Randomization-based inference and the choice of randomization procedures (Q2423167)

It is known that the randomization tests constitute a method of inference for analyzing data from randomized experiments. In particular, regarding the testing of significance of treatment effects in randomized clinical trials, the randomization-based inference is distinguished from the population-based parametric and nonparametric inference (t-test or permutation tests) taking into account the preservation of type I error, the relation of power to the randomization procedure as well as the flexibility in choosing the test statistic. The aim of the present paper is to revisit rationale of the above-mentioned properties by providing justification through simulations. Through that process the authors managed to find out that the choice of randomization procedures as well as the analysis of randomized clinical trials can be facilitated by the application of randomization-based inference.