Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Joanne Flavel <flav0031@flinders.edu.au> |
To | "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu> |
Subject | st: Predicted probabilities from gllapred using mlogit |
Date | Thu, 16 Aug 2012 19:23:47 -0700 |
Dear Statalisters I have estimated a 2 level multinomial logit for a model with a categorical dependent variable with 3 outcomes using gllamm . I am now trying to estimate average predicted probabilities but am having some trouble so was hoping for some advice My model is given by: gllamm depvar varlist, i(id) nrf(1) eqs(cons) base(2) link(mlogit) family(binom) robust trace I then used gllapred gllapred prob1, mu marginal outcome(1) gllapred prob2, mu marginal outcome(2) Outcome 3 is my base outcome I understand that for a logit, the average predicted probability can be obtained by the following: gen phat=exp(prob1)/(1+exp(prob1) I am stuck on how to calculate this for the multinomial logit. Can anybody please advise on the correct way to do this? Any help will be much appreciated Kind regards Joanne * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/