Dear all,
I am trying to use Coarsened Exact Matching (CEM) in Stata and have run into a challenge.
I have a relatively large control group compared to my treatment group ( ~30.000 vs. 500 observations). In order to make my statistical test not overly sensitive I would therefore like to restrict the number of observations in each strata. I know this possibility exists using K-2-K matching but this allows me to do 1 to 1 matching as far as I know. Is there a way to define the number of units in each strata say to 5 or 10?
Any help is much appreciated.
Have a great day,
Lorenz
I am trying to use Coarsened Exact Matching (CEM) in Stata and have run into a challenge.
I have a relatively large control group compared to my treatment group ( ~30.000 vs. 500 observations). In order to make my statistical test not overly sensitive I would therefore like to restrict the number of observations in each strata. I know this possibility exists using K-2-K matching but this allows me to do 1 to 1 matching as far as I know. Is there a way to define the number of units in each strata say to 5 or 10?
Any help is much appreciated.
Have a great day,
Lorenz
