Hello everyone,
As the title mention, I have an endogenous explanatory variable, which happens to be binary. But my main model is not linear; it is a fractional response model.
So, I'm not sure if the following approach could be adequate:

Code:
probit Y2 z1 z2 z3
predict my_e, resid

glm Y1 Y2 my_e x1 x2 x3, link(probit) family(binomial)
Is this correct? Thank you for your advice.