3 Level Propensity score matching

[newbie_ari]

Hi,

In my study I've 3 levels, 2 treatment and 1 control effect.  I want to do propensity score matching at 3 level. Could anyone please guide me how to proceed doing PSM. Also, if there's any literature I can refer to.

Thanks a lot!

[EastwoodDC]

You need a multinomial model to estimate the probability of being assigned to each of the three groups. From that you create TWO propensity score variables, and use both in the matching. 

Macros for greedy or optimal mathing are available here:  http://www.mayo.edu/research/departments-divisions/department-health-sciences-research/division-biom...

Greedy matching is very easy to use, and works fine for large data sets. Optimal matching may be a better choice for smaller data sets.