09-15-2015 04:52 PM
I hope everyone is doing very well. I am fitting some log linear models using proc Catmod. The data we have is all categorical. All variables are being treated as independent variables, our outcome is the cell counts. The data that I am using comes from different studies with different study designs (case/control etc) and populations. Therefore, if we just pool the data and analyze it we are afraid that we could introduce some study/cohort effect in the results. For this reason, we would ike to adjust for study/population differences in the proc catmod model. Does anybody have any experience or suggestions on how to approach this? I was looking into the population statement in proc catmod--will this do it?
Thank you so much for any insight you can offer,