Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: prediction using a random effects logit


From   Xiaodong Chen <cxdgis@gmail.com>
To   statalist <statalist@hsphsun2.harvard.edu>
Subject   st: prediction using a random effects logit
Date   Tue, 12 May 2009 11:32:52 -0400

Hi stata users,

I want to apply a random effects logit model to predict the
probability for observations where the dichotomous outcomes are not
observed. If I assume no random effects in the prediction, I can
simply use "exp(xb)/(1+exp(xb))". If I do not assume no random
effects, I can use “gllamm y x1 … xk, i(clusterid) link(logit)
family(binom) nip(30) adapt” and “gllapred pr, mu” to predict the
probability for observations where the dichotomous outcomes are
observed. But how can I predict the probability for observations where
the dichotomous outcomes are NOT observed without assuming that there
is no random effects?

Thanks,
Xiaodong

*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index