Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down at the end of May, and its replacement, statalist.org is already up and running.


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

Re: st: treatreg vs ivreg revisited


From   Austin Nichols <austinnichols@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: treatreg vs ivreg revisited
Date   Wed, 15 May 2013 07:37:58 -0400

Andrea Menclova <andrea.menclova@canterbury.ac.nz>:
You should never "rely on the inverse Mills ratio for addressing
endogeneity/selection" though if you have excluded instruments,
-treatreg- can get
estimates with lower variance if the maintained assumptions are
correct. If any of the extra assumptions required by -treatreg- fail,
your estimates can have substantially higher mean squared error with
-treatreg- relative to IV.

On Tue, May 14, 2013 at 12:22 AM, Andrea Menclova
<andrea.menclova@canterbury.ac.nz> wrote:
> Dear Statalist community,
>
> I have a “treatment” model with individual-level data (and a large number of
> observations) where the dependent variable is “birth weight” and the
> treatment variable is “delivery by a midwife” (versus delivery by an
> obstetrician).  Having read several previous posts on this listserv (with
> helpful responses from Mark Schaffer – thanks!), I still can’t decide
> whether to use treatreg or ivreg to estimate the effect of midwife delivery
> on infant health. So far, all comments I’ve received from colleagues and
> readers of my manuscript were focused on the strength/validity of my
> instruments but only recently someone suggested I could use treatreg and
> rely on the inverse Mills ratio for addressing endogeneity/selection. Could
> someone please give me advice about which is preferable?

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


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