I have used xttest3 to test for heteroskedasticity in a fixed effects model (xtreg, fe) based
on a large dataset (44th over 13 years). However, in one of the help-notes I have found that
xttest3 is not very accurate for large N and small t. In addition to
this, the dataset is very unbalancedand with lots of missing values.
Would it be correct to use xttest3 to test for heterosced in this type of dataset?
Many thanks,
Svetlana
On 26 May 2004 at 0:35, Clive Nicholas wrote:
> Svetlana Mira wrote:
>
> > Clive, unfortunately, if I set the memory to 30m and matsize 6500, I
> > cannot open the dataset
> > (the usual no room to add more observations). If I further increase the
> > memory, the matsize
> > decreases correspondingly. Any other ways of sorting this problem out?
>
> Well, I would either follow up Mark's very neat suggestion or to fire an
> e-mail to technical support ([email protected]) and see if they can
> help.
>
> > Mark, thanks for the fixed effects and cluster-robust standard errors
> > suggestion ;-). I will check it out
>
> CLIVE NICHOLAS |t: 0(044)191 222 5969
> Politics |e: [email protected]
> Newcastle University |http://www.ncl.ac.uk/geps
> *
> * 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/
*
* 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/
