GSEM probit estimate - where can I get rho?

[Juan Jung]

Dear Stata Forum: I am estimating a 3-probit equation model through GSEM. However, I want to be sure if it is correct to estimate this as a structural model, rather than 3-separated probits.

How can I know this? I think that I need the rho estimate as reported in Biprobit (and the Wald test of rho=0), to find our if the residuals are correlated.

Can anybody help me if there is a postestimate that allows me to do this after gsem?

Many thanks

Juan

[Joseph Coveney]

Fitting bivariate probit and bivariate ordered probit regression models is illustrated in a few posts on Statalist over the past several years, for example, here, here, here (in the context of panel data and here (ordered probit).

A Wald test (H0: covariance is zero) is produced by -gsem- at the bottom of its regression output table..