On Thursday 07 December 2006 21:35, Claude Francoeur wrote:
> Our model contains a dichotomic dependent variable. Our data is organized
> in a panel (100 firms over 5 years). One of our revisors recommends a panel
> regression instead of the OLS that we did. We had used EViews to do our
> regressions. EViews does not permit panel regressions with fixed or random
> effects in a binary model. I own Stata 9 and I am in the process of
> learning how to use it. Is there anyway, with Stata, to do a panel
> regression while controlling for fixed or random effects, considering our
> binary dependent variable? Any other way to test the robustness of our
> results?
Yes, there is.
Try out the command
. xtlogit depvar indepvarlist, re
for the random effects specification, and
. xtlogit depvar indepvarlist, fe
for the fixed effects specification (aka conditional logit).
I don't know Eviews. However to use Stata's panel commands your data must be
organised in a specific way. You might take a look at -help xt- and -help
reshape-, to start with.
Uli
--
Ulrich Kohler
[email protected]
030/25491-361
*
* 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/
