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

# st: xtivreg2 - Incomplete computation of first stage regression

 From "Lim Boon Leong" To Subject st: xtivreg2 - Incomplete computation of first stage regression Date Sun, 28 Feb 2010 16:44:30 +0800

```Dear Statalisters:

I have a regression equation which is consisted of 5 suspected endogenous
variables with a list of excluded instrument variables.  I use xtivreg2,
fixed effect, cluster (id) small function to run the regression analysis.

Xtivreg2 lsalary ceodual ... (ownexr ownex2r ownex3r L.ldiv L.debt = lrisk
lsale lsale2......), fe cluster(id) small

The regression runs smoothly and provides test results such as
underidentification test, weak identification test, and Hansen J statistic.

But when I run the first function to get the first stage regression as
below:

Xtivreg2 lsalary ceodual ... (ownexr ownex2r ownex3r L.ldiv L.debt = lrisk
lsale lsale2......), fe cluster(id) small first

, it only managed to provide first stage regression results for the first 3
endogenous variables namely ownexr, ownex2r, and ownex3r.

And for the fourth first stage regression result it is reported as such:

First-stage regression of _000007:

..........................

r(111) which implies missing variables.

My questions are as follows:

1.        If I have 5 endogenous regressors, can I expect to have 5 first
stage regression results for all these endogenous variables?

2.       Is there any methods to compute the first stage regression for all
these 5 endogenous regressors?

Thank you very much.

Best regards,

Lim Boon Leong

*
*   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/```