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

Re: st: Sample selection variable for bivariate probit selection

From   Maarten buis <[email protected]>
To   [email protected]
Subject   Re: st: Sample selection variable for bivariate probit selection
Date   Fri, 25 Jan 2008 12:16:22 +0000 (GMT)

--- Carlos Gamero <[email protected]> wrote:
> In your previous messages you advice me to use heckprob command in
> order to manage the situation I describe. However, heckprob command
> is designed to correct sample selection bias when both, main
> regression (job satisfaction) and selection equation, are probit
> regressions. But, what happens when you do not have a probit as
> selection equation but a bivariate probit as it is in my case?

I read your problem as follows: You are substantively interested in
work satisfaction (satisfied/not satisfied) and you think you have two
selection equations: 1) do you work or not, and 2) are you in education
or not.

If that is the case, than you have just one selection equation: The
selection is that you cannot be satisfied of dissatisfied with your job
if you don't have a job. In education or out of education is not a
selection equation because you can still be satisfied with a job if you
are in education as long as you also have a job. You are less likely to
have a job when you are in education, but that just means that being in
education should be an explanatory variable in your selection equation.

Hope this helps,

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

Sent from Yahoo! Mail - a smarter inbox

*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index