Dear all,
I have a panel data set and have tested for presence of hetroskedasticity
using xttest3 and autocorrelation using xtserial, the results of which
show presence of both heteroskedasticity and autocorrelation . I would
like to run a fixed effect model for this panel controlling for
Heteroskedasticity and Autocorrelation.
Tried running an xtgls command with cross-section dummies and panels(het),
corr(ar1) force option but the result does not match
with that of the fixed effect model without controlling for het and ar1.
Almost all the terms in the xtgls model are significant at all levels
whereas this is not the case in the fixed effect model without controlling
for het and ar1. I do not understand why this should be the case.
Can one use the prais command with cross-section dummies for this purpose.
Helpful comments are most welcome.
Thanks
Chidambaran
*
* 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/