--- Zou Hong wrote:
> I estimate a tobit model and then use "mfx" to compute
> marginal effects of variables. Strangely, the computed
> marginal effects for all the variables are the same as
> the estimated cofficient in the probit model. It
> seems "mfx" does not function.
There are a variety of marginal effects possible and of
interest after -tobit-:
o the effect on the expected value of the latent variable
o the effect on the expected value of the dependent,
variable conditional on it being larger than the lower
bound, and
o the effect on the probability of being larger than the
lower bound.
For more on this see: (Breen 1996).
By default Stata chooses the effect on the latent variable,
which are exactly the same as the coefficients estimated by
-tobit-. You will have to specify the -predict()- option in
-mfx- to get the other marginal effects. See the example
below, and -help mfx- and -help tobit postestimation-
*--------- begin example ----------
sysuse auto, clear
tobit mpg pric wei, ll(15)
mfx
mfx, pred(e(15,.))
mfx, pred(pr(15,.))
*----------- 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 )
Hope this helps,
Maarten
Richard Breen (1996) "Regression Models: Censored, Sample
Selected, or Truncated Data". Thousand Oaks: Sage.
-----------------------------------------
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/
-----------------------------------------
*
* 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/
