A modified Dunnett and Tamhane step-up approach for establishing superiority/equivalence of a new treatment compared with k standard treatments

Abstract

There are still some open questions whether the existing step-up procedures for establishing superiority and equivalence of a new treatment compared with several standard treatments can strongly control the type I familywise error rate (FWE) at the designated level. In this paper we modify one of the three step-up procedures suggested by Dunnett and Tamhane (16 (1997) Statist. Med. 2489-2506) and then prove that the modified procedure strongly controls the FWE. The method for evaluating the critical values of the modified procedure is also discussed. A simulation study reveals that the modified procedure is generally more powerful than the original procedure. © 2001 Elsevier Science B.V.