Dear Statalist users,
Has somebody tried to program the Maximum Likelihood for an ordered probit
or logit?
I'm having problems in programming it. Any help would be really appreciated.
Thanks
Josefina
As Nick notes, it has already been done with ologit and
oprobit. But, if you have some other reason for wanting to see the
coding, check out -oglm- from SSC. oglm will estimate ordered logit
and probit and 3 other links, and also makes it possible to control
for heteroskedasticity. For more info, see