Previous: Quick Overview | Up: Preprocessing via Matching | Next: Exact Matching |

To run any of the examples below, you first must load the library and and data:

> library(MatchIt) > data(lalonde)

Our example data set is a subset of the job training program analyzed
in Lalonde (1986) and Dehejia and Wahba (1999). MATCHIT includes a
subsample of the original data consisting of the National Supported
Work Demonstration (NSW) treated group and the comparison sample from
the Population Survey of Income Dynamics (PSID).^{3.1} The variables in this
data set include participation in the job training program
(`treat`, which is equal to 1 if participated in the program,
and 0 otherwise), age (`age`), years of education (`educ`),
race (`black` which is equal to 1 if black, and 0 otherwise;
`hispan` which is equal to 1 if hispanic, and 0 otherwise),
marital status (`married`, which is equal to 1 if married, 0
otherwise), high school degree (`nodegree`, which is equal to 1
if no degree, 0 otherwise), 1974 real earnings (`re74`), 1975
real earnings (`re75`), and the main outcome variable, 1978
real earnings (`re78`).

- Exact Matching
- Subclassification
- Nearest Neighbor Matching
- Optimal Matching
- Full Matching
- Genetic Matching
- Coarsened Exact Matching

Gary King 2010-12-11