I'm trying to add a constraint on the cut points of a ordered probit model. The command for my unconstrained model is

oprobit y x

where $y \in \{0, 1, 2\}$. I'm having no problems fixing the second cut point with the commands

constraint 1 [cut2]_cons = 4
oprobit y x, constraints(1)

But what I really want to do is let the second cut point vary by individual and be determined by my variable L2_individual. My guess would be to write

constraint 2 [cut2] = L2_individual
oprobit y x, constraints(2)

but, while neither of those commands produce any errors, the resulting regression output is identical to that of my unconstrained model, which feels fishy.

Do you have any suggestion?


That syntax won't work, and you will have received a warning that that constrained was ignored after your oprobit command.

Instead you can add L2_individual as an explanatory variable, constrain the coefficient for that variable to be -1 and constrain the second cut value to be 0.

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.