Abstract

Numerous academic studies and reports indicate that as many as half of all students cheat on exams. Cheating on exams undermines the central purpose of a university, corrupts the meaning of grades as a measure of subject matter mastery, and significantly harms honest students. Although instructors are aware that many students cheat and they clearly oppose the behavior, they do little to punish cheaters. Accusing, prosecuting and convicting cheaters are time intensive, stressful and potentially costly activities for which faculty members receive few rewards. In this paper, we derive an equation to estimate the benefit that can be gained by a student who copies on a multiple choice exam. We then propose an exam design that not only eliminates the benefit, but also proportionately punishes cheaters, with little to no cost to instructors. Moreover, the exam system we propose can allow an instructor to determine, with a high degree of certainty, the odds that any student seated anywhere in the classroom cheated on any part of the exam.