March 10, 2021
In this paper, soon-to-be-Doctor Weidmann used an administrative database to assess the magnitude of attrition bias in a collection of education randomized trials. There was not much to be found. Weidmann then extended this work, proposing a simple sensitivity analysis one might use on any trial of one's own or even read about to gauge how much attrition may have played a role in a found impact or lack thereof. To view, please click here.