Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: Predict residuals after xtpcse


From   <lmg2005@columbia.edu>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: Predict residuals after xtpcse
Date   Mon, 30 Jul 2007 17:49:45 -0400

Dear Stata listservers:

I would like to recover the residuals after an xtpcse estimation using the
Stata command "predict". However, I am unable to do so because xtpcse does
not allow "predict varname, residuals" as a post-estimation command after
xtpcse. I could get the predicted values and compute the residuals as the
difference between the dependent variable and the predicted values but I
wondered if anyone had written an ado file in order to perform this step.
More to the point, I am interested in using the residuals in order to test
for serial correlation in a panel model with fixed effects estimated with
panel corrected standard errors using xtpcse. Wooldridge (2002, p. 275) lays
out a method for doing this in a panel model with fixed effects. Has anyone
written an ado file to perform this test? The test is somewhat different
than the existing xtserial which assumes that the dependent variable is
first differenced and follows Wooldridge (2002, p. 282-3). 

With all best wishes and many thanks! Lucy

Lucy Goodhart
Assistant Professor of Political Science
Columbia University
lmg2005@columbia.edu

*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index