On 1/31/06, Richard Goldstein <[email protected]> wrote:
> 2. robust reports Huber-White t-statistics; cluster takes this
> to another level when you have clustered data; what is N when
> one has clustered data? That is, I think, the reason, why there
> is no adjusted R2 and is also the reason not to use the adjusted
> R2 from regress without the cluster option
If you are using -robust- or -cluster-, you admit that you have
problems with your second moments: either non-constant variance, or
correlated error terms, so the sum of squares per se does not tell you
that much about the real variability in your data. So anything based
on that sum of squares, like R^2 or F-test, would not be used in the
domain where it is applicable (i.i.d. errors, preferrably normal).
--
Stas Kolenikov
http://stas.kolenikov.name
*
* 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/
