Re: st: Implementation of Propensity Score Matching, BalancingProperty

[David Radwin <radwin@berkeley.edu>]

You and other Stata users interested in matching might look at this 
recent paper, which comes with accompanying open source software for 
Stata and R. Instructions for installing the software are at 
http://gking.harvard.edu/cem/ .

Right now I am not in any position to pass judgment on its relative 
strengths and weaknesses compared to -psmatch2- or other 
alternatives, but maybe someone else has an opinion.

David
-----------
Stefano M. Iacus, Gary King, and Giuseppe Porro, "Matching for Causal 
Inference Without Balance Checking"; copy at 
http://gking.harvard.edu/files/abs/cem-abs.shtml.

Abstract: We address a major discrepancy in matching methods for 
causal inference in observational data. Since these data are 
typically plentiful, the goal of matching is to reduce bias and only 
secondarily to keep variance low. However, most matching methods seem 
designed for the opposite problem, guaranteeing sample size ex ante 
but limiting bias by controlling for covariates through reductions in 
the imbalance between treated and control groups only ex post and 
only sometimes. (The resulting practical difficulty may explain why 
most published applications do not check whether imbalance was 
reduced and so may not even be decreasing bias.) We introduce a new 
class "Monotonic Imbalance Bounding" (MIB) matching methods that 
enables one to choose a fixed level of maximum imbalance, or to 
reduce maximum imbalance for one variable without changing the 
maximum imbalance for the others. We then discuss a specific MIB 
method called "Coarsened Exact Matching" (CEM) which, unlike most 
existing approaches, also explicitly bounds through ex ante user 
choice both the degree of model dependence and the treatment effect 
estimation error, eliminates the need for a separate procedure to 
restrict data to common support, meets the congruence principle, is 
robust to measurement error, works well with modern methods of 
imputation for missing data, is computationally efficient even with 
massive data sets, and is easy to understand and use. This method can 
improve causal inferences in a wide range of applications, and may be 
preferred for simplicity of use even when it is possible to design 
superior methods for particular problems. We also make available open 
source software which implements all our suggestions.

At 10:03 AM -0600 9/10/08, Pena,Anita wrote:

> Dear Statalist,
> I am struggling with implementing propensity score matching. 
> Specifically, I am stuck with the balancing property and differences 
> between pscore followed by att*, psmatch2 with pstest, and nnmatch. 
> I have tried several specifications using pscore for which the 
> balancing property test consistently fails.  Using pstest after 
> psmatch2, however, balancing seems to be satisfied for the same 
> specifications that fail under pscore.  I would greatly appreciate 
> if someone who has recently implemented propensity score matching in 
> STATA could provide pointers on how to get balancing and also any 
> insight on the advantages/disadvantages of the three alternative 
> commands.
> Many thanks!
> ----------------------------------------
> Dr. Anita Alves Pena
> Department of Economics
> Colorado State University
> <http://lamar.colostate.edu/~aalves>
>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/

--
David Radwin // radwin@berkeley.edu
Office of Student Research, University of California, Berkeley
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/