# st: Re: multivariate count regression

 From "Scott Merryman" To Subject st: Re: multivariate count regression Date Thu, 14 Aug 2003 21:04:14 -0500

```----- Original Message -----
From: "Ralf Dewenter" <ralf.dewenter@unibw-hamburg.de>
To: <statalist@hsphsun2.harvard.edu>
Sent: Thursday, August 14, 2003 12:44 PM
Subject: st: multivariate count regression

> Hi all,
>
> Does anybody know, where to find an ado-file
> for multivariate count regressions?
> Thanks.
>
> Regards
>
> Ralf Dewenter
>

Would -poisson- or -nbreg- with -suest- be sufficient?  I believe it is similar
to -mvreg- in that it estimates the between-equation covariance so you can test
coefficients across equations.

Example:

. clear

. set obs 100
obs was 0, now 100

. gen x1 = invnorm(uniform())

. gen x2 = invnorm(uniform())

. gen xb = 1 + 2*x1 - x2

. genpoisson y, xbeta(xb)

. genpoisson y2, xbeta(xb)

. qui poisson y x1 x2,score(s1)

. est store A

. qui poisson y2 x1 x2, score(s2)

. est store B

. suest A B

Simultaneous results for A, B
Obs      =     100

------------------------------------------------------------------------------
|               Robust
|      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
A_y          |
x1 |   1.432079   .1199582    11.94   0.000     1.196966    1.667193
x2 |  -.7417589   .0804557    -9.22   0.000    -.8994491   -.5840686
_cons |    1.45254   .0961629    15.10   0.000     1.264064    1.641015
-------------+----------------------------------------------------------------
B_y2         |
x1 |   1.418806   .1223085    11.60   0.000     1.179085    1.658526
x2 |  -.7914187   .0901748    -8.78   0.000     -.968158   -.6146794
_cons |   1.459112    .097084    15.03   0.000      1.26883    1.649393
------------------------------------------------------------------------------

. test [A_y]x2 = [B_y2]x2, common

( 1)  [A_y]x2 - [B_y2]x2 = 0

chi2(  1) =    1.48
Prob > chi2 =    0.2236

Scott

*
*   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/
```