The influence of classroom context on the probability of being caught cheating is compared between face-to-face classes and online classes. A decision tree model assigned in the context of a management science class presents alternatives, including unethical choices, risks and rewards, and a decision facing a potential ethical dilemma. Part of the student response to the assignment is estimation of the subjective probability of being caught copying on homework. Student-estimated probabilities of being caught for both "real" (face-to-face) and "virtual" classrooms are compared. The same information was collected from students at points in time ten years apart to study the change in technological context on the probability of being caught. Broadly speaking, student respondents felt that only about one third of cheaters get caught, leading to a dilemma where sometimes grades can be improved by cheating. More specifically, the probability of being caught was higher in a face-to-face classroom than a virtual classroom. That difference was significant in the earlier time period but was no longer significant ten years later.