Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: AW: question about Heckman and general Heckit models


From   "Martin Weiss" <martin.weiss1@gmx.de>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: AW: question about Heckman and general Heckit models
Date   Fri, 8 May 2009 17:29:58 +0200

<> 

As you can replicate the first stage easily, you have access to the full
post-estimation capabilities for -probit-. For comparability, it is better
to model via - heckman, twostep-, I think. 

Note that the option -first- to -heckman- shows you exactly this probit
equation, including the LR test against the naive model.


*************
webuse womenwk, clear

generate byte wageseen = (wage < .)
heckman wage educ age,  /// 
select(wageseen = married children educ age) /// 
two first


//first stage only 

probit wageseen married children educ age

//diagnostics
capt ssc inst fitstat
fitstat 

//graphs
lsens, name(lsens, replace) nodraw
lroc, name(lroc, replace) nodraw

gr combine lsens lroc, cols(1)

//classification table
estat classification
*************

The collinearity thing could easily be checked by capturing the mills ratio
with the -mills- option to -heckman- and then running the second stage by
hand.

HTH
Martin


-----Ursprüngliche Nachricht-----
Von: owner-statalist@hsphsun2.harvard.edu
[mailto:owner-statalist@hsphsun2.harvard.edu] Im Auftrag von Ed Levitas
Gesendet: Freitag, 8. Mai 2009 17:13
An: statalist@hsphsun2.harvard.edu
Betreff: st: question about Heckman and general Heckit models

Statalisters,

This is more of a general question about the Heckman Selection procedure
(aka Heckit) but also has obvious implications for the HECKMAN procedure
in Stata.

How can one assess the "suitability" of the first stage selection model?
That is, are there tests that indicate that the selection equation
does/does not effectively model selection?

Wooldridge (2002) among others notes the importance of testing for
collinearity between the inverse Mills' ratio and the explanatory
variables in the outcome equation.

Can one, for example, also perform a likelihood ratio test between the
first stage unrestricted probit, and a constant only probit?

Any suggestion would be greatly appreciated.

Thanks
Ed

****************************************
Edward Levitas, PhD
Associate Professor
Sheldon B. Lubar School of Business 
University of Wisconsin-Milwaukee
3202 N. Maryland Ave.
Milwaukee, WI  53211
ph: (414) 229-6825
fx: (414) 229-6957

*
*   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/


*
*   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/



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