Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: Balance for PSM
From
"Garrido, Melissa" <[email protected]>
To
"'[email protected]'" <[email protected]>
Subject
Re: st: Balance for PSM
Date
Tue, 3 Dec 2013 16:01:33 +0000
Carlos,
A few other thoughts on your discrepant results:
You asked:
"1) The first results I had with psmatch2.ado were wrong (unbalanced)?"
Not necessarily; the first run with nearest neighbor matching indicates balance in your matched sample via standardized differences (a much better test of balance than t-tests). You might want to run a few other balance tests (including graphing distributions of variables in the treated and comparison groups) before and after matching your sample by propensity score before deciding that you have sufficient balance.
"2) If the answer is no, do I have to get a better PC to process Radius Matching with psmatch2.ado?
3) If the answer is yes, why psmatch2.ado worked without Radius and did not worked with Radius?"
With your radius matching code, it appears you did not specify a caliper. Try specifying .2(standard deviation of the of the logit of your propensity score) and see if that makes a difference.
The two matching strategies work differently - nearest neighbor will match a control person to a treated person, no matter how terrible the match. Radius matching will match a comparison observation to a treated observation only if their propensity scores are within a certain distance (caliper) from each other.
4) Is it possible that my sample is not good for PSM?
Only you can make that determination. As Jorge pointed out, there may be other variables related to both treatment and outcome in your sample that are not included in your propensity score. If so, it is likely that you
will obtain a biased treatment effect.
Hope this helps!
Melissa
--------------------------------------------------------------------
Melissa Garrido, PhD
Research Health Science Specialist
GRECC, James J Peters VA Medical Center, Bronx, NY
Assistant Professor
Brookdale Department of Geriatrics & Palliative Medicine
Icahn School of Medicine at Mount Sinai, New York, NY
718-584-9000 x 3804
[email protected]
*
* 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/