Re: st: how to test for multicollinearity with xtreg

[David Jacobs <jacobs.184@sociology.osu.edu>]

Since OLS with dummies gives identical estimates to "xtreg, fe" or "areg", 
you can compute the case-specific (and time-specific if you are including 
them) dummies perhaps using tab (one way) generate.  Then run a standard 
OLS model with all dummies included and use Stata's regression diagnostics 
(like VIF).

You should be warned, however.  VIF isn't a strong indicator (because it 
ignores the correlations between the explanatory variables and the 
dependent variable) and fixed-effects models often generate extremely large 
VIF scores.

Dave Jacobs

At 06:20 PM 7/29/2005 -0500, you wrote:

> Dear STATA listserve,
>
> I'm running regressions on panel data using the xtreg command but the
> VIF command does not seem to work. I was wondering if anyone knows of
> any STATA commands to test for the presence of multicollinearity as an
> option with the xtreg command. Particularly, since there is little
> within country variation with one of my variables, but significant
> between country estimation, I suspect the fixed effects specification
> is causing collinearity with the fixed effect for each cross section
> and the variable mentioned above.
>
> I would be extremely grateful for any tips.
>
> Sincerely,
> Ross Hanig
>
>
> --
> Department of Agricultural and Consumer Economics
> 1301 West Gregory Drive
> 405 Mumford Hall
> Urbana, IL 61801
> rhanig2@uiuc.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/

*
*   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/