Discriminant Analysis Within-Group versus Separate-Group Covariance Matrices Predict


New Member

I'm hoping someone can help. I have ran a discriminant analysis along with the prediciton rates for the 5 groups in my study. I am wondering the reason why I get different prediction rates when using within-group versus separate group covariance matrices even though my coefficients are the same. And, what is the difference between these two.

Many thanks for any help.