Dear Statalisters,
Following William Gould's guideliness on how to construct a simple Chow test
(http://www.stata.com/support/faqs/stat/chow.html), I would like to ask if this
procedure is safe for reproduction following a panel data estimation, such as
the Between Effects estimation of:
y_it = a + bX_it + C_i + e_it
That is to say, estimate the above model as -xtreg, be wls- for the pooled model
t=1,2,...,10 and then estimate the same model for two separate periods
t1=1,2,..,5 and t2=6,7,..,10. Then, compute their error sum of squares and
compute the simple linear Chow test.
The problem as I understand it, is that the pooled model will be based on
averages of i over the entire t=1,2,...,10, where the two sub-estimations will
be based on averages of i over t1=1,2,..,5 and over t2=6,7,..,10.
Do you think that the computation of the Chow test as I described it above for a
BE estimation is correct? If not, I would appreciate it if you could offer some
advice or even guide me through some useful references. All comments are
welcome!
many thanks,
Dimitris
--
This mail sent through http://webmail.bangor.ac.uk
*
* 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/
