Statalist


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

st: nnmatch: Use only some observables for matching, and the others as regression controls?


From   "Sergio I Prada" <sprada1@umbc.edu>
To   statalist@hsphsun2.harvard.edu
Subject   st: nnmatch: Use only some observables for matching, and the others as regression controls?
Date   Thu, 11 Feb 2010 18:29:39 -0500 (EST)

Here a great summary on matching:

http://www.biostat.jhsph.edu/~estuart/Stuart-MatchingMethods-StatSci-Dec09.pdf

You can use as many variables as you want in your algorithm to find the
best balanced sample. Once you have it, use those weights and run
regression models with all the covariates again to adjust for the
remaining bias.

See this advice in page 537
Matthew Blackwell, Stefano Iacus, Gary King, Giuseppe Porro (2009). cem:
Coarsened exact matching in Stata. The Stata Journal, Volume 9 Number 4:
pp. 524-546
 http://www.stata-journal.com/article.html?article=st0176

-- 
Sergio

*
*   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   |   What's new   |   Site index