06-29-2017 10:02 AM
I would like to know if it is possible with SAS to do a generalized ordered logit (different parameters for each response categories, such as in a multinomial logit).
With stata, I think it is gologit2, but I didn't find the equivalent function with SAS.
06-30-2017 10:06 AM
I'm not sure why, but when I add this option, it says : "The UNEQUALSLOPES option is ignored."
And the output is the same as a normal glogit. How should I use it?
proc logistic data=edu.educationClean; class cohort4(ref='2') cntr_regr(ref=first) /param=ref; model edu3(descending)= cntr_regr cohort4 cntr_regr*cohort4 /link=glogit UNEQUALSLOPES=edu3; run;
Need further help from the community? Please ask a new question.