Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
| From | Maarten Buis <maartenlbuis@gmail.com> |
| To | statalist@hsphsun2.harvard.edu |
| Subject | Re: st: predict(pu0) after clogit |
| Date | Tue, 27 Mar 2012 12:14:23 +0200 |
On Tue, Mar 27, 2012 at 11:55 AM, Ori Katz <ori.k66@gmail.com> wrote: > What do you think about using "margins, dydx(*) predict(pu0)" in my > case, after the clogit? > > In the results table it says "Pr(choice1|fixed effect is 0)", but I > don't understand what does it means by "fixed effect is 0". I don't > have fixed effects in the model, only majors income and interaction > between major income and the ethnic origin of the individuals. You have fixed effects: that is what -clogit- is giving you. If you do not want fixed effects than -clogit- is the wrong command. I don't think that the - predict(pu0)- option in -margins- is what you want as it is a rather weird hybrid between average marginal effects and marginal effects at average values of explanatory variables. The fixed effects are implemented such that the value 0 represents the average (which is why you have a constant in -clogit-). So your margins command gives you marginal effects averaged over your observed variable but at the same time fixing the fixed effects at its average. The best thing you can do is work with odds, odds ratios, and ratios of odds ratios. Hope this helps, Maarten -------------------------- Maarten L. Buis Institut fuer Soziologie Universitaet Tuebingen Wilhelmstrasse 36 72074 Tuebingen Germany http://www.maartenbuis.nl -------------------------- * * 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/