Interface Andrew G. Dean, EpiInformatics.com, and Roger Mir Pairs of Cases and Controls matched 1:1 on defined characteristics are evaluated for exposure to a risk factor and distributed into the four cells of the table representing the four possible combinations of outcome in a pair. The coun...
Briefly, this method allows one to exactly match on a small number of covariates (in this sense the method is sparse because it does not consider all possible matched pairs). The matching algorithm allows the near fine balance on discrete covariates between the intervention and control group (...