Bookmark and Share

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]

st: Propensity score matching and multiple imputation


From   Michela Coppola <Coppola@mea.mpisoc.mpg.de>
To   "," <statalist@hsphsun2.harvard.edu>
Subject   st: Propensity score matching and multiple imputation
Date   Thu, 17 May 2012 11:51:03 +0000

Dear all,



I'm using psmatch2 to estimate with a propensity score matching the effect of a treatment on a certain outcome.



I'm using a multiply-imputed dataset and I have 5 implicates.



As  I understand, psmatch2 does not work with the "mi" suite (i.e. mi estimate: ...) so I have run the estimation separately on the 5 implicates so that I obtained the estimated effect and the standard errors in each of the 5 imputed datasets.



Then I combined (in excel) the estimates and the standard errors using the Rubin's rule.



Now the combined standard error using the Rubin's rule is  smaller than the standard errors in each of the 5 implicates! I expected bigger standard errors not smaller ones (as the variability introduced through the imputation is now taken into account).



How is it that possible (I checked several times the formulas, so I'm sure that there are no errors when combining in excel the standard errors)?


What is wrong in my procedure?

Thanks!



Michela



-- 



Dr. Michela Coppola

Research Fellow



MEA - Munich Center for the Economics of Aging

Max-Planck-Institute for Social Law and Social Policy

Amalienstr. 33

D- 80799 Munich



Office      +49 89 38602 360

Home office +49 89 3741 0259

Fax.        +49 89 38602 390

E-mail      coppola@mea.mpisoc.mpg.de








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