Type I Error Inflation in Unexpected Event During Survey Designs

• In Event: Experimental Methodology

Brief Overview

In this paper, we evaluate the false-positive rate of one of the most frequently used causal inference methods, the UESD. We find an inflated error rate and propose a fix for it. We demonstrate the utility of our approach with two replications.

Authors

• Joris Frese, European University Institute

• Sascha Riaz, European University Institute