Hello,

I am dealing with an intreg model, which does not meet the assumption of homoskedasticity. I ran a likelihood ratio test comparing the orginal model with an exponential model of heteroskedasticity. Therefore I added the het() option and included all independent variables in the het part. The likelihood ratio test confirms heteroscedasticity.
Now I wonder, whether I have to include all independent variables in the het option.
I noticed that if I leave out some of those variables which are not significant in the part below Insigma (in the model with the het option), it sometimes has an impact on whether my variable of interest is significant or not.
>Do I have to to include all independent variables in the het() option?
> If not, how can I determine which variables I have to add to the het option?

Many thanks in advance!

Kind regards
Anna