Online Proctoring Discount: The Role of Measured Stressors

Abstract

We examine the exam score effect of online proctoring with a comprehensive split sample of proctored and unproctored students in an upper-level economics class. Ours is the first study to measure the online proctoring effect, to our knowledge, that combines student in-class performance metrics, measured student perceptions and stress, and administrative data. We first document a proctoring discount (i.e., the tendency for lower scores on proctored exams versus on unproctored exams). We also find a relationship between students’ perceptions of the online proctoring environment and their performance on online exams. Our analysis highlights the importance of student perceptions and beliefs in determining performance and suggests that educators need to address student perceptions in order to accurately assess student mastery of course materials.Using ordinary least squares regressions and multilevel mixed-effects models, our findings show that students’ perceptions and characteristics can affect online exam performance and, potentially, contribute to the observed proctoring discount. Assuming that the proctoring discount is caused, at least in part, by undesirable phenomena (i.e., academic dishonesty and/or student discomfort with online proctoring), we suggest policy implications that the elimination of this proctoring discount is a worthwhile goal of the educational community in an environment where online teaching and learning is prevalent and growing due to the COVID-19 pandemic.