1. Department of Mathematics , Kumanioto University , Faculty of Science, Kumamoto, 860, Japan;2. Department of Mathematics and Informatics , Tokyo Gakugei University , Tokyo, Japan, 184
Abstract:
We consider the problem of constructing a fixed-size confidence region for the difference of means of two multivariate normal populations It is assumed that the variance-covariance matrices of two populations are different only by unknown scalar multipliers Two-stage procedures are presented to derive such a confidence region We also discuss the asymptotic efficiency of the procedure.