Hello,
I have discovered in R that there is a way to allow for conditional heteroskedasticity when estimating a Tobit regression (https://cran.r-project.org/web/packages/crch/index.html). This can be used to limit misspecification in the likelihood function leading to less misspecification. Post-estimation inference can also be performed, such as clustering with the sandwich estimator.
Is it possible to do something similar using STATA?
Thanks!
I have discovered in R that there is a way to allow for conditional heteroskedasticity when estimating a Tobit regression (https://cran.r-project.org/web/packages/crch/index.html). This can be used to limit misspecification in the likelihood function leading to less misspecification. Post-estimation inference can also be performed, such as clustering with the sandwich estimator.
Is it possible to do something similar using STATA?
Thanks!
Comment