On the performance of some statistics in discriminant analysis based on covariates

In this paper we study the behavior of three statistics suggested for testing the hypothesis, H0 : [mu]1 = [mu]2, in the two sample case, in the presence of covariables. Power comparisons are made in the case when [delta]2, the difference of the mean vectors in the covariates, is not equal to zero. This extends an earlier paper of the authors [Sanklya Ser. B35 51-78], where [delta]2 was assumed to be equal to zero. The results reiterate those obtained in the above cited paper that for low observed values of Dq2 one would use t2 otherwise t3 would be recommended. The statistic t1 does not seem to be appropriate for testing this hypothesis.