Scott
Thanks a lot for your suggestion. Your method seems to work well for the
coefficients of the variables but not for the intercepts which is what I
am really interested in. Because I am quite sure that the variance of
the two groups is not equal and since the Hausman test is indefinite I
would like to estimate two different models and compare the intercepts
of the two models. Is there a way to do that? Furthermore, is it
possible to generalize your approach for more than two groups of
individuals?
Thanks again.
Kind regards,
Daniel
*** Summary of the prior emails:
> I would like to estimate a fixed effects panel regression for two
groups
> of individuals (men and women, say). Then I would like to compare the
> coefficients of the two models. However, the -suest- command seems not
> to work in this instance
Yes, you could interact the variables with a dummy variable and test if
the
interacted variable of interest is 0.
For example:
webuse grunfeld, clear
mark gr if com <7
foreach v of varlist kstock mval {
gen gr_`v' = gr*`v'
}
xtreg invest kstock mval gr_*, fe robust
test gr_kstock
Scott
*
* 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/