Hi everyone,
I am using a biprobit model with two dichotomized outcome variables (health and mortality). I know that pcond1 generates predicted probability for depvar2=1 given depvar1=1 (i.e., Pr(depvar2=1 | depvar1=1).
I wonder if there is any post-estimation that allows me to include more conditions. Specifically, I am looking for something like this:
pr (depvar1=1 | X=1 & depvar2=0), where X is an independent variable used in the model.
Best,
Nader