meologit for STATA 12

[Sridhar Telidevara]

I understand that MEOLOGIT is not part of STATA 12. Can someone please help me how to estimate mixed effects ordered logistic regression using STATA 12. Thank you in advance.

Best Regards,

Sridhar

[Clyde Schechter]

At version 12 you have two options. Sophia Rabe-Hesketh and Andres Skrondal's -gllamm- (available from SSC) can do this. And David Roodman's -cmp- (available from Stata Journal) can estimate the closely related mixed-effects ordered probit regression (but not ordered logit as far as I know).

-cmp- will not give you exactly what you are looking for here, but its syntax is closer to that of the native Stata commands for mixed effects models. And qualitatively the conclusions drawn from an ordered probit regression are usually the same as you get from the corresponding ordered logit (though the magnitudes of the coefficients will be different.)

-gllamm-'s syntax is somewhat "arcane," but it does cover ordered logit.

[Sridhar Telidevara]

Highly appreciate your helpful inputs. Thank you.