Dear Kumiko,
Unfortunately, margin is not helpful if you estimate a selection model other than
heckman/heckprob. I think you should rely on mfx. Probably this involves writing
an ado which provides the predicted values you are interested in.
Tamas
> Thank you for your reply. Yes I know about the excellent program
> but it
> won't allow me to compute dy/dx when E(y|x) = E(y|y>0,x)*Pr(y>0|x)
> willit?
>
> K.
>
> > Kumiko,
> >
> > I'll steal a march on Tamus Bartus here by suggesting his excellent
> > -margin- package:
> >
> > -net search margin-
> >
> > That will do the job for you much more quickly than -mfx compute-,
> > especially if your dataset is obscenely huge!
> >
> > C.
> >
> >> Hello,
> >>
> >> If anyone is aware of any do/ado file that allows me to obtain
> marginal>> effects after user-defined multiple equation estimation
> does -mfx
> >> compute,
> >> I would appreciate your help. I am aware of -mfx compute, predict
> >> (yexpected)- but this seems to work only after -heckman-. I
> have a
> >> two-part model with -logit- as 1st part and -xtgee- as 2nd part.
> >>
> >> Thank you for your help,
> >>
> >> Kumiko
> >> [email protected]
------------------------------------------------
Tamas Bartus, PhD
Assistant Professor, Department of Sociology and Social Policy
Budapest University of Economic Sciences and Public Administration
1093 Budapest, Fovam ter 8.
Phone: +36-1-217-51-72 Fax: +36-1-217-44-82
Homepage: http://www.bke.hu/bartus
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
