Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: Heckprobit or Heckman Selection model (ML)?


From   Maarten buis <[email protected]>
To   [email protected]
Subject   Re: st: Heckprobit or Heckman Selection model (ML)?
Date   Wed, 20 Dec 2006 15:29:42 +0000 (GMT)

--- GJIPALI Arsena <[email protected]> wrote:
> Doing my estimations on labour market returns to education, I am
> facing two concerns.
> 
> 1) I have a very low value of R square (18%) for the earnings
> equation with OLS technique.

I think 18% is very acceptable. It does mean that a lot of variation
remains unexplained, but look around you and you'll see a lot variation
in wages that has nothing to do with any of your explanatory variables.
As I remarked earlier, explaining human behavior is a tough job.
 
> 2) I want to estimate the probability of being in employment,
> accounting for participation in the labour force. I am applying the
> Heckman selection model (ML) wchich provides the inverse Mill's Ratio
> calculation for the selection bias of participation in the labour
> force. STATA also provides estimation of HeckProbit. I want to be
> assured that the former model is the one appropriate for me,
> given the censoring because of the effect of education on the labour
> force participation and than in employment likelihood.

I don't know about STATA, but Stata doesn't have a command
-HeckProbit-. (see:
http://www.stata.com/support/faqs/res/statalist.html#spell ) Stata does
have a command -heckprob-, and this seems to me the command you need.
The difference between -heckman- and -heckprob- is that the dependent
variable in the equation of interest is continuous for -heckman- and
dichotomous for -heckprob-. 

>      Similar to this, I also estimate the probability of being
> enrolled in tertiary education of the group age 19-22. However, these
> are to be selected from the pool of young that have completed upper
> secondary. Hence, I am in a dilemma, whether I should choose the
> Heckman selection model (ML) or the HeckProbit.

-heckprob- for the same reason as above.

> What would be the commands for the partial effects in these models?
> And what about any diagnostic tests?

-help mfx- and -help heckman postestimation-

Hope this helps,
Maarten

-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands

visiting address:
Buitenveldertselaan 3 (Metropolitan), room Z434

+31 20 5986715

http://home.fsw.vu.nl/m.buis/
-----------------------------------------


		
___________________________________________________________ 
Try the all-new Yahoo! Mail. "The New Version is radically easier to use" � The Wall Street Journal 
http://uk.docs.yahoo.com/nowyoucan.html
*
*   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 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index