Page 1 of 1

McFadden R-Squared with categorical dummies

Posted: Mon Sep 10, 2012 4:51 am
by RoaStudent1984
I have an estimation: Probit with continuous and categorical dummy variables.

If I add the dummy variables to the estimation I have to leave the constant out, otherwise the output gives me NA everywhere. (Right?)
My problem is that I get no McFadden R-squared if I leave the constant out of the estimation. Is there another way to do the estimation correctly, and get a McFadden R-squared?

Thanks!

Re: McFadden R-Squared with categorical dummies

Posted: Mon Sep 10, 2012 6:46 am
by EViews Gareth
Leave the constant in, but drop a dummy.