On some nonparametric generalizations of Wilks' tests forH M, HVC andH MVC, I
- 42 Downloads
This paper is concerned with the nonparametric generalizations of the well-known likelihood ratio tests, proposed and studied by S. S. Wilks  (also see Votaw ), for testing the hypothesis of compound symmetry, i.e., equality of means (H M), equality of variances (H V), and equality of covariances (H C) of a multinormal distribution. In this part of the paper, some nonparametric rank order tests are offered for testing the hypothesisH M of equality of location parameters of a multivariate distribution of unspecified form. In the second part, the general problem of nonparametric tests for the hypothesesH VC andH MVC will be considered.
KeywordsAsymptotic Normality Noncentrality Parameter Permutation Distribution Rank Order Test Normal Score Test
Unable to display preview. Download preview PDF.
- Z. Govindarajulu, L. LeCam and M. Raghavachari, “Generalizations of Chernoff-Savage theorems on asymptotic normality of nonparametric test statistics,”Proc. Fifth Berkeley Symp. on Math. Statist. and Prob., Univ. Calif. Press, 1 (1965), 608–638.Google Scholar
- J. L. Hodges, Jr. and E. L. Lehmann, “Comparison of the normal score and Wilcoxon tests,”Proc. Fourth Berkeley Symp. on Math. Statist. and Prob., Univ. Calif. Press, 1 (1960), 307–317.Google Scholar
- M. Loeve,Probability Theory. D. Van Nostrand Co., New York, 1962.Google Scholar
- P. K. Sen, “Nonparametric tests for multivariate interchangeability, Part I, Problems of location and scale in bivariate distributions,”Sankhyā, Ser. A, 29 (1967), (in press).Google Scholar
- P. K. Sen, “On a class of aligned rank order tests in two way layouts”, submitted toAnn. Math. Statist., 1966.Google Scholar