2
$\begingroup$

I have 50 control subjects and I would like to get 150 treatment subjects (i.e., a 1:3 ratio between the control and treatment groups) using propensity score matching and a few covariates. After performing the matching I would like to get indices of matched subjects (i.e., indices for both control and treatment groups) - this is my main goal.

I would like to do this in R. I see that R has the Matchit and Matching packages. However, I am not sure what the difference is between the two and which one is more appropriate for my purpose.

I am also not sure about the code and what output I need to be looking at to get indices of the matched control and treatment subjects.

Also, could I use the match() command without the outcome (i.e., Y) variable? The experiment is not yet complete and thus I do not have the outcome variable. Would it be appropriate to use match(Y=NULL, Tr=.....)?

$\endgroup$
  • $\begingroup$ I know that Match in the Matching package does not require you to supply a Y variable. $\endgroup$ – Charlie Jul 9 '13 at 16:34
3
$\begingroup$

Have a look at an overview here: http://www.biostat.jhsph.edu/~estuart/propensityscoresoftware.html

Match package does, as Charlie mentioned, let you exclude the outcome variable and will provide you with matching only then.

$\endgroup$

We're looking for long answers that provide some explanation and context. Don't just give a one-line answer; explain why your answer is right, ideally with citations. Answers that don't include explanations may be removed.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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