# st: problems with marginal effects of biprobit

 From "Natalia Ferreira" To statalist@hsphsun2.harvard.edu Subject st: problems with marginal effects of biprobit Date Fri, 25 Feb 2005 04:12:22 +0000

```Dear all

I´m estimating a biprobit to my child labor problem:
1. Child just study
2. Child study and work
3. Child just work
4. Child neither study or work

So the first probit is regarding to Study: 1 if child study and zero
otherwise. The second probit refers to work probability: 1 if child work and
zero if doesn´t.
My problem occours when I want to get the marginal effects of the two
equation and all the 4 output probabilities (p11, p10, p01, p00). When I
type the command "mfx" after the estimation I get the marginal effects but
just for the default probability (p11). As show the table below.
I already tried "mfx, predict p10", and all kinds of possible combination,
but it didi not work....
Does some one knows how  can I get the same marginal effects and estimate
probability for all possible combinations, and not just for p11?

Thanks for the help, natalia

Marginal effects after biprobit
y  = Pr(bi_prob1=1,bi_prob2=1) (predict)
=  .01582596

variable       dy/dx    Std. Err.     z    P>z	[    95% C.I.   ]	     X

id_f          .0112753      .00026   43.61   0.000    .010769  .011782
11.9961
f_cor*       .0003155      .00076    0.42   0.677   -.001167  .001798
.290431
novo         .0034677      .00031   11.22   0.000   .002862  .004074
.904704
velho         .0015147      .00051    3.00   0.003   .000525  .002505
.549959
edu_r       -.0012177      .00011  -10.91   0.000   -.001437 -.000999
6.91298
edu_c       -.0010851      .00012   -9.37   0.000   -.001312 -.000858	6
.73947
migint_r*   .0062618      .00259    2.42   0.016     .00119  .011333
.079524
migint_c*  -.0033373       .0017   -1.96   0.050     -.00667 -4.5e-06
.08513
nrpc_sm    .0030452      .00017   18.18   0.000     .002717  .003374
5.18521

(*) dy/dx is for discrete change of dummy variable	from 0 to 1

_________________________________________________________________
MSN Messenger: instale grátis e converse com seus amigos.
http://messenger.msn.com.br

*
*   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/
```