``experiment: R Package for Designing and Analyzing Randomized Experiments''



This package is still in its infancy. I'm planning to add more functionalities and improve the documentation over the next few years.


The package provides various statistical methods for designing and analyzing randomized experiments. One main functionality of the package is the implementation of randomized-block and matched-pair designs based on possibly multivariate pre-treatment covariates. The package also provides the tools to analyze various randomized experiments including cluster randomized experiments, randomized experiments with noncompliance, and randomized experiments with missing data.

  Current Version

The current version is 1.1-0 (built on 07/24/07) which runs on R 2.0.0 and higher. The source code and the documentation are available for download from The Comprehensive R Archive Network (CRAN).


Regardless of one's eperating system, the software can be installed by typing the following command at R prompt:

> install.packages("experiment")

This will install the latest version from CRAN.

© Kosuke Imai
Last modified: Fri Jun 12 11:22:45 EDT 2009