meeri at sas.upenn.edu wrote: > Hello, I have a question about designing the G matrix and adding a > single covariate of no interest. Does it matter what scale we use > for our covariate? Not at all -- try rescaling it and you'll see you get the same results. dan