Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Nick Cox <njcoxstata@gmail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | Re: st: xtreg, fe and xtlogit, fe |
Date | Thu, 13 Oct 2011 15:11:16 +0100 |
You are fitting quite different models here. Without knowing anything about your data, it is quite possible that at least one makes little sense, depending on the nature of your response variable. But even if both make sense, discrepant results don't seem surprising at all. Orthogonally to all that: 1. It is difficult to comment on your results when you just describe them verbally. 2. Plot your results in as many ways as you can to try to see what is going on. Nick On Thu, Oct 13, 2011 at 3:04 PM, natasha agarwal <agarwana2@googlemail.com> wrote: > I was trying to estimate the following model > > xtreg x a y y*g, fe vce(robust) > xtlogit x a y y*g, fe > > I get that the both y and the interaction term is insignificant in the > xtreg, fe model while y is significant and the interaction term is > insignificant in the xtlogit, fe model. > > I fail to understand the discrepancy I observe in the results. * * 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/