2 answers: L1 should be treated as a relative measure -- relative to the
variables in the model and the particular data set. so there isn't an
absolute level of L1 that is acceptable.
more interesting tho, with CEM, you don't need to worry about measures of
balance after the fact since you can control the maximum imbalance that
there can be by choosing a level of coarsening ex ante. So I'd focus your
efforts on choosing and justifying the level of coarsening. Check the
remaining n, but aside from that, everything else is automatic.
Gary
--
*Gary King* - Albert J. Weatherhead III University Professor - Director,
IQSS - Harvard University
GKing.Harvard.edu <http://gking.harvard.edu/> - King(a)Harvard.edu -
@kinggary<http://twitter.com/kinggary>- 617-500-7570 - Asst 495-9271 -
Fax 812-8581
On Sun, Mar 13, 2011 at 3:52 PM, Evann Smith <egsmith(a)fas.harvard.edu>wrote;wrote:
Hi,
I'm using CEM to match my data. What is a reasonable value of the L1
statistic when testing the balance of matched data? I have heard elsewhere
that a value of 0.25 or lower is acceptable, but I am unsure.
Thanks,
Evann
====================================
Evann Smith
Ph.D. Student
Harvard University, Department of Government
egsmith(a)fas.harvard.edu
====================================
--
MatchIt mailing list served by Harvard-MIT Data Center
List Address: matchit(a)lists.gking.harvard.edu
Subscribe/Unsubscribe:
http://lists.gking.harvard.edu/?info=matchit
MatchIt Software and Documentation:
http://gking.harvard.edu/matchit/
Browse/Search List Archive:
http://lists.hmdc.harvard.edu/lists/matchit/