st: xtlogit, margins & xtrho

[mmamiti <mmamiti@uic.edu>]

Hello everyone,

I am running xtlogit models and in some cases I have noticed that there is a
large discrepancy between the predictive margins obtained from -margins,
predict(pu0)- and the marginal probability obtained from -xtrho- (with the
-xtrho- estimate being much closer to the actual probabilities).  I
understand that -margins, pred(pu0)- sets the random intercept at 0 (the
mean), while -xtrho- includes sigma_u in its computations.  My question is,
what does this discrepancy suggest?  Maybe are the random effects not
normally distributed?  Is there a way to diagnose the normality of random
effects?  Does this mean I shouldn't use this  model?

Thank you,

Mary E. Mackesy-Amiti 
-- 
View this message in context: http://statalist.1588530.n2.nabble.com/xtlogit-margins-xtrho-tp5623255p5623255.html
Sent from the Statalist mailing list archive at Nabble.com.
*
*   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/