lalonde

This is one of the data sets from the National Supported Work Demonstration used by Dehejia and Wahba (1999) to evaluate propensity score matching methods. This and other related data sets are available at <a href="http://users.nber.org/~rdehejia/data/nswdata2.html">http://users.nber.org/~rdehejia/data/nswdata2.html</a>.

datasets

Includes functions for the construction of matched samples that are balanced and representative by design.  Among others, these functions can be used for matching in observational studies with treated and control units, with cases and controls, in related settings with instrumental variables, and in discontinuity designs.  Also, they can be used for the design of randomized experiments, for example, for matching before randomization.  By default, 'designmatch' uses the 'GLPK' optimization solver, but its performance is greatly enhanced by the 'Gurobi' optimization solver and its associated R interface.  For their installation, please follow the instructions at http://user.gurobi.com/download/gurobi-optimizer and http://www.gurobi.com/documentation/6.5/refman/r_api_overview.html.  We have also included directions in the gurobi_installation file in the inst folder.

Jose Zubizarreta

designmatch

Construction of Optimally Matched Samples for Randomized
Experiments and Observational Studies that are Balanced and
Representative by Design

lalonde function

A data frame with 445 observations, corresponding to 185 treated and 260 control subjects, and 10 variables. 
		The treatment assignment indicator is the first variable of the data frame: <code>treatment</code> (1 = treated; 0 = control).
		The next 7 columns are the covariates:<ul>
<li> <code>age</code>, measured in years;</li></ul>
<li> <code>education</code>, measured in years;</li>
<li> <code>black</code>, indicating race (1 if black, 0 otherwise);</li>
<li> <code>hispanic</code>, indicating race (1 if Hispanic, 0 otherwise);</li>
<li> <code>married</code>, indicating marital status (1 if married, 0 otherwise);</li>
<li> <code>nodegree</code>, indicating high school diploma (1 if no degree, 0 otherwise);</li>
<li> <code>re74</code>, real earnings in 1974;</li>
<li> <code>re75</code>, real earnings in 1975.
		</li>
The last variable of the data frame is <code>re78</code>, the real the earnings in 1978.

Format

This is one of the data sets from the National Supported Work Demonstration used by Dehejia and Wahba (1999) to evaluate propensity score matching methods. This and other related data sets are available at <a href = 'http://users.nber.org/~rdehejia/data/nswdata2.html'>http://users.nber.org/~rdehejia/data/nswdata2.html</a>.

lalonde: Lalonde data set

Description

Usage

Arguments

Format

References