st: How to obtain level 2 confidence interval from multilevel modeling with binary outcomes

["Song, Lin" <Lin.Song@kingcounty.gov>]

Hi Stata users,

I'm running a random intercept model using xtmelogit with covariates on both levels.  After talking to a few people online, my impression is that it is not possible to obtain SE and confidence interval at level 2 (e.g. school, hospital, or district) from xtmelogit post-estimation.  A couple of possible solutions seem to be using bootstrapping  or gllamm.

If my original command is:

-xtmelogit smoking var1 var2 || school:		/*var1 is individual level and var2 is school level covariates */

Is it possible to obtain the confidence interval  at the school level using xtmelogit post-estimation?

What would be the command for bootstrapping xtmelogit to obtain confidence interval at the school level?

What would be the command to obtain confidence interval using gllamm?

Any answer to the above questions should  be helpful.

Thanks a lot.

Lin

Lin song
Public Health - Seattle & King County
 





*
*   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/