Abstract

This paper provides a model for testing the relation between a particular action (cheating) and ethics education. The test is for a difference in the incidence of cheating (answer copying) between two groups: students who have and students who have not taken a course in ethics. The model facilitates testing by obtaining a relation between the unobservable variable (cheating) and an observable variable (a wrong answer on the target question which is the same as the answer of a nearby student). The required sample size is large but roughly comparable to the size that has been used with an answer copying index. Unlike an answer copying index, the model does not rely on extensive copying by individual students. The model is best suited for faculty who use in-class, multiple-choice exams