A simulation study was conducted to compare the levels of significance and power between Schuirmann's and nonparametric two one-sided tests procedures for a 2 x 2 crossover design under different combinations of sample sizes, intrasubject variabilities, and underlying distributions. Empirical results suggest that Schuirmann's two one-sided tests procedure is robust to minor departure from the assumption of normality.
No takes yet. Share an insight, caveat, or question.
Liu et al. (1993) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: