| Content-Type: |
text/plain; charset="iso-8859-1" |
| Date: |
Thu, 22 May 2014 09:23:01 +0200 |
| Reply-To: |
|
| Subject: |
|
| Content-Transfer-Encoding: |
quoted-printable |
| Message-ID: |
|
| MIME-Version: |
1.0 |
| Sender: |
|
| From: |
|
| Parts/Attachments: |
|
|
My name is Daniel Suryadarma, I am the senior scientist in charge of impact assessment at the Center for International Forestry Research (CIFOR).
I am replying to the question by Amadou Binta Ba about impact evaluation methods (Message 13). Basically, identification of propensity score matching (PSM) requires two assumptions:
First, the unconfoundedness assumption. That is: after controlling for all observable characteristics, no other characteristics - observed or unobserved - influence both participation in the program and the outcomes on which the impact is being evaluated.
Second, common support. That is: for all observable characteristics, we can observe participants and non-participants.
Looking at the two assumptions above, the first one is the most difficult to accept. For this reason, most quantitative impact evaluations no longer use matching. There are other evaluation methods working on weaker identifying assumptions, and matching should only be used as a last resort.
Dr. Daniel Suryadarma
Senior Scientist - Impact Assessment
Center for International Forestry Research
P.O. Box 0113 BOCBD
Bogor 16000
Indonesia
www.cifor.org
e-mail: d.suryadarma (at) cgiar.org
www.danielsuryadarma.com
twitter.com/dsuryadarma
[To contribute to this conference, send your message to [log in to unmask] For further information, see http://www.fao.org/nr/research-extension-systems/res-home/news/detail/en/c/217706/ ].
########################################################################
To unsubscribe from the Impact-L list, click the following link:
https://listserv.fao.org/cgi-bin/wa?SUBED1=Impact-L&A=1
|
|
|