From what i understand, conventional OLS heteroskedasticity tests can't be applied for Tobit regressions. I have looked at the help "tobit post-estimation" command and i still am unable to find heteroskedasticity tests for tobit.

From reading woldbridge (2003) i have seen that conducting a probit model for my data can be used as an informal test for heteroskedasticity, am i correct in thinking this?