I need a procedure to compare several mean vectors simultaneously. We have e.g. three multivariate normal distributed random vectors with mean m1,m2,m3 and the same covariance matrix Sigma. Each random vector has 4 variables. I want to implement a test on Mathematica.
Now we want to test simultaneously
of these three random vectors and not of each variable. In univariate case it is the Scheffe, Tukey and so on, but in multivariate case I don`t know.
Anyway, which teststatistic, I would need the teststatitic and the quantiles or how to compute p-values for that test.
So, please if this, is not a problem for you, please help me and send your answer directly to me:
hope to hear from you soon
Institute for Mathematics, University of Innsbruck, Austria e-mail:W.Hitzl@uibk.ac.at