New IES-funded project - Quantifying the Robustness of Causal Inferences

Joshua Rosenberg


This project is still a ways off from starting, but I’m beginning a new IES-funded methodological project. This work is on quantifying the robustness of causal inferences in educational research. It builds on work that Ken Frank at Michigan State University did to develop a widely-used approach for doing just that: quantifying how sensitive an inference about an effect is to sources of bias. My role on this work has been to develop statistical software, including the konfound R package and Konfound-It! Shiny app. I like to stay sharp methodologically and I like this approach, having used it in several papers (i.e., this and this paper).

I also like the team and am excited that an outstanding doctoral student here at UTK with unsurpassable R chops, Sarah Narvaiz, has agreed to work on the project for the next year.

More on the project is here: