We discuss the power of a certain multivariate one-sided test and a certain multivariate two-sided test for normal mean vectors based on the likelihood ratio test under the assumption that the covariance matrix is unknown. We calculate the power by using conservative critical values for a specified significance level derived by upper bounds of the distributions of the likelihood ratio test statistics. We give some numerical examples regarding critical values and power of the test.
Proceedings of the School of Science of Tokai University 48 1-9, 2013-03-20
Tokai University. School of Science