RE: st: Xtlogit: two different groups, two different RE

["Marsh, Kimberly" <k.marsh07@imperial.ac.uk>]

Hi - In follow up to my original question (below), thanks for the
suggestion to look at xtmelogit. I have had a look at the code, however,
and it seems to be fitting nested RE (or crossover), whereas my two
random effects (HHID and clinics) are not nested in one another/not a
crossover. These measurements of prevalence in the groups come from
completely different sampling frameworks/studies. 

Any further suggestions on how to handle this in stata would be very
helpful. Thanks in advance.

Original post:

Hi - I wonder if anyone has any ideas on how STATA can handle this
scenario: 

I need to fit an xtlogit model to compare trends in HIV prevalence
estimates (as measured over three points in time - Round 1, Round 2 and
Round 3) between two different groups (Group 1; Group 2). 

Because of the differences in the sampling frameworks in the two groups
(the first group is comprised of people in households, the second group
is comprised of people attending clinics), I therefore have two
different random effects (HHID in Group 1; Clinic_num in Group 2). 

Any suggestions on what the STATA code for the xtlogit command would
look like when there are two different random effects?




*
*   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/


*
*   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/

*
*   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/