On 2004-10-12, at 17.01, Stas Kolenikov wrote:
Yes, -xtreg, [re|mle]- implicitly assumes MAR and provides consistent
and asymptotically efficient estimates
Thanks for confirming that assumption.
both -xtreg, re- and -gllamm- will be of
little help as they assume independence of observations taken at
different time points conditional on the random effect.
I know very little about econometrics but I realise that this might be
a problem also in psychological/experimental longitudinal research. I'm
currently reviewing a paper that use SAS (I assume PROC MIXED). The
authors just state that they use a compound symmetry correlation
structure with no motivation and no test. I have seen this before in
papers using PROC MIXED and GEE (exchangeable correlation). Do you
think this is a reasonable assumption or should I ask them to motivate
their choice?
Thanks
Michael
*
* 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/
