NYU Prof Vows Never to Probe Cheating Again—and Faces a Backlash:
A New York University professor’s blog post is opening a rare public window on the painful classroom consequences of using plagiarism-detection software to aggressively police cheating students.
The result was an education in “how pervasive cheating is in our courses,” Mr. Ipeirotis wrote. By the end of the semester, 22 out of the 108 students had admitted cheating.
The professor’s blog post described how crusading against cheating poisoned the class environment and therefore dragged down his teaching evaluations.
Worse, Mr. Ipeirotis’ campaign aroused mistrust. Students were anxious, discussions contentious. He found teaching to be exhausting rather than refreshing. Dealing with the 22 cheating cases sucked up more than 45 hours “in completely unproductive discussions,” forcing him to focus attention on the least-deserving students, Mr. Ipeirotis said.
I think there are valuable lessons to be learnt from this experiment.
- Focus on those students who really want to learn, and work to provide an environment where opportunities for learning are maximised. Looking beyond education, companies should focus on rewarding and engaging their good customers, rather than pursuing and punishing those who are not. The music & computer game industries immediately spring to mind.
- As I observed very early in my own education, cheaters exist and obtain unfair short-term advantage. But cheaters also fail to learn, and in the long term they fall well behind. There are many opportunities to evaluate performance where the opportunities for cheating are minimised, such as closed book exams and class presentations. Perhaps, rather than providing the automatic plagiarism detection report to the lecturer, the report could be provided first to the student, so that student could amend their assignment prior to final submission.
- Be careful what you measure. If you base teacher rewards on positive student evaluations, then you may have created some unintended consequences. Fewer students will fail (since a failing student is unlikely to be a happy student). And plagiarism and other transgressions may tend to be overlooked.
Pingback: Moral hazards of plagarism detection | Stephen Tyler