(with James Honaker and Gary King)

Amelia is an R package for the multiple imputation of incomplete data. Multiple imputation is a method to overcome the computational problem of missing data while maintaining good statistical properties. Amelia uses a bootstrap-based algorithm that increases speed and robustness. In addition, Amelia includes a graphical user interface that requires no knowledge of R at all.

causalsens is an R package to implement the selection bias approach to sensitivity analysis for causal effects as introduced in Blackwell (2014). This approach allows researchers to evaulate the effect of unmeasured confounders on their estimated effects varying both the strength and direction of the confounding.

(with Stefano Iacus, Giuseppe Porro and Gary King)

CEM is a package for R, Stata, and SPSS that implements the method of coarsened exact matching. CEM improves causal inferences and reduces model dependence by making observations more comparable. I helped to write the Stata and SPSS packages.

- To install the CEM package for Stata, you can use:
`net from https://www.mattblackwell.org/files/stata net install cem`

- CEM for Stata Manual
- CEM for SPSS

*(with Avi Acharya and Maya Sen)*

DirectEffects is an R package to estimate controlled direct effects. As of now, the only model supported is sequential g-estimation, but we plan to expand to other models, including doubly robust estimators, in the future. For more information on how CDEs can be useful for applied research and a brief introduction to sequential g-estimation, see our 2016 APSR. Note that this package is still in alpha stages and is under rapid development.

*(with Nicole Pashley)*

`factiv`

is an R package to analyze 2^{K} factorial experiments with noncompliance. Currently, it implements the finite-population and superpopulation inference approaches described in Blackwell and Pashley (2020). Note that this package is still in early stages. Learn how to install it on the github page.

*(with Michael Olson)*

`inters`

is an R package with tools to estimate interactions. Currently, it implements the post-double-selection approach to estimating interactions described in this working paper by Matthew Blackwell and Michael Olson. Note that this package is still in alpha stages and is under rapid development. Learn how to install it on the github page.