Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: psmatch2 question


From   Alistair Windsor <alistair.windsor@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   st: psmatch2 question
Date   Tue, 28 Dec 2010 14:46:00 -0600

Dear Statalisters,

I am using psmatch2 to generate a matching group to estimate the average effect of the treatment on the treated. The group produced by these matching techniques is a proxy for the effect of not treating the treated (a counterfactual). There are various options to psmatch2, choices for stratification, and alternative matching approaches that I wish to compare. My first step is look at the summary statistics of the propensity scores and covariates over the treatment and matched groups.

Here is my problem: After running psmatch2 with the ties option

psmatch2 treatvar covariate_list, outcome(outcome_var) common ties ate

I expect to get _weight that sums to the number of observations in the treatment group (actually to the number of observations in the treatment group after common support is imposed using common). In my case this is 98. However when I type

sum outcome_var [aweight = _weight] if _treated==0

I get a sum of weights of 92.924 (I am also unsure of what psmatch2 is storing in _weight for treated observations - as far as I can see nothing is mentioned in the documentation). I also expected the mean here to give me the estimate for the effect of not treating the treated - it is close but not equal to that returned by psmatch2. I certainly would like to be able to duplicate the output so I can be sure I know what is being done by the script.

I am sure that I am missing something. Can anyone enlighten me?

Alistair

*
*   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/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index