There are a lot of ways to do power calculations, but I wanted to show off a package, PUMP, that my collaborators and I recently set out into the world. The original purpose of the package was to provide tools for power analyses that account for multiple outcomes under a variety of common experimental designs—an …

A simulation and R tutorial demonstrating how to recover interpretable treatment effects from logistic regression.

When estimating average treatment impact for a multisite randomized trial, recent literature (see Bloom et al. (2017)) has suggested using the so-called FIRC (Fixed Intercept, Random Coefficient) model. There are some technical details that make fitting this model slightly tricky in R; this post talks through these …

You’ve taken the leap, and you wrote a manuscript using R Markdown. Good job! I support your endeavor, especially as it is a good practice for reproducible research. You’ve finished your work, and you want to submit your manuscript for publication. Unfortunately, the journal has various style and formatting …

Does becoming President of the United States cause you to age faster than you would have otherwise, if you had not become President? Media sources have claimed that presidents age more quickly while in office. The British tabloid The Daily Mail asserted “American presidents age TWICE as fast while in office” (Daily …

This is a live document that is subject to updating at any time. Introduction This document, the first of a three part series, is an introduction to matching in R using the optmatch and RItools packages. We focus on unit-to-unit matching, as opposed to other recent approaches that focus on obtaining overall balance in …