Abstract
Closed test procedures are constructed based on Follmann's test (1996) and its nonparametric rank sum analog for testing multiple endpoints. These procedures are based on the stepwise closed test strategy of Lehmacher, Wassmer & Reitmeir (1991). The experimentwise type I error rate and power performance of the test procedures are compared to those of the Westfall and Young (WFY) bootstrap resampling closed procedure (1993) by simulation studies. The comparison focuses on one-sided alternatives. Both the Follmann-type and the bootstrap-WFY closed test procedures control the experimentwise type I error rate, and their power performances are similar.
∗The opinions expressed are those of the authors and not necessarily those of the U.S. Food and Drug Administration.
∗The opinions expressed are those of the authors and not necessarily those of the U.S. Food and Drug Administration.
Notes
∗The opinions expressed are those of the authors and not necessarily those of the U.S. Food and Drug Administration.