Using student self-reported cheating admissions and answers from a hypothetical cheating scenario, this paper analyzes the effects of individual and situational factors on potential cheating behavior. Results confirm several conclusions about student factors that are related to cheating. The probability of cheating is associated with younger students, lower GPAs, alcohol consumption, fraternity/sorority membership, and having cheated in high school. Student perceptions of the certainty and severity of punishment appear to have a negative and significant impact on the probability of cheating (...) on in-class assignments. Students who report a belief that cheating is never acceptable appear to be significantly less likely to cheat in any circumstance. This study illustrates the context-dependent nature of academic dishonesty, and the associated difficulty in understanding the relationships between measurable factors and cheating behavior. (shrink)

This article investigates the propensity for academic dishonesty by university students using the partitioning method of decision tree analysis. A set of prediction rules are presented, and conclusions are drawn. To provide context for the decision tree approach, the partition process is compared with results of more traditional probit regression models. Results of the decision tree analysis complement the probit models in terms of predictive accuracy and confirm results previously found in the literature. In particular, students’ moral character—whether they believe (...) cheating is acceptable—is found to be the most important factor in determining the propensity for academic dishonesty. (shrink)