Re: st: Bug in margins

[Sue Kay <professors@gmail.com>]

Hi,

I don't know whether it is due to a bug, or I am missing something, but 
I have some problems with margins command after I run my model with 
meqrlogit (with two random slopes).

Even I add the prefix i. to the factor variables in my model, I still 
got the dreaded "default prediction is a function of possibly stochastic 
quantities other than e(b)" message (which I get only if I don't use i. 
prefix with the variable I try to calculate my margins after xtlogit)

Here is a sample of what I do:
meqrlogit y i.x1 i.x2 i.x3 x4 x5 x6 x7 || cluster: x1 x2

And when I try to do this:
margins x1

I receive:
default prediction is a function of possibly stochastic quantities other 
than e(b)

I tried running my model using xtlogit (just with random intercept of 
course), and the margins command works without a problem.

Am I missing something?
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/