GOLOGIT command for panel data (XTOLOGIT)?

[Gurpreet Singh]

Hello everyone,

I am working with ordinal outcome model with specification. Hence, I am using XTOLOGIT command for my work.

However, as is the case, ordered logit models often tend to violate parallel regression assumption. Would that be the case with XTOLOGIT model (i.e. ordered logit model in panel set up) as well?

If yes, what are the options? Is there a gologit command analogous to panel ordered regression?

Thanks for your help!

[Joseph Coveney]

Is there a gologit command analogous to panel ordered regression?

Not to my knowledge, but you could probably set something up using menl.

Have you considered the vce(cluster clustvar) option of the user-written gologit2 available from SSC? It might suffice for your purposes.

[Richard Williams]

For some ideas, see the section in the gologit2 troubleshooting FAQ entitled "Can I do a random effects or multilevel model with gologit2?".

http://www3.nd.edu/~rwilliam/gologit2/tsfaq.html