All,
I posted this a few days ago but saw no responses, perhaps
because there was flood of activity on the list immediately
subsequent. Or perhaps because I didn't state it clearly
enough. I'll try to address the latter, at least.
-xtlogit, re- estimates a random intercept logistic
regression model, and returns estimates of the variance
of the random (group) error term u0. What I want to get
are the group residuals u0 for each panel, or at least
a predicted value that includes them.
After -xtmixed-, there is a -predict, fitted- that produces
predictions with group level error terms included. However,
-predict- after -xtlogit,re- only allows a linear prediction
which assumes u0 = 0, and a probability prediction which
assumes u0=0. What I want is a linear or probability prediction
without this assumption.
Given that -predict- doesn't do this, does someone know a
trick for getting u0 for each group? Or a theoretical reason
that I shouldn't be able to? I think I could roll my own,
but it would be easier to detour through a software package
which will give me these directly.
thanks,
Jeph
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/