Dear Users, I am conducting a function analysis with 6 groupings and 11 predicter variables. fit<-lda(Groups~ x1 + x2+ ....+ x11, data=my.data, na.action="na.omit") My resulting misclassification rates are low, and although this is one of the goals, I would like to test assumption of equality among the covariance matrices for groups. I have been attempting to find a way do do this, but have been unsuccessful. I am a relatively new R user, and perhaps I am overlooking the obvious. Could anybody guide me through the procedures for conducting such a test for equality of covariance matrices, either from the resulting output from the lda model or from the original dataset? Thank you very much. Cheers, Rob -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Robert Lonsinger Wildlife Grad. Research Assistant New Mexico State University Dept. of Fish, Wildlife, & Cons. Ecology rob.lonsinger@gmail.com 484.459.1977 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ [[alternative HTML version deleted]]