st: Handling two-way clustering

[Edward White <e.white@yale.edu>]

Hi, Statalisters.

I'm looking for a method to handle clustering on two factors: repeat  
binary measures from individuals who are in small groups.

I can use -xtgee- to handle clustering by individual, but it seems  
that  it'll only handle one level of nesting. But how to handle the  
potential correlation between group members? Upon trying a simulation  
of the data I anticipate collecting there won't be sufficient  
variation to use -gllamm-. And I'm not really wanting to understand  
the variance contributed by those groups, just to account for it. Any  
suggestions? Thanks!


Ted

Edward White, PhD
Postdoctoral Fellow
Yale Center for Interdisciplinary Research on AIDS

135 College St #200
New Haven CT 06510
United States
+1-203-764-8119
fax
+1-203-764-4353


*
*   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/