Over the winter break, all the students in a 300-person computer science class at Australian National University (ANU) were told that their grades would be docked 30% due to rampant cheating and plagiarism on their final project.
According to the ANU computer science convenor Hanna Kurniawati, this was because there was “massive academic misconduct reports” and that they school was unable to identify who the perpetrators were.
Her letter went on to say that they students should not complain to the teaching staff, but instead “you should complain to your colleagues who were trying to outsource” the project.
According to research done by the Australian Broadcast Corporation (ABC), the issue came to light after the school learned of ads offering to complete the assignment in exchange for payment.
This course of action prompted a strong backlash, in particular from the school’s Computer Science Students Association and the school quickly reversed course after news of the punishment broke.
According to the school’s director of computer sciences, Tony Hoskings, the grade adjustment was not in response to the alleged infractions but because the marks were skewed, and their standard procedure called for them to be shifted.
Nonetheless, in order to “alleviate any possible suspicion of the connection between the routine mark adjustments and the detection of widespread academic misconduct” the school reversed the changes.
While this is good news for the students that were in the class, it does beg one question: Why not punish a whole class for widespread plagiarism or academic fraud?
The reason is because it’s a terrible idea.
The Definition of Perverse Incentives
Let’s pretend that 60 students in the class of 300 plagiarized or otherwise cheated on their assignment. That would represent 20% of the class and could certainly be perceived as widespread or rampant cheating.
However, that still leaves 240 students that didn’t cheat and completed the assignment to the best of their abilities.
Cheating and plagiarism is, fundamentally, about putting outcomes ahead of processes. If a student doesn’t care about process of completing the assignment and what they hope to learn or gain from it, they may be tempted to take shortcuts.
However, such students must weigh the risks before they decide to cheat. They balance the risk of getting caught and the consequences of getting caught versus reward of taking a shortcut on an assignment.
Now let’s assume that cheating was discovered and only two options present themselves (though there are always more options):
- Punish all of the students
- Punish none of the students
For the cheaters, this isn’t a loss. They cheated because they wanted an advantage over their peers and what they would have gotten without the unethical help. Since the penalty is applied uniformly, that advantage remains. That 30% reduction still gives them a higher score than if they didn’t cheat.
However, the real tragedy is for the 240 students that didn’t cheat. Many of them weighed the same risks and rewards only to come up with a different answer. Punishing them removes the risks of cheating, since they can be punished for doing nothing wrong, and leaves only the rewards.
In short, punishing a whole class in this manner doesn’t punish the cheaters but it does encourage the honest students to cheat in the future. Cheaters still gained an advantage and honest students no longer fear the risks as they are just as real if they’re honest.
This creates a perverse incentive that encourages cheating. Though the hope is that the non-cheating students will pressure the cheaters or that guilt of harming their peers will motivate cheaters to either change their ways or, ideally, turn themselves in.
But the honest students may not know who the cheaters are to apply such pressure and the cheaters aren’t likely to grow a conscience.
In short, all such action does is encourage honest students to cheat while failing to punish those that deserve it.
Much of this might seem elementary, akin to questions about why you don’t punish the whole class to stop one student’s misbehavior. That’s because it is.
Punishing honest students alienates them much the same way punishing well-behaved kids alienates them. All the while, it does little to stop the behavior that you want to end.
However, as frustration with plagiarism and cheating grows and stories like the Harvard cheating scandal of 2012 reminds us that cheating isn’t always the work of a lone student, this approach becomes more tempting.
This puts the focus back on good policies, creating plagiarism-resistant assignments and having robust tools for detecting cheating.
While all of those are difficult things in the middle of a pandemic, that doesn’t mean it’s wise for schools or faculty to turn to the nuclear option when it comes to enforcing academic integrity.