Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: Sensitivity when it comes to dropped variables due to collinearity?


From   CM <cmhsieh54@yahoo.com>
To   statalist@hsphsun2.harvard.edu
Subject   st: Sensitivity when it comes to dropped variables due to collinearity?
Date   Mon, 14 Jun 2004 20:47:32 -0700 (PDT)

Hi all,

This is a more basic question "following" up my
previous post.

I am running a Heckman specification.  At the outset,
it tells me it will drop variables due to
collinearity.  Instead of outputting "dropped" in the
line of the guilty variable, the line is either
entirely gone or filled with .'s  When I try to
regress the guilty variable on all the other variables
according to the help file
(http://www.stata.com/support/faqs/stat/drop.html), I
should get a "." for the test statistics of the
variables it's collinear with.  Instead, regressing
the guilty variable shows that it indeed is NOT
collinear to the others.

Why is Stata still telling me that it's collinear?  Is
there some kind of "sensitivity" option that I can
control?  I've searched -findit- and -search-, there
appears to be no ready answer.

Thanks,
CM
*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index