ADJUSTING PAIRWISE NONPARAMETRIC EQUIVALENCE HYPOTHESIS TESTS AND CONFIDENCE INTERVALS FOR PERIOD EFFECTS IN 3×3 CROSSOVER TRIALS