to the full data.
|