st: Re: use mfx to compute marginal effects in tobit

["Zou Hong" <hzou@ln.edu.hk>]

Dear All,

Further to my previous query, 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. But I found it does 
function following a probit model (i.e., generating the same marginal 
effects as that is generated from "dprobit").

I enclose the model output using two variables for your information.

Thanks
Hong  Zou
----- Original Message ----- 
From: "Zou Hong" <hzou@ln.edu.hk>
To: <statalist@hsphsun2.harvard.edu>
Sent: Wednesday, August 08, 2007 9:35 AM
Subject: st: Re:tobit model robust standard errors and marginal effects



> Dear All,
>
> I want to report the results from estimating a Tobit model (left censored 
> at zero) using robust standard errors and marginal effects (which by 
> default are evaluated at the sample mean of variables).
>
> I have done a search and found that somebody suggesting using "intreg" to 
> get the robust standard errors and p-values. For marginal effects, 
> somebody suggested using dtobit, dtobit2 and/or mfx. However, it seems 
> these routines do not allow for the use of robust standard errors.
>
> I wonder how I can report a Tobit model using robust standard errors and 
> marginal effects at the same time? Can I use "intreg" and request "robust" 
> option, followed by "mfx"?
>
> Many thanks
> Hong Zou
>
> ----- Original Message ----- 
> From: "Nick Cox" <n.j.cox@durham.ac.uk>
> To: <statalist@hsphsun2.harvard.edu>
> Sent: Wednesday, August 08, 2007 4:27 AM
> Subject: RE: st: RE: mlogit problem with "predict"
>
>
>
> > -stripplot- from SSC will let you
> > compare variables in groups.
> >
> > stripplot pc1-pc4, by(tax)
> >
> > gives you one graph, not four. See
> > the help for further options.
> >
> > Nick
> > n.j.cox@durham.ac.uk
> >
> > Maarten buis
> >
> > > --- Sheela Athreya <athreya@neo.tamu.edu> wrote:
> > > > The idea that maybe I have an (almost) perfect prediction occurred
> > > > to me too, but I don't know how to investigate it. Perhaps there is
> > > > an equivalent test to -tab tax pc- that allows me to look at
> > > > quantitative variables for anything that is off?
> > >
> > > One option is to look at dotplots:
> > > dotplot pc1, over(tax) median
> > > dotplot pc2, over(tax) median
> > > dotplot pc3, over(tax) median
> > > dotplot pc4, over(tax) median
> > >
> > > They give you simultaneously an idea about the strength of the
> > > association and the number of observations in each category.
> > *
> > *   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/

Attachment: tobit marginal effects output.pdf
Description: Adobe PDF document