Dear all,

How can I compute the marginal effects for ordered logit/probit models based on the output provided by gretl? Anything similar to Stata's "margins" command ?

Hi Artur,

Allin's function package "lp-mfx" should do the trick, I guess.

Let me also mention the package "a_eff" by Oleh (Komashko), right now I don't know the difference between these packages with respect to your context.

(BTW, I guess at least the help text of a_eff can be updated since it starts with warnings about a transition period after gretl 2018c which should be obsolete by now. But that's a different issue.)

cheers

sven