[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
st: Endogenous variables in Survival analysis
I am trying to estimate a survival duration model for 'firm'-level
data. The issue is how to deal with potentially endogenous regressors
(like leader's education level) in a pseudo-likelihood model like the
one estimated by -stcox-.
Until now, I was using a first stage linear predictor of the
endogenous regressor (i have some instruments) in my estimation
equation but someone recently pointed out that this 2SLS-type
procedure may not be entirely valid in the case of non-linear models.
So my question is, is there a way to deal with endogenous regressors
(like-ivprobit-) in Stata Survival suite?
* For searches and help try: