add_model_matrix: Rebuilds the model matrix for MatchIt objects

In MatchItSE: Calculates SE for Matched Samples from 'MatchIt'

Description

In order to use estimate Standard Errors with the Abadie and Imben's method the MatchIt object needs to have a model matrix. The model matrix is created based on the subclasses given by the full matching procedure.

Usage

add_model_matrix(fit)

Arguments

fit	MatchIt Object

Details

This function is experimental! Most methods for estimating standard errors are only documented for NN matching. Please use with caution!

Value

MatchIt Object with added model matrix

Examples

## Not run: 
library(MatchIt)
data('lalonde')
m.out  <- matchit(treat ~ educ + black, data = lalonde, method = 'full')
att(obj = m.out, Y = lalonde$re78)
abadie_imbens_se(m.out, lalonde$re78)  # FAILS!
m.out <- add_model_matrix(m.out)
abadie_imbens_se(m.out, lalonde$re78)

## End(Not run)

MatchItSE documentation built on May 29, 2017, 4:30 p.m.