By Dr. Ashley Norris
Cheating in college is a problem and if you follow the news, frequency of cheating has increased dramatically during the pandemic. Derek Newton who frequently writes on this topic and publishes The Cheat Sheet reported in a recent issue that at least 32 schools had publicly reported cheating incidents in the previous 90 days. That’s one every 72 hours. Naturally, academics and parents are worried about how this diminishes the value of a degree or even worse, leaves a student ill prepared for their working life.
What’s more troubling to me, however, is the issue of equity
With all the attention we’ve given to the cheating scandals at West Point and Dartmouth Medical School, Texas A&M and dozens of other universities, even high schools, I am most bothered by how cheating further separates the haves and have-nots. How those who have the chutzpah or financial means to game the system are crushing those who follow the rules.
As Julianne Malveaux, former President at HBC Bennett College for Women in North Carolina, wrote recently, paid cheating sites are, “out of the budget for many of these students, but it also lowers the quality of education across the board by requiring less actual learning in exchange for quick and easy good grades. Unlike their white counterparts, many of these students do not have a safety net to fall back on when they lose access to quality education.”
Whether the incidences of cheating are increasing or whether we are simply paying more attention to them is immaterial. Malveaux is right; cheating, especially paid cheating, is an overlooked equity issue.
What truly matters is that if one student cheats in a class they are not only gaining an unfair advantage for themselves, they are squelching opportunities for their classmates. The matter compounds itself when a course is graded on a curve, or for gateway courses that have to be passed for a student to move on in the program. There are dozens of programs like this where a student will not move into highly coveted programs if they haven’t scored in the top percentile. Cheating in these gateway courses is widening the social equity divide we want so desperately to close.
We took an important step recently to try and level the playing field in exam security. It’s a small step but one that–while it initially seems counterproductive to preventing cheating–is significant because it is the right thing to do.
Our company, which is the largest proctoring company in the world, just announced that we will no longer support or sell proctoring services that rely solely on technology to flag incidences of suspected cheating. To be able to explain why that is important, it is critical to understand the difference between online monitoring and proctoring.
Online monitoring systems for exams are fully automated, meaning that they rely on AI to watch for suspected cheating. They do not have human proctors either watching the exam in real time or reviewing the exam afterwards. To reiterate, online monitoring does not include a human being in the process, people who are trained on what to look for and how to discern when someone sneezes as opposed to when a person looks repeatedly off to one side as if checking their phone for answers.
Conversely, proctoring an exam is what we see in-person when a faculty member or teaching assistant monitors an exam either in the classroom or a testing center. At our company, the proctors we employ go through six weeks of training before they can proctor exams and that comprehensive training continues throughout their tenure. Furthermore, each proctor is supervised by seasoned professionals, people who have proctored hundreds of exams and are very experienced with the process. Third-party audits add yet another layer of oversight.
Systems like ours that have humans in the process are a very different thing from online monitoring of exams. For a small portion of our business, we offered online monitoring but as stated above, we aren’t offering it anymore.
Online monitoring relies on the software technology to flag suspect events and then a faculty member, or TAs and support staff need to review those flags to determine if the suspected event is an incidence of cheating. In an ideal scenario, this could be a perfectly acceptable way to deal with suspected cases of cheating, however something very different is happening. In truth, very few of these incidents are being reviewed by faculty or other university staff.
According to our very real and applied data, less than 11 percent of the technology-only exam monitoring videos are being reviewed by faculty, and that includes those that have been flagged as being suspect. Our data is consistent with that reported by the University of Iowa stating that just 14 percent of recorded test sessions were checked.
To be fair, no teacher or college instructor got into teaching to catch cheaters
They have been busy enough, especially this year translating their courses to an online format and keeping up with the other complications wrought by COVID-19. Faculty also did not get into teaching to watch videos of their students taking tests either in real time or after the fact. Teachers want to teach. Just as troubling are the highly publicized mishaps by faculty who have gotten this wrong. It really does take a highly trained individual to recognize the signs of real cheating.
Which brings me back to equity. The reason why ProctorU is dropping exam monitoring services that do not include human oversight is twofold. The first reason is the lack of follow-through when those videos are turned over to college campuses, and second is the fact that the need for exam equity overrides the need for convenience.
The AI technology used to monitor and flag suspect actions is not well-enough refined to be left without human supervision. It may be convenient and less expensive to rely on technology to watch exams but the concerns about AI bias are significant enough that we cannot afford the risk. And, as more students are being shut out of opportunity either because of their cheating classmates or because of potential bias in AI-only monitoring, the right thing to do is to end the practice. We’re setting the standard here by eliminating this aspect of our business.
Our call to action for the rest of the exam monitoring and proctoring industry is to follow suit and adopt the principle that every exam being delivered online or via a computer always include a person who is extensively trained and certified on the proper ways to proctor an exam and has been trained to recognize and prevent their own biases from entering the process.
In getting back to my statement that cheating on exams further widens the gap for those who are already at a disadvantage by ethnic, racial, gender, socioeconomic, geographic, ability or other factors, my final point is about numbers. We have to protect the integrity of exams even if it is for one person because one person at a disadvantage is one to many.
So rather than fixate on whether cheating numbers are really increasing or whether we just are more in tune to the issue, we have to do something to protect each individual from their classmates who might be cheating. Potential solutions to reduce cheating are numerous and varied, and all worth deep consideration. Among them are less reliance on exams and demonstrating knowledge acquisition in more real and practical ways. Course redesign is also significant not only for delivering individual courses but for entire degree-granting programs.
However, there will always be a need to give an important exam and to ensure its integrity so that each individual taking the exam is on the same footing, which is essential to making quality education accessible and equitable. For our part, we have drawn a line in the sand on the only kind of exam proctoring that is acceptable given the current state of our technological understanding and anti-bias programming in artificial intelligence. And on that side of the line is the requirement that humans be a part of every exam we proctor. That’s the only kind of exam proctoring that we’ll do and we hope the industry will follow suit.
Dr. Ashley Norris is an expert in academic integrity, regulatory and accreditation compliance, and assessment. She spent nearly 15 years in higher education as both a faculty member and administrator across major institutions including University of Alabama, Samford University, and the University of Phoenix. She is currently the chief academic officer and chief compliance officer at ProctorU.