Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.

# st: Test of Exogeneity in STATA

 From nil sen <[email protected]> To [email protected] Subject st: Test of Exogeneity in STATA Date Fri, 1 Oct 2010 19:50:15 -0500

```Hello,

My questions are preceded with a ***** sign.

*****How to conduct test of exogeneity in a binary outcome model?

I have two equations,

hval = a0 + a1*rval+a2*faminc + a3*reg2 + a4*reg3 + a5*reg4 + epsilon1 ----(1)
rval = b0 + b1*hval + b2*urban + epsilon2 ----(2)

where hval is continuous and rval is binary.

I think, I can test for exogeneity of rval in equation (1) as following:

. regress hval faminc reg2-reg4 urban
. predict h_res, res
. probit rval hval urban h_res
. test h_res

*****Please let me know, if this is correct.

But I do not know, how to test for exogeneity of hval in equation (2).
If I use the following,

.probit rval urban reg2-reg4 faminc ---(3)

then I do not know how to obtain the residual from (3) in STATA,
since, it is a probit model.

*****Could you please let me know, how to obtain the residual here.

I think, once I can obtain the residual in (3) as 'r_res', then I can
proceed as following,

. regress hval faminc reg2-reg4 r_res
. test r_res

*****Is this correct?

Regards,

Nil

*
*   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/
```