RE: Re:st: RE: marginal effects interaction and squared variables ordered probit

[Nick Cox <n.j.cox@durham.ac.uk>]

If you upgrade (not -update-!), I think you can only upgrade to Stata 12. So, what you should do is look at the help for 

http://www.stata.com/help.cgi?oprobit_postestimation

and see what is supported by official Stata in the latest version. 

Nick 
n.j.cox@durham.ac.uk 

Vincenzo Carrieri

I am on stata 10. I understand your point that user-written are fairly
limited in comparison with official stata support, but I actually
don't know which is the command that does what I need.
Is there a way to estimate marginal effects in ordered probit models
with interaction and squared terms on stata 11 or 12? Eventually,
which is the name of this command?
I could think to update my version in such case.

Nick Cox 

Presumably you are on Stata 10 (why not say so?).

Official Stata support for this kind of thing has surged ahead since
then, and user-written add-ons are fairly limited by comparison.

- -inteff- is user-written (SJ).

Vincenzo Carrieri

I am estimating an ordered probit model with interaction terms  and
squared variables as covariates:

Y= age+age^2+age^2*income

Is there a way to get marginal effects and standard errors for all
variables (both single, interacted and squared variables)? I have
realized that inteff does not work with ordered probit and margin
seems to not work with ordered probit model, too. I have not stata 11,
so I cannot test directly margin, but stata help form margin does not
include oprobit in the examples.

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/