hi, I had a problem with the variance when I run IVPROBIT regression. I have a number of binary variables (whether the respondent belongs to a particular country or not). The PROBIT method results in a very large and irrational variance. There is no complete multicollinearity and in addition when running the same regression in the OLS method, the resulting variance is small and logical. From this I conclude that the problem is in the way of the PROBIT calculation, but do not know how to solve the problem.