New article in JOSS - konfound: An R Sensitivity Analysis Package to Quantify the Robustness of Causal Inferences

For many years, I’ve been a part of a project focused on developing statistical methods and statistical software tools for sensitivity analyses or techniques for understanding how robust a statistical inference is. I’ve played a role in developing the R package for the project. I’ve also played a role in developing an interactive (Shiny) web app to make the functionality of the R package even easier to use. Here are some links:

• The R package for the project is here: https://github.com/konfound-project/konfound

• You can find the Shiny here: https://konfound-project.shinyapps.io/konfound-it/

• Some info on the wider project is here: http://konfound-it.org

We recently published an article in the Journal of Open Source Software that is intended to serve as the source of record for the R package. The article is brief, as the review process focused as much (or more) on the R package’s code base than on the article’s text. In any case, the brief article can be found here. It was immensely challenging but fun to dive back into the R code; the last time I had was when I was a graduate student more than five years ago! I am glad to still be a part of the collaboration.