Statalist


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

RE: st: stata code for two-part model


From   "Shehzad Ali" <[email protected]>
To   <[email protected]>
Subject   RE: st: stata code for two-part model
Date   Sat, 16 Aug 2008 07:10:56 +0100

Thanks, Martin, but as I understand the two-part model that I am trying to
estimate is little different from Heckman model because Heckman includes an
Inverse Mills Ratio on the right hand side in the second part of the model
whereas two-part model does not. That's what I understood from the following
article:

Dow & Norton (2003). Choosing between and interpreting the Heckit and
two-part models for corner solutions. Health Services and Outcomes research
methodology. 4: 5-18. Page 10

Please feel free to correct me if I am wrong.

Regards,

Shehzad


-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Maarten buis
Sent: 16 August 2008 06:35
To: [email protected]
Subject: Re: st: stata code for two-part model

--- Shehzad Ali <[email protected]> wrote:
> I was wondering if someone can help with stata code for calculating
> marginal effects after two-part models for say, cost of care. Here,
> first part is a probit model for seeking care or not, and the second
> part is an OLS model of cost of care, conditional on decision to seek
> care. 

Looks to me you are trying to reinvent -heckman-. If you use -heckman-
you will both do a better job at modeling the process, and you will
have many -predict- option which you can feed into -mfx-, so you can
look at the various types of marginal effects that are possible in this
type op models. 

> Here is the simplified code:
> 
> probit care $xvar
> 
> reg cost $zvar if care==1
> 
> mfx
> 
> I understand that mfx after the second part gives us the marginal
> effects for the OLS part only, and not the conditional marginal
> effects.

The OLS part is already conditional, the problem here is that this is
not a two-part model as you are not using anything from the -probit-
model, in other words you could just as well leave that -probit- away
and your results will not change. I am not sugesting that you do that
though, you have selection on the y and you will have to model it. It
is just that the model you are proposing just isn't the way to do it.
Use -heckman- instead. 

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/
-----------------------------------------

Send instant messages to your online friends http://uk.messenger.yahoo.com 
*
*   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/

Internal Virus Database is out of date.
Checked by AVG - http://www.avg.com 
Version: 8.0.138 / Virus Database: 270.4.6/1540 - Release Date: 08/07/2008
06:33

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