Hi - I have been using logistic regressions (PROC SURVEYLOGISTIC) to calculate AORs on replicated-weighted population survey data, such as the Tobacco Use survey of the Current Population Survey.  However, I have a new faculty member who prefers Risk Differences (with confidence limits, of course).  I have found a couple old references that describe doing this with PROC GENMOD, but that doesn't hand survey data.  Can anyone point me to a way to do this?  thanks -  ... View more