Hi there, I was wondering how any gologit2 users have plotted ROC curves and obtained an estimate of AUROC for ordinal outcomes in STATA. My ordinal outcome has 3 levels (mild, moderate, severe), I have predicted probabilities too. I've created new variables mild vs. mod/severe, mild/mod vs. severe and mod vs mild/severe for my outcome variable and predicted probabilities. I'm not sure what the appropriate command is that I can use. Lroc does not seem appropriate. Any pointers would be very welcome.