st: Re: xttest2 doubt

[Christopher F Baum <baum@bc.edu>]

Prof. Vallejo,

I doubt that any results from the test have any value. Check the "obs per 
group" in the xtreg output. You will probably find that the min==2. The 
test is based
on a correlation matrix, calculated over the observations which are 
available for each unit. Apparently in your case there are only two time 
points at which all units are available. That does not prevent computation 
of the fe model, but the test based on a correlation matrix with two 
observations per unit is surely worthless.

Thanks
Kit Baum


--On Monday, June 17, 2002 17:37 -0500 Hernan Vallejo 
<hvallejo@uniandes.edu.co> wrote:


> Dear Sir,
>
> I have downloaded your xttest2 in order to test for heteroskedasticity in
> an fixed effects panel regression.
> However, when I have used it, it states
>
> "Based on 2 complete observations"
>
> �What that does mean and what does that imply in terms of the validity of
> the test?
> Kind regards,
>
> Hern�n Vallejo
> Profesor Investigador
> CEDE Facultad de Econom�a
> Universidad de los Andes
> Bogot�, Colombia

*
*   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/