Academic Integrity in Traditional Vs Online Undergraduate Medical Education Amidst COVID-19 Pandemic

Due to the coronavirus disease (COVID-19) pandemic, undergraduate medical education made a drastic paradigm shift towards online learning and assessment. This study was aimed at developing a statistical approach to empirically analyze the academic integrity of the online exams. Data were collected retrospectively from academic records of Physiology and Anatomy courses of a private medical college, in which students had attempted sequential on-campus and online modular exams. We developed a statistical model to predict on-campus and online exam scores of undergraduate medical students, from their previous academic records. Hypothesizing that the predictor model contains comparable explanatory power for both exams, we utilized the explanatory power variation (R(2) statistics) of the improvised model to predict academic dishonesty behavior in traditional vs online exams. Reduced explanatory power (R(2 )statistics) of the model for any mode of the exam, in which students scored considerably different from their previous academic record, was interpreted as indirect evidence of academic dishonesty. Our model explained a large proportion of variation (R(2)) in overall scores of on-campus and online Physiology and Anatomy modules. Whereas, the model could explain only a small proportion of variation in scores of online theory exams, with a moderate effect size (adjusted R(2)). Reduced explanatory power for both Physiology and Anatomy online theory exams implies the chance of academic dishonesty in this particular component of the online exam. Explanatory power variation of a predictor statistical model can be further explored and utilized to monitor academic integrity in future online exams.