EViews.com

Hey,

I was wondering if i can still use the reported R² as the goodness of fit in a TSLS estimation. It seems to me that you can't say that the model will explain a certain negative % of the real data. Doesn't their exist an alternative value or test we can use to determine which model fits best?, when estimating TSLS.

Thanks,

Annelies

Do I have to use the MacKinnon J test to determine which value fits the best? 'cause it looks like the J statistic is the only one who is reported after an TSLS and not after an OLS regression...
I would be relieved if someone could help me out here.
Thanks a lot!
Annelies