Subject: | |
From: | |
Reply To: | |
Date: | Sun, 30 Nov 2008 16:51:05 -0500 |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
The best tests for this are those suggested by Rosenbaum in his reply
to the Angrist, Imbens and Rubin 1996 JASA. He discusses this
further in his statistical science paper in 2002 and in JRSS A in 2005
with Guido Imbens and in more recent JASA papers.
He suggest using a nonparametric sensitivity test. This allows you to
quantify how sensitive your estimate is to the exclusion restriction.
The nice thing about these tests is they provide correct coverage on
the estimate by causing the confidence interval on the IV estimate to
slide out possibly to infinity which can happen when the instrument is
weak.
Luke Keele
Assistant Professor
Dept. of Political Science
Ohio State University
On Nov 30, 2008, at 4:03 PM, Carola Herrera wrote:
> Dear list,
>
> I need to run a test to find out how sensitive my estimates are to
> violations of the exclusion restriction of the instrumental variable
> I am using. Could anyone let me know what the code is in STATA10?
> Or perhaps suggest an article that walks one thru the steps?
>
> Thank you.
> Carola
>
>
>
>
>
>
> **********************************************************
> Political Methodology E-Mail List
> Editors: Melanie Goodrich, <[log in to unmask]>
> Delia Bailey, <[log in to unmask]>
> **********************************************************
> Send messages to [log in to unmask]
> To join the list, cancel your subscription, or modify
> your subscription settings visit:
>
> http://polmeth.wustl.edu/polmeth.php
>
> **********************************************************
**********************************************************
Political Methodology E-Mail List
Editors: Melanie Goodrich, <[log in to unmask]>
Delia Bailey, <[log in to unmask]>
**********************************************************
Send messages to [log in to unmask]
To join the list, cancel your subscription, or modify
your subscription settings visit:
http://polmeth.wustl.edu/polmeth.php
**********************************************************
|
|
|