Re: st: running regressions for sub-samples

[Philipp Rehm <Philipp.Rehm@gmx.de>]

You could generate an identifier variable and condition your regressions 
 on it (or use -statsby- or the like).
You could generate a variable containing the median for each year of 
variable x3it:

bys year: egen median=median(x3it)

Then, you could condition your estimations:

reg yit x1it x2it x3it yr* if x3it < median
reg yit x1it x2it x3it yr* if x3it >= median & median != .

Using a similar logic, you could generate an identifier variable (say, 
id), and then run the regressions as: "bys id: reg ..." or using 
-statsby- (or something similar).

HTH,
Philipp

tabreez shams wrote:

> Hi,
> I am using the following two models on an unbalanced
> panel data over period 1997 to 2004:
>
> 1. reg yit x1it x2it x3it yr*
> 2. xtreg yit x1it x2it x3it yr*, fe small
>
> Now I want to re-run the regression for sub-samples
> based on the following criteria:
>
> if x3it is less than the median value of x3it in each
> year. Obviously, the other one is if x3it is equal or
> more than the median value of x3it in each year.
>
> Much appreciate if anyone can kindly provide the
> statement/macro to run the models for the sub-samples
> in Stata/SE 8.2.
>
> Regards,
> Shams
> PhD Candidate
> Accounting and Finance Dept.
> Monash University,
> Australia
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam?  Yahoo! Mail has the best spam protection around 
> http://mail.yahoo.com 
> *
> *   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/