Conditional logit for multinomial log regression

Posted by edocoppo on Sep 24, 2014; 11:17am
URL: http://spssx-discussion.165.s1.nabble.com/Conditional-logit-for-multinomial-log-regression-tp5727378.html

Dear listers, I have run a series of multinomial logistic regressions for different country-year surveys. The DV is not ordered. I was told I should conduct some conditional logit analysis to confirm that each category of the DV is well assessed. I guess I can do that by running different multinomial logistic regressions with a different reference category of the DV each time, but this may take too long. Some IVs are continuous and some categorical. Do you know some way to proceed? Some step-by-step tip? Thank you.