There are several combination-based permutation tests for testing hypotheses within the framework of Randomized Complete Block design (Arboretti, Corain and Ragazzi (2010)). The method is based on extensions of the nonparametric combination of dependent permutation tests (Pesarin (2001)) will be compared. After a critical revision of the methods proposed in the literature, we compare such methods and the one proposed through an exhaustive simulation study in terms of power and under H 0. We refer to a typical psychometric study when using ordered categorical response variable within the framework of randomized complete block designs such as in the case of sensorial evaluations.