RE: st: Statistically significant difference in R Squared

["Santos Silva, J.M.C." <jmcss@essex.ac.uk>]

Dear Ani,

I guess that what you can do is to test if the variance of the errors is larger
in one period than in the other. In the spirit of the old Goldfeld?Quandt test, 
you can just do a BP test where you regress the square of the residuals on a 
constant and on a dummy that is 1 for the second period only. Then you can 
simply do a one sided t test on the coefficient of the dummy. Of course, your 
base model should allow the coefficients to differ in the two periods and 
therefore you should include the interactions that Dave suggested (or just 
estimate the model for the two periods separately).

All the best,

Joao

> Hi all,
> 
> I am running a time-series regression model over a 10 year period. I would be 
> interested in splitting the sample into two five year periods and finding out 
> whether the model has a statistically significantly higher R^2 during the 
> second period as compared to the first. Is there a test for this, and if so 
> is it possible to implement in Stata?
> 
> Many thanks for any help.
> 
> Kind regards,
> Ani


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/