A Monte Carlo study was done to find the Type I error rates for three nonparametric procedures for making k − 1 many-one comparisons in a one-way design. The tests (t) were the Silverstein and Steel many-one ranks tests and the two-sample Wilcoxon rank-sum test, k = 3, 5, 7, and 10 treatments were crossed with n = 7, 10, and 15 replicates with 1000 simulations per k, n combination. Analyses of four Type I error rates showed that: (1) The Wilcoxon test had the best comparisonwise error rates; (2) none of the tests functioned well as protected tests; and (3) the Silverstein test had the best experimentwise error rates and was the recommended procedure for many-one tests for a one-way layout.
References
1.
BernhardsonC. S.Type I error rates when multiple comparison procedures follow a significant F test of ANOVA. Biometrics, 1975, 31, 229–232.
2.
ChurchJ. D.WikeE. L.Two Monte Carlo studies of Silverstein's nonparametric multiple comparison tests. Psychological Reports, 1980, 46, 403–408.
3.
DunnettC. W.A multiple comparison procedure for comparing several treatments with a control. Journal of the American Statistical Association, 1955, 50, 1096–1121.
4.
GamesP. A.Multiple comparisons of means. American Educational Research Journal, 1971, 8, 531–565.
5.
HollanderM.WolfeD. A.Nonparametric statistical methods. New York: Wiley, 1973.
6.
MillerR. G.Jr.Simultaneous statistical inference. New York: McGraw-Hill, 1966.
SteelR. G. D.A multiple comparison rank sum test: Treatments versus control. Biometrics, 1959, 15, 560–572.
9.
WikeE. L.ChurchJ. D.Analysis of variance methods for the design and analysis of Monte Carlo statistical studies. Bulletin of the Psychonomic Society, 1977, 10, 131–133.
10.
WilcoxonF.KattiS. K.WilcoxR. A.Critical values and probability levels for the Wilcoxon rank sum test and the signed rank test. Pearl River, NY: American Cyanamid Co., & Lederle Lab., 1963.
11.
WinerB. J.Statistical principles in experimental design. (2nd ed.) New York: McGraw-Hill, 1971.
