What about this?
*---------- begin example -----------------------
sysuse auto, clear
recode rep78 1/2=3
oprobit rep78 foreign
mfx, predict(pr outcome(#1))
matrix list e(Xmfx_dydx)
matrix b = e(Xmfx_dydx)
matrix se = e(Xmfx_se_dydx)
mfx, predict(pr outcome(#2))
matrix b = b \ e(Xmfx_dydx)
matrix se = se \ e(Xmfx_se_dydx)
mfx, predict(pr outcome(#3))
matrix b = b \ e(Xmfx_dydx)
matrix se = se \ e(Xmfx_se_dydx)
matrix bci = b - invnormal(.975)*se, b, b + invnormal(.975)*se
matrix colnames bci = lb b ub
matrix rownames bci = outcome1 outcome2 outcome3
matrix list bci
drop _all
svmat bci, names(col)
gen outcome = _n
twoway rcap lb ub outcome || ///
scatter b outcome , ///
xlab(1 2 3) xscale(range(.5 3.5)) ///
xtitle("outcome") yline(0) ///
ytitle("Pr(outcome|foreign)-Pr(outcome|domestic)") ///
legend( order( 2 "point" "estimate" ///
1 "95%" "Conf. Int.") )
exit
*-------------------- end example --------------------------------
(For more on how to use examples I sent to the Statalist, see
http://home.fsw.vu.nl/m.buis/stata/exampleFAQ.html )
--- Tom Boonen <[email protected]> wrote:
> I have a binary X covariate and a discrete (five category) outcome
> variable. All i need is a way to store these predcited changes (and
> most importantly the CIs on the predicted changes). Any help on how
> to
> do this is highly appreciated. Spost commands do not seem to provide
> this functionality.
>
> Thanks,
> Tom
>
> On Jan 11, 2008 10:01 AM, Maarten buis <[email protected]>
> wrote:
> > --- Tom Boonen <[email protected]> wrote:
> > > For an ordered probit or logit model, is there any way to save
> > > marginal effect estimates produced via the suite of spost
> commands?
> > > For example:
> > >
> > > oprobit Y X1
> > > prvalue, x(X1=0) save delta
> > > praccum , using(Xmat) xis(0)
> > > prvalue, x(X1=1) dif delta
> > >
> > > * this produces nice marginal effect estimates (ie. changes in
> the
> > > PR(Y=1|X1=0) - PR(Y=1|X1=1)) including CI intervals.
> >
> > These are usually called discrete change coefficients, and not
> marginal
> > effects. Marginal effects is the first derivative of the
> probability
> > curve with respect to x1 evaluated at some value chosen by you.
> > Discrete change coefficients are the differences between two
> predicted
> > probabilities.
> >
> > What kind of graph do you want to create (what is on y-axis and
> > x-axis?)
> > What kind of explanatory variables do you have (continuous or
> > categorical)?
> >
> > -- Maarten
> >
> > -----------------------------------------
> > Maarten L. Buis
> > Department of Social Research Methodology
> > Vrije Universiteit Amsterdam
> > Boelelaan 1081
> > 1081 HV Amsterdam
> > The Netherlands
> >
> > visiting address:
> > Buitenveldertselaan 3 (Metropolitan), room Z434
> >
> > +31 20 5986715
> >
> > http://home.fsw.vu.nl/m.buis/
> > -----------------------------------------
> >
> >
> > __________________________________________________________
> > Sent from Yahoo! Mail - a smarter inbox http://uk.mail.yahoo.com
> >
> >
> > *
> > * 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/
> >
> *
> * 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/
>
-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands
visiting address:
Buitenveldertselaan 3 (Metropolitan), room Z434
+31 20 5986715
http://home.fsw.vu.nl/m.buis/
-----------------------------------------
__________________________________________________________
Sent from Yahoo! Mail - a smarter inbox http://uk.mail.yahoo.com
*
* 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/
