0
$\begingroup$

I am new to Mahalanobis distance matching and trying to see for information on how I can assess whether the matches I got using Mahalanobis matching are good matches.

gen long matchD3 = .
mahapick hhh_age55older hhh_edu nfemale1859 nmale1859 nmale05 nmale615 nmale61up nfemale05 
nfemale615 nfemale61up x1_17, idvar(a01) treated(program)pickids(matchD3) ///
genfile(matchD3) matchon(hhh_sex wealth provincename)  score

After I executed the above stata commands, I merged and stacked all the necessary files together, and now I would like to evaluate whether my treated observations and all their matches are sound.

Can someone please guide me on how to do this? I know for PSM we have to check the common support, balance, and etc to make sure if meets the PSM assumptions. For Mahalanobis, I am unsure of what post-matching method to take in order to check the quality of my resulting matches. Any guidance you may have will be helpful. Thank you!

$\endgroup$

1 Answer 1

0
$\begingroup$

You evaluate the quality of matches after Mahlanobis distance matching exactly the same as you would after propensity score matching: by assessing balance on the covariates and ensuring the effective sample size of the matched sample is adequate. The form of matching has nothing to do with how balance is assessed. The propensity score is not relevant for balance assessment and is only a heuristic for assessing overlap and therefore is not at all necessary for evaluating matches.

If your question is about actually how to write a Stata command to do that, that is a programming problem, not a statistics problem, that should be asked on StackOverflow.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.