I would like to get within correlation for years, which has repeated measures in them also. I would appreciate any help. In detail: The dependent variable is milk yield on a test day. I have parity, which is 1 to 5 (like year) and Testday, which is repeated 9 times within parity. Subject is the id number for goats. When I specify subject=id(parity) with type=CS and rcorr, it gives me the correlation between test days, within parity. What I would like to do is to get the correlation between parities. I cannot add the numbers of the testdays to get one number per parity (which would make life a lot easier), because the testdays are different and there is the day effect that must be corrected for. Do you know of any ways to get within correlation for parity?