0
$\begingroup$

I ran a 2SLS regression in STATA, and there is no R squared value given in the output (only a dot where the value should be). What could this be interpreted as? Is it missing because it may be a negative value? And does this imply the output is not to be trusted?

$\endgroup$
0
$\begingroup$

The $R^2$ has no meaning with 2SLS/IV regression since the residual sum of squares is not restricted to be smaller than the total sum of squares. So to answer your question: Yes, it might be negative. Stata supresses printing a negative $R^2$ when you use the ivreg command.

For a detailed discussion of this question, see the detailed answer in Stata's FAQs which can be found here.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.