Addressing the computer science scandal

Some students enrolled in Computer Science 201 found themselves especially anxious Wednesday evening. The cause wasn’t an upcoming midterm exam, but an email from the computer science department notifying students of an investigation into cheating in the class. It was especially noticeable on Yik Yak, where students took advantage of the application’s anonymity to comment on the news and express their concern. The investigation into cheating raises questions about the terms of acceptable collaboration and the ramifications for violating Duke’s community expectations.

In his comments with The Chronicle, Owen Astrachan, co-director of undergraduate studies and professor of the practice of computer science, admitted that there is a fine line between collaboration and cheating in computer science—online and in person. A tremendous amount of resources are available to students online, and the process of problem solving in computer science is not so much about pure inspiration as it is about creatively applying existing methods. For students working on problem sets together, the question of what types of collaboration are deemed acceptable is also pervasive. Can students discuss and puzzle through problems together and work on similar ideas and codes if the work is ultimately their own? With programs like MOSS—a tool used by the computer science department to check code by scanning the underlying data structure so as to find similarities and identify patterns that suggest duplicated work—the divide between acceptable and unacceptable collaboration is increasingly brought to the fore.

Yet, the issues arising in this computer science scandal evoke concerns that pervade across other disciplines of academia as well. In the humanities, the boundaries of collaboration appear clearer—to plagiarize is to copy work ad verbatim from a text or to present another’s original ideas as your own, among others. Yet, these delineations become blurred in many of the STEM field courses. Because the nature of the work is different in, for example, computer science than in physics, the terms and expectations of collaboration rightly fluctuate. Each department has different methods of combating students submitting unoriginal work, and each department should, like the computer science department, supplement the Duke Community Standard with their own clear expectations that are in turn communicated to students.

As the investigation unfolds, we question whether the current approach to sanctioning students found in violation of the policies is adequate. Currently, the department is offering first-time offenders the chance to admit any wrongdoing. In return, students can settle the issue through faculty-student resolution, which isn’t reported on students’ permanent records. On the one hand, stricter punishments could jolt the academic climate toward the “police state” that stifles student collaboration, an invaluable aspect of university life, according to Astrachan. Yet the other extreme is equally, if not more, undesirable: the possibility that relaxed consequences will cause students to feel able to routinely compromise the Community Standard. Cheating is an unacceptable behavior that dilutes a student’s ability to learn, and we strongly support the computer science department’s efforts to investigate the matter.

Ultimately, it is the prerogative of the department to decide the limits of acceptable behavior. But as they decide sanctions, it is important to note that their decisions will weigh heavily on the minds of all students and will send a message about cheating at large. At an elite academic institution like Duke, it is each student’s responsibility to ensure that your work reflects the integrity of one’s own efforts.