01-05-2016 05:17 PM
I am conducting a survival analysis using the link=cloglog for interval censored time to event data. Most of my covariates are time dependent and measured at the interval prior to the outcome being measured. I need to test multicollinearity but am not sure how to do this with this type of analysis. Normally I would use proc reg with the vif and tol options but since I have repeated measures data I don't think that is appropriate. I have participants with up to 4 visits of data. I can check collinearity at each visit separately but am also not sure if that would be technically correct. I know for survival analysis I should be able to just use proc reg with my event time as the dependent variable but I don't have event time as a variable, I have a binary yes/no outcome. Any suggestions would be greatly appreciated.