# st: Re: Simulating Instrumental Variable Probit Model

 From Sachin Chintawar To statalist@hsphsun2.harvard.edu Subject st: Re: Simulating Instrumental Variable Probit Model Date Tue, 21 Apr 2009 11:41:31 -0600

```Thank you (Dr) Martin
I think that was the problem. I had totally missed the -ivprobit- syntax
Regards
Sachin
PS: Considering you may be a Doctoral Candidate I included the
brackets around the Dr.

On Tue, Apr 21, 2009 at 11:08 AM, Sachin Chintawar
<schint1@tigers.lsu.edu> wrote:
> Dear Statalist Users
> Well adding to my earlier program I did find a few silly mistakes that
> I have cleaned up. Find a newer version of the program with more
> doubts.
> --------------------------------------------------Start
> Example---------------------------------------------------------------------
> global numobs 500             // sample size N
> global numsims "1000"         // number of simulations
> set seed 123456789
>
> capture program drop endoprob
>
> program endoprob, rclass
>        version 10.1
>        drop _all
> set obs \$numobs
>    generate u = rnormal(0)
>    generate mu = rnormal(0)
>    generate x1 = rnormal(0)
>    generate z1 = rnormal(0)                      // 4 Instruements
>    generate z2 = rnormal(0)
>    generate z3 = rnormal(0)
>    generate z4 = rnormal(0)
>                                                // Also be written *drawnorm u mu x1 z1 z2 z3 z4
>    generate a = 0.5*u
>    generate x2 = z1 + z2 + z3 + z4 + a
>    regress x2  z1 z2 z3 z4 mu                // endogenous regressor
> with four instruments
>    predict px2
>
>    generate y = 0.5 + 2*x1 + px2 + mu + u>4    //Reduced Form Equation
>    ivprobit y x1 px2
>    return scalar b2 = _b[x1]
>    return scalar b3 = _b[px2]
>    return scalar se2 = _se[x1]
>    return scalar se3 = _b[px2]
> end
> simulate b2=r(b2) b3=r(b3) se2r=r(se2) se3r=r(se3), ///
>    reps(\$numsims): endoprob
> mean b2 b3 se2r se3r
> -------------------------------------------End
> Example---------------------------------------------------------------------------------
> 1. Adding to my earlier problem if we run this program an error occurs
> suggesting that there is no endogenous variable  and that I use
> -probit- . But this totally negates the question I want to answer -
> the Amount of bias as we increase the number of instruments?
> Any suggestions would be appreciated
>
>
> I
>

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
```