| Previous: Full Matching | Up: Examples | Next: Coarsened Exact Matching |
Genetic matching automates the process of finding a good matching solution (Diamond and Sekhon, 2005). The idea is to use a genetic search algorithm to find a set of weights for each covariate such that the a version of optimal balance is achieved after matching. As currently implemented, matching is done with replacement using the matching method of Abadie and Imbens (2007) and balance is determined by two univariate tests, paired t-tests for dichotomous variables and a Kolmogorov-Smirnov test for multinomial and continuous variables, but these options can be changed.
Genetic matching can be performed with MATCHIT by setting method = "genetic", which automatically loads the Matching (, ) package. The following example of genetic matching (using the estimated propensity score based on logistic regression as one of the covariates) can also be run by typing demo(genetic):
> m.out <- matchit(treat ~ age + educ + black + hispan + married + nodegree +
re74 + re75, data = lalonde, method = "genetic")