RE: Re: st: PSM: Balance diagnostics using the empirical sampling distribution of the standardized difference

["Radwin, David" <dradwin@rti.org>]

To clarify, -pstest- (part of -psmatch2-, Leuven and Sianesi, SSC) does use t-tests. 

An objection to t-tests as a measure of balance in matching is articulated in section 7.2 of

Imai, K., King, G., & Stuart, E. A. (2008). Misunderstandings between experimentalists and observationalists about causal inference. Journal of the royal statistical society: series A (statistics in society), 171(2), 481-502. http://imai.princeton.edu/research/files/matchse.pdf

David

--
David Radwin
Senior Research Associate
Education Studies Division
RTI International
2150 Shattuck Ave., Suite 800
Berkeley, CA 94704
Phone: 510-665-8274

www.rti.org


> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-
> statalist@hsphsun2.harvard.edu] On Behalf Of Ariel Linden, DrPH
> Sent: Thursday, August 15, 2013 7:02 AM
> To: statalist@hsphsun2.harvard.edu
> Subject: RE: Re: st: PSM: Balance diagnostics using the empirical sampling
> distribution of the standardized difference
> 
> A correction and some additions:
> 
> -ebalance- (http://fmwww.bc.edu/RePEc/bocode/e) is not for checking
> balance.
> It is a program that uses an entropy weighting approach to achieve balance
> in covariates.
> 
> -pstest- (http://fmwww.bc.edu/RePEc/bocode/p) is a post-estimation command
> that comes part of the -psmatch2- program
> 
> -imbalance- (http://fmwww.bc.edu/RePEc/bocode/c) is a post-estimation
> command that comes part of the -cem- program
> 
> 
> 
> Date: Wed, 14 Aug 2013 08:42:54 -0400
> From: Richard Goldstein <richgold@ix.netcom.com>
> Subject: Re: st: PSM: Balance diagnostics using the empirical sampling
> distribution of the standardized difference
> 
> there are at least 3 user-written routines that might be related to this
> article; use -findit- to find and install these
> 
> ebalance
> pbalchk
> mpbalchk (version of preceding for MI data)
> 
> in addition for those using -cem- for matching, there are routines from
> the -cem- authors also
> 
> Rich
> 
> On 8/14/13 8:31 AM, Lukas Borkowski wrote:
> > Dear list,
> >
> > based on this statalist post
> (http://www.stata.com/statalist/archive/2013-01/msg00516.html) I have read
> Austin, P.C. (2009) on how to check the balancing properties of the
> covariates between matched treated and untreated observations after PSM
> using the empirical sampling distribution of the standardized difference
> as
> alternative to t-tests. I was wondering whether there is a Stata command
> (user-written or else) to run this approach? My search was not successful.
> >
> > Best, Lukas
> >
> > Austin, P.C., 2009. Balance diagnostics for comparing the distribution
> of
> baseline covariates between treatment groups in propensity-score matched
> samples. Statistics in Medicine, 28(25), pp.3083-3107.
> >
> > #
> > Lukas Borkowski
> > University of London, School of Oriental and African Studies (SOAS)


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/