A field experiment into the cheating habits of undergraduate students sheds new light on the use of threats and sanctions to curb dishonest scholastic behavior. Georgia Tech School of Economics professor Daniel Dench, along with Theodore Joyce at Baruch College, published the research online in the February 14, 2022 issue of Journal of Economic Behavior and Organization. They conclude that a professor informing students that the school has the means to detect cheating has little effect. But notifying students that they have been caught and are on a watch list reduces subsequent cheating attempts by 75 percent.
Dench wrote, “As a first step in devising a strategy to deter cheating, we designed an experiment to determine whether direct messages to students regarding our ability to detect cheating would curb it or whether we needed some form of sanction before students appreciated the seriousness” of dishonesty. To address these questions, 3,515 students were randomly divided in each of four courses into groups A and B. Students were unaware which group they were in.
“We first amended the course syllabus statement on academic integrity. We reminded students to submit their own work, but we explicitly informed them that we had software to detect any work copied from another student,” Dench wrote. Furthermore, the software could identify both the sender and the user of the plagiarized material. This information was available to both groups A and B; one week before the first assignment was due, Dench and Joyce emailed Group A, reminding students to submit their own work and that software could detect any work they copied from another spreadsheet. Those caught cheating on the first assignment would be put on a watch list for subsequent assignments. Further violations of academic integrity would involve their course instructor for further disciplinary action.
Group B received the same email one week before the second assignment, and all students flagged for cheating in either of the two assignments were informed that they were currently on a watch list for the rest of the semester’s assignments.
The study found that warning students about the software’s ability to detect cheating has a practically small and statistically insignificant effect on cheating rates, but flagging cheaters, and putting them at risk for sanctions lowers cheating by approximately 75 percent.
The study contributes to the literature on cheating among college students in several novel ways. First, it measures the extent of cheating with software.
A second contribution is that the study tested whether students who are informed that they used another student’s work on their first assignment, are less likely to do so on subsequent assignments.
Finally, they suggestive evidence on how warning and demonstrating ability to detect cheating has an impact on cheating in subsequent semesters. They show that in semesters following the experiment, cheating rates were dramatically lower in first assignments before any sanctions to students who cheated could have occurred. Even without any explicit punishments such as reports to the Dean’s office or grade impacts, showing students you have the ability to detect their cheating behavior deters students in future semesters from cheating.
The study differs from previous plagiarism research (Dee and Jacob, 2012) in that it focused on two aspects of deterrence: informational warnings regarding our ability to detect cheating and explicit demonstration of our ability to catch cheating. In addition, Dench and Joyce’s measure of cheating is more objective, relying on detection of copying directly from other students rather than combinations of words.