Examining the Effect of Proctoring on Online Test Scores

Abstract

Online education continues to grow, bringing opportunities and challenges for students and instructors. One challenge is the perception that academic integrity associated with online tests is compromised due to undetected cheating that yields artificially higher grades. To address these concerns, proctoring software has been developed to address and prevent academic dishonesty. The purpose of this study was to compare online test results from proctored versus unproctored online tests. Test performance of 147 students enrolled in multiple sections of an online course were compared using linear mixed effects models with nearly half the students having no proctoring and the remainder required to use online proctoring software. Students scored, on average, 17 points lower [95% CI: 14, 20] and used significantly less time in online tests that used proctoring software versus unproctored tests. Significant grade disparity and different time usage occurred on different exams, both across and within sections of the same course where some students used test proctoring software and others did not. Implications and suggestions for incorporating strategic interventions to address integrity, addressing disparate test scores, and validating student knowledge in online classes are discussed.