How to get standardized coefficients in generalized ordered logit model?

[Qun Wang]

Dear experts,

I am doing a generalized ordered logit model. I use gologit2. My command is “gologit2 opinionnew gender i.ageg i.edug i.perinc4 medinsur i.healthnew i.level, autofit lrforce gamma or”。
I need to get standardized coefficients from the generalized ordered logit model. But I google for some time and got no answer. Does any expert know how to do it? Thank you so much!

[Phil Bromiley]

If you mean the parameter divided by its standard error, you can always calculate this using the output from the estimator. In regression, it is also transforming the dependent variable to have a standard deviation of one but I don't see how this would make any sense in an ordered logit. After most estimations, issuing return list and ereturn list will tell you what matrices and macros that estimation has saved. You also can generally access the parameter for xwith _b[x] and _se[x] although whether these work with that user written procedure is not clear to me.

I don't know if this works for nonlinear estimators, but you can also simply standardize the variables using egen before the estimate with most linear estimators.