Dear authors,
I was trying to go through your analysis example for using MatchIt and Zelig,
and after entering demo(analysis), I get the following error after a few
steps:
> ate.all <- c(s.out1$qi$att.ev, -s.out2$qi$att.ev)
Error in -s.out2$qi$att.ev : invalid argument to unary operator
and the example stops.
I am using Zelig 2.4-5 and MatchIt 2.2-11.
Don't know if this is my error or part of the package.
Vinh Nguyen
--
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/
with least squares as your analysis model, and exact one-to-one matching
you will get what you expect - identical results for the diff in means and
regression adjustment. MatchIt is doing a slightly different version of
exact matching, allowing multiple controls to be matched to a single
treatment if they're all exactly matched. That will typically not be very
different from the result above except in the case where the effects of
your explanatory variables are very large, which does appear to be the
case in your data.
I'm CCing our listserv in case my coauthors or others have anything to
add.
Gary King
---------- Forwarded message ----------
Date: Wed, 13 Sep 2006 09:29:31 -0500
From: Greg Blevins <gblevins(a)marketsolutionsgroup.com>
To: <King(a)Harvard.edu>
Subject: Question pertaining to MatchIt results
Dr. King,
My question is contained in the attached Word document which also includes
relevant statistical output.
Any guidance would be much appreciated.
Gregory L. Blevins
Senior Research Consultant
The Market Solutions Group, Inc.
Suite 1800
701 4th Avenue South
Minneapolis, MN 55415
612 392-3163 Direct
612 333-5400 Main
612 344-1408 Fax
612 251-0232 Cell
gblevins(a)marketsolutionsgroup.com
Gregory L. Blevins
The Market Solutions Group, Inc.
gblevins(a)marketsolutionsgroup.com
Office phone: 612 392-3163
Cell phone: 612 251-0232
"What is at the beginning but a small error, swells to huge proportions at the close." Aristotle