Keynes M. Smith--
Kindly address Stata questions to Statalist. The cluster option
estimates SEs robust to arbitrary intra-cluster correlation, assuming
you have enough clusters to allow correct inference, with 50 or more
being a good rule of thumb. Thus, if you are worried about serial
correlation within US states, and you have data on 50 states and the
District of Columbia, you can
xtreg y x, cluster(state) fe
but -xtivreg2- offers a panoply of SE options. -findit xtivreg2- and
-findit ivreg2- and install, then read the help files. As for
heteroskedasticity, you may be interested in -xtscd- and -xttest3-
(both user-written).
You may also want to invest in a copy of Baum's book:
http://www.stata.com/bookstore/imeus.html
On 4/19/07, Keynes M. Smith <[email protected]> wrote:
Hi,
Could you please suggest me how to correct autocorrelation in fixed
effect model?
Besides, how can I check heteroskedastic in fixed effect model.
Regards,
A. Sura
*
* 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/
