What’s interesting is how institutions have gravitated towards different levers. The University of Michigan’s undergraduate program has focused on the detection method for assignments. Since the mid ’90s, students have submitted essays (and computer programs) online. These submissions are compared against all previous submissions and a broader internet search. Even though the existence of this detection system is highly publicized, several students are caught plagiarizing every year. I can distinctly remember one student received a failing grade for submitting two lines of plagiarized computer code buried in a program of over 10,000 lines. From my own experience, honest students appreciate the fair objectivity of a good detection system.

Most universities utilize the prevention method when administrating exams. Usually exams are proctored by professors and student assistants. In some cases paper and pens are provided for the exam to prevent students from cheating with crib sheets. By their implicit nature, these prevention techniques lead to a distrustful environment and generate a fair amount of student anxiety. Ultimately, fraud prevention interferes with the learning process.

The Darden honor system is exclusively focused on the penalization method. Given the strong message sent to students by the single sanction system (expulsion), prevention and detection are virtually unnecessary. Students and professors benefit from mutual trust and a healthy cooperative learning environment. Additionally, students are free of the anxiety created by prevention and detection methods and professors do not have to spend time managing fraud.

So why doesn’t every institution rely on penalization? In my opinion there are two reasons:

1. Culture is a hard thing to change. The University of Virginia stands behind the Honor System because it has worked since 1842. For another university to adopt this system would require a significant adjustment. For example, students who expected to receive a failing grade if caught cheating would actually be expelled. Also, the faculty may be comfortable with and unwilling to abandon the prevention and detection methods they have used for several years.
2. Not everyone is comfortable with a single sanction system. While expulsion may seem fair for a student that copied a friend’s final exam answers, the fairness of expulsion for “lesser” acts is sometimes questioned. I think people get stuck here by focusing on the act of fraud rather then the underling dishonesty. If there are no gradients in honesty (i.e. half-honest), then there probably shouldn’t be any gradients in punishment.

Personally, in an environment of trust, a take home exam is certainly more appealing then a proctored exam.