st: gllamm, diagnostics and model fit random intercept model

[Martina Brandt <m.s.brandt@web.de>]

gllamm [vars], i(person country) link(logit) family(binom) nip(10) adapt 
robust
(random intercept models; ~3000 dyads, ~2400 persons, 10 countries, one 
level 3 variable per model)

4 questions:
1. what kind of diagnostics are necessary?
2. how should the model-fit of a multilevel model of that kind be assessed?
3. how can (non-hierarchical) models with different level 3-variables be 
compared? Is BIC adequate?
4. And is there any possibility to compare models with different level 
3-variables in different subgroups with differing n?

thanks in advance!

martina
*
*   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/