Guest post by Derek Newton of The Cheat Sheet: Lowering the stakes of assessment to reduce cheating

Featured guest post by Derek Newton, publisher and founder of The Cheat Sheet.   You can listen to our conversation with Derek about the ever inventive world of exam cheating in Episode 3 of The VICTVS Podcast. 

When the pandemic hit and educators were essentially forced into remote learning and remote assessments, they also had to confront the realities of rampant academic misconduct spawned and fueled by distance, technology adjustments, isolation, and increased opportunity.

In those environments, many teachers prioritized mitigation. Empathy and the constraints of the possible motivated teachers to be more flexible, to adjust their policies and expectations, to do what they could to keep the learning journey moving during some impossible circumstances.

One way many educators sought to ease the very real burdens of the time was by reducing the stakes of their assessments. Instead of a typical arrangement in which perhaps a third of a course grade was based on a midterm exam and another third on a final exam, teachers broke assessments into smaller, less valuable bites. Twenty smaller items of assessment each worth 5%, for example.

It was believed that at the time that lowering the stakes of assessments would not only reduce the pressure of make-or-break exams, but that lower stakes assessments would reduce cheating. Many educators and administrators recommended lowering assessment values during remote learning, believing and outright advising that a reduced relationship between the value of an assessment and a student’s final grade would reduce the pressure on students and, by extension, cheating.

In a 2021 podcast interview, Natasha Jankowski, a consultant on student learning and a lecturer at New England College said that changing assignments and making them “less high-stakes” was “the answer to more cheating.”

In another example, The University of Texas, at Austin gave advice to their faculty stating explicitly that, “instructors are encouraged to consider a testing plan where there are many short tests throughout the semester and where each test has a relatively low impact on the final grade. Dropping the lowest quiz or exam grade is another way to further decrease test anxiety. Reducing the stakes for quizzes and tests does not guarantee ethical behavior on these assessments.”

Turns out, not only does lowering the stakes of assessments “not guarantee ethical behavior,” reducing the stakes of assessments likely increased the likelihood of academic misconduct.

That conclusion is based mostly on inference and observation. There’s no direct test of the relationship between assessment value and rates of misconduct. At least not as far as I know. But there is decent evidence nonetheless.

As I’ve noted previously, in their foundational book “Cheating in College,” researchers McCabe, Butterfield and Trevino, found that students who engage in cheating behavior routinely say that “the assignment counted as a very small portion of my overall grade” as justification for cheating and that many students report that the low stakes nature of an assignment meant that their actions weren’t actually cheating at all.

And just a few weeks ago, I reported on a great article from a high school in California. In it, a student is quoted as saying, “I only have Canvas tests in two of my classes, and in one of them pretty much everyone Googles the answers, but —- are care-free because the quizzes have nearly no impact on our grades to begin with.”

This makes perfect sense.

In lowering the value of any given assessment, the consequences of being discovered for deception are low too. The reward may be low; but the risk is too. And because we know that most decisions related to engaging in misconduct are calculated and rational, the balance in risk and reward plays a major role.

That’s not to say that people don’t cheat – or try to cheat – on high stakes exams and assessments. They do. But most often, if the stakes are high, the fall of failure is a low way down. Consider a very high stakes assessment such as a legal bar exam or medical licensing exam. These exams represent the culmination of years of study and significant financial investment, exceptional social expectations – your friends and family know you’ve spent years in medical school. In exams such as those, the reward for cheating is high; you’re a doctor. But the risk is very high too. People know that. And they tend to act accordingly, not wanting to take the risk.

Further supporting the idea that lower stakes equals more cheating is the research that shows that efforts by teachers to reduce cheating signal seriousness. When teachers proctor exams, or review test data, or issue consequences for cheating, incidents of cheating go down. Not just because the risk calculations change, but because students infer the exam is important. The research also found that when students infer that assessments are not important, they infer permission to cheat. And lowering assessment values clearly signals a lack of importance.

But there’s also actual evidence too. As online learning retreated and schools could start to review its blessings and failures, the University of California system – one of the largest and best systems in the county – issued a memo. It said, in part, that “Prior strategies to provide multiple, lower-stakes assessments as a way to reduce academic misconduct may have been less effective during the pandemic.”

They don’t outright say that it does not work. But they did say that lowering the value of assessments as a cheating mitigation during the pandemic was not a solid approach.

Underlying the idea that reducing the value of performance metrics will reduce cheating is the idea that misconduct springs from stress – that students would not cheat if they were not pressured to do so. That the value of the tests themselves and the competitive nature of education and grades drive cheating. If we just stopped putting so much value on exams and grades, the idea goes, cheating would subside.

As you may expect, there’s no evidence to support that view. As mentioned, most often, cheating is a rational choice of risk and reward. Sure, the reward is the grade. But there is also reward in just not doing the work. There’s also reward in demonstrating success in a peer group – earned or not.

To support the idea that the rewards of social status and work avoidance are significant and motivational, consider that people cheat in situations in which the rewards are zero, where there are no grades and no financial incentives at all.

In 2021, cheating on the video game Call of Duty was so rampant that the manufacturer had to update the game and ban cheaters. The stakes of a video game and the pressure to succeed are highly limited, if not absent entirely. And yet cheating was pervasive. People will cheat when they can cheat. Even when the stakes are set to zero.

In other words, reducing the value of assessments as a prevention strategy for academic misconduct is very unlikely to work. Worse, the evidence shows that lowering assessment values may in fact encourage and incentivize cheating.

    Share this article

Leave a Reply

Your email address will not be published. Required fields are marked *

Fill out this field
Fill out this field
Please enter a valid email address.
You need to agree with the terms to proceed

error: Content is protected