Dear All
I have run the hausman test and this rejected random effects in favour of
fixed effects. However, my panel of data is suitable for xtgls regression
and I have tried this out [with a preferred version using p(h) and c(p)].
My questions are as follows:-
1. Does the Hausman test referred to above rule out xtgls?
2. If not, can I effectively compare model performance between xtreg FE
and xtgls i.e. how do I distinguish between Rsquared of 0.06 (FE) and log
likelihood of -2984.167 (xtgls)?
Many thanks in advance.
Ron Dorsey
*
* 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/
