Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

re: Re: st: panel data hausman negative


From   "Mark Schaffer" <M.E.Schaffer@hw.ac.uk>
To   statalist@hsphsun2.harvard.edu
Subject   re: Re: st: panel data hausman negative
Date   Tue, 30 Sep 2003 16:39:20 +0100

Paula,

To:             	statalist@hsphsun2.harvard.edu
From:           	paula garcia <paulagarciag@latinmail.com>
Date sent:      	Tue, 30 Sep 2003 11:31:47 -0400
Subject:        	st: X-Mailer: LatinMail v3.0 -- http://www.latinmail.com.com
Send reply to:  	statalist@hsphsun2.harvard.edu

> Mark, thanks again. I have looked for the xthausman command, but stata
> says "(help hausman for replacement)". I have also tried with:
> 
> xtreg y x, fe
> est store fixed
> xtreg y x, re
> hausman fixed .
> 
> but the hausman statistic is still negative. How can I use the suest
> command if xtreg does not support the score option?

It looks like I have to bow out, because I still (!!) haven't yet 
upgraded to Stata 8.  It should be any day now, but that isn't good 
enough for this thread. :(

Anybody else out there have any ideas?

--Mark

> 
> Related to your second check, the sigma_u is different from zero.
> 
> Thanks,
> Paula
> 
> 
> > Mark, thanks for your answer. The problem is the following.
> > I want to know if the best estimation is a fixed or a random
> > effects. So I use the Hausman test:
> > 
> >  xtreg y x, fe
> >  hausman, save
> >  xtreg y x, re
> >  hausman
> > 
> >  I want to run three versions of models, and they have more
> >  or less the same variables. In the first two, the p-value of
> >  Hausman is 0.0000, so I use xtreg, fe. However, when I 
> run the third model, the HAusman statistic is negative.
> 
> A couple of checks:
> 
> - Do you get the same results if you use the built-in -
> xthausman- command?
> 
> - Have you checked to see if your random effects estimation has
> inadvertently reduced to pooled OLS?  This will show up as sigma_u=0.
> 
> >  I have read that this means that with my sample, I have 
> no evidence to reject the null, so I would have to make a random
> effects for this third model.
> 
> >  However, for me is strange the result on the third model 
> >  (random effect), since more or less the variables are the 
> >  same ones than in the first two models, and the population is, of
> >  course, the same. So I would like to test 
> the hypothesis of random effects by means of other command,
> like suest.
> > 
> >  Thank you in advance.
> >  Paula
> 
> 
> _________________________________________________________
> http://www.latinmail.com.  Gratuito, latino y en espaņol.
> 
> *
> *   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/


Prof. Mark E. Schaffer
Director
Centre for Economic Reform and Transformation
Department of Economics
School of Management & Languages
Heriot-Watt University, Edinburgh EH14 4AS  UK
44-131-451-3494 direct
44-131-451-3008 fax
44-131-451-3485 CERT administrator
http://www.som.hw.ac.uk/cert

*
*   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/



© Copyright 1996–2021 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index