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

st: Gauss Newton Regression and Testing for Omitted Time-Invariant Regional Effects in a Binary Response Model


From   "David Richmond" <drichmon@wfubmc.edu>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: Gauss Newton Regression and Testing for Omitted Time-Invariant Regional Effects in a Binary Response Model
Date   Wed, 5 May 2004 10:50:33 -0400

Has anyone used Gauss-Newton regression method ala Davidson and
MacKinnon for a logistic regression in Stata? If so, do you have any
code?

This is the code I am using:

logit y x
predict yhat
gen multiplier = yhat*(1-yhat)
gen e = y - yhat
gen weight = 1/sqrt(multiplier)
gen w_e = weight * e
gen w_x = weight * x
gen w_z = weight * z
regress w_e w_x w_z
test w_z

If the coefficient on w_z is not significantly different from 0, then I
can omit it from the regression.

David Richmond

*
*   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