SAMPLE SIZE DETERMINATION FOR MULTIPLE MANY-ONE AND PAIRWISE COMPARISONS OF PROPORTIONS USING THE ARCSIN TRANSFORMATION

Our paper provides sample size formulas for multiple comparisons of k treatments with a control as well as with pairwise comparisons of k treatments when observing proportions. The test statistics we regard use the arcsin transformation of the proportions. We determine sample sizes with preassigned all-pairs, any-pair and per-pair power considering the control either of the comparisonwise or of the familywise type I error rate.