Page 1 of 1

VIF scores panel data

Posted: Wed Sep 17, 2014 8:49 am
by daan
Hello,

I have a question concerning VIF scores and Eviews.

I have read that VIF scores aren't available in Eviews for panel data, is this still true?
Additionally i read that you can compute the VIF score by this formula VIF= 1/(1-R^2) and you should use the R^2 that you get from regressing your independent variables on each other.

Is this correct?

When regressing my independent variables on one another, i for some reason get a negative R^2, this should be impossible. Could it be that this happens because i haven't selected fixed or random effects?

Manny thanks

Re: VIF scores panel data

Posted: Wed Sep 17, 2014 9:12 am
by EViews Gareth
It is true that VIFs aren't available for panel estimation in EViews.

The other questions are of an econometric nature, so I'll let someone else answer.