Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

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

RE: st: psmatch2 with exact matching

From   "Caskey, Judson" <>
To   "" <>
Subject   RE: st: psmatch2 with exact matching
Date   Tue, 22 Nov 2011 07:41:43 -0800

> From: Charles Stoecker <>
> Subject: st: psmatch2 with exact matching
> Hi Statalist,
> I'm using the add-on psmatch2 (by Barbara Sinansei and Edwin Leuven -
> ssc install psmatch2) to conduct propensity score matching.  I would
> like to use it to conduct an exact match on state and then further
> match by propensity scores within state.  In the helpfile for the
> command it indicates this can be done with:
>         g att = .
>         egen g = group(groupvars)
>         levels g, local(gr)
>         qui foreach j of local gr {
>                 psmatch2 treatvar varlist if g==`j', out(outvar)
>                 replace att = r(att) if  g==`j'
>         }
>         sum att
> This will provide the estimate of the difference between treatment and
> control groups.  My question is, how do I get the standard error of
> the estimated difference?

See the following message post for one way to get exact matches:

What you can do is match on a variable that has integer codes for state, and p-score in the decimal positions. You use caliper to ensure exact matches on your state code.


Judson Caskey
UCLA Anderson School of Management
110 Westwood Plaza, D416
Los Angeles, CA  90095
Office:                  (310)206-1503
Mobile:                (310)775-0080

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index