Hi everyone,
Within the multistate life table framework, I am trying to compute transition probabilities among my states of interest (healthy, unhealthy, dead). Given the scale of my outcome variables (i.e health and mortality status), I used the multivariate probit model (i.e., biprobit command) to get the hazard coefficients. I wonder to know how can I get the transition probabilities between states of interest for specific covariates using hazard coefficients. Does post-estimation command such as "predict" or "margin" do it?
Thanks,
Nader