Stanford Faculty Vote to Allow Exam Proctoring
Plus, a little Hullabaloo, Tulane returns to paper and pen exams to thwart cheating. Plus, a TV segment in New Zealand you really need to see, or keep an eye on.
To join the 3,270 smart people who subscribe to “The Cheat Sheet,” enter your e-mail address below. New Issues every Tuesday and Thursday. It’s free:
If you enjoy “The Cheat Sheet,” please consider joining the 13 amazing people who are chipping in a few bucks a month via Patreon. Or joining the 12 outstanding citizens who are now paid subscribers! I don’t know how that works, to be honest. But I thank you!
Stanford University Faculty Vote for Exam Proctoring
The Stanford University Faculty Senate has voted to start proctoring exams, beginning next semester.
The move is very big, both at Stanford and within the academic integrity community.
As I try to summarize and give context, there are two good sources on the vote - the student paper, The Stanford Daily (SD) coverage and over at Inside Higher Ed (IHE). Though the article at The Stanford Daily does lean heavily in the direction of the students, which you may expect.
Starting backwards, the move is important for the academic integrity community because, well, because it’s Stanford. And because Stanford is one of the few schools in the country with a true Honor Code (see Issue 204 for a quick distinction). At Stanford, faculty are disallowed from being in the room during exams; students are left alone to demonstrate their integrity and report any violations among their peers. Remote exams were open book, open note and, again, entirely unsupervised. Not even a lock-down browser was allowed. At Stanford, that Honor Code had been in place for more than a century.
So, faculty leaders voting to amend it and begin supervising exams, essentially immediately, is huge. And the reasons are pretty clear. Like nearly every other school on Earth, Stanford has a cheating problem.
According to the Stanford Daily, the move to reexamine the Honor Code began in 2019:
Amid increasingly frequent reported instances of academic dishonesty
For more, here’s a short review of articles in The Cheat Sheet:
Issue 32, June of 2021 - Cheating Citations at Stanford Rose 114%
Issue 36, also June of 2021 - Stanford Grad, “I cheated.” The author of the article wrote:
Stanford’s Honor Code system is without a doubt faulty, and it’s hurting the student body.
Issue 54, September 2021 - The Daily Stanford Previews Changes to Honor Code and writes:
The proposal comes at a time of increased scrutiny of the Honor Code at Stanford. Over the past academic year, documented cheating soared across the University.
Issue 182, January of 2023 - Chegg Hires Stanford Athlete as Brand Ambassador
Issue 184, January 2023 - A Stanford Survey Shows “Scores” of Students Used ChatGPT on Their Finals
Then there’s the coverage of this most recent faculty vote, in which IHE says:
While students have grown accustomed to the freedom that unproctored exams grant them, instructors argue that they’ve taken advantage of that trust; incidents of cheating and failure to report infractions by fellow students, as the honor code dictates, are rampant, they say. Only two of the 720 honor code violations reported at Stanford between 2018 and 2020 came from students, according to the university.
[A] chemistry graduate student, favors proctoring precisely because he has witnessed so much cheating in his courses.
“No one respects the honor code in its current form—not graduate students, not faculty, not undergraduates,” he said, adding that cheating has become “part of the fabric of the university.”
You may want to read that again. Rampant. Only two of the 720 academic integrity violation reports came from students. Cheating, part of the fabric of the university.
The SD reports:
Following its vote of approval, the [Graduate School Student Group] issued a statement in support of the Faculty Senate taking action on academic integrity. The statement asked “the Faculty Senate, in this specific instance, to reclaim its authority over the Honor Code and pass this motion,” the statement read.
“Graduate students have been observing many issues with academic dishonesty and cheating – both inside the classroom in our teaching roles as teaching assitants [sic], in the national news, and even at the highest levels of this institution,” Lawrence Berg, a fourth-year chemistry Ph.D. student and member of the GSC, wrote in a statement to The Daily. “Stanford quite clearly has a problem with academic integrity.”
Stanford clearly has a problem with academic integrity. Their “trust the students” Honor Code simply was not working.
And, to repeat, this is not a Stanford problem. But it is apparent that whatever is driving the international trend of academic misconduct, the policy of disallowing assessment supervision is not the answer.
On campus, the issue is convoluted. I’ll try to boil it down.
Basically, because Stanford was under water with cheating, they set up a committee to review their policies. The committee included faculty, student and administrator stakeholders. The committee researched, debated, and struggled - as committees do.
Eventually, the committee offered a proposal to study proctoring, allowing it in limited use as part of a trial. To be enacted, the plan needed approval from several governing bodies, including the undergraduate student assembly and the faculty Senate. The student group rejected the plan, twice, essentially killing it. Students said they did not want proctoring. They did not even want to study it.
With that backdrop, the faculty Senate essentially retook control over academic policy at the University and passed the newsworthy resolution. In part, the faculty resolution says:
the current mechanisms are insufficient to ensure the academic integrity of our degree programs
Again, that’s true. And it’s clear the faculty sensed they needed to do something. It continues:
To foster a climate of academic honesty, effective learning, and fair assessment, instructors have the right to engage in reasonable proctoring of in-person exams.
Just in-person exams. So, the bare minimum.
With the faculty resolution, the student group can either reverse course and accept the committee’s proposal to study changes - the one they voted down - or get full-on proctoring, at individual teacher discretion, immediately.
Based on early reaction, there’s no indication the student group will reconsider. Their response, as represented in the SD, was a mixture of offense and horror.
The SD quoted a member of the student governance body saying the faculty vote:
disenfranchises all undergraduate students and is a very dangerous precedent to set
“I am just extremely disappointed and upset by the Faculty Senate’s decision to pass this amendment because it essentially corners us just because they disagree with us,” [the student] said after the meeting. “It also, unfortunately, demonstrates a disregard for undergraduate student voices and shared governance as it is a complete abuse of power.”
An abuse of power - sure. Plus, the student leaders really don’t want their exams proctored. One student leader told IHE:
“It creates this sort of atmosphere or environment that makes students feel like they’re cheaters and they’re not academically honest,” he said. “We don’t want students being shadowed with this looming figure over them, pressuring them.”
The same student also said:
“…. we just didn’t want anything to do with proctoring.”
There’s much more to unpack on this. But I have two final, quick things I’ll share.
One is that IHE reported about the committee that proposed the study:
The committee’s research—which included outreach to students, instructors and other institutions of higher education—indicated that there was resistance among students to the idea of proctoring, but it wasn’t universal. Just under half the students they spoke to said they were against proctoring, though members stressed they did not conduct a scientific study.
Just under half. They may not be as vocal, but some students do actually want their work protected, they want the cheating to stop. They want to achieve and be assessed, they just want it to be fair. I think it’s worthy of noting that not even the majority of students oppose proctoring, not even at Stanford - unscientifically anyway.
My final point is this argument against exam proctoring:
But students have argued that the majority of Honor Code violations do not occur in the exam room, so proctoring is irrelevant
You see the error here. As always, lack of violations does not mean lack of cheating. Usually, it means no one is watching. At Stanford, that’s literally been exactly the case for a long time. That may be about to change.
To Curtail Cheating, Tulane University Reverting to Paper Exams
The student newspaper at Tulane University (Louisiana) has a recent story about the school moving back to paper and pen exams in order to limit cheating.
The paper, by the way, is named the Tulane Hullabaloo - which is awesome.
teachers are finding new ways to adapt to online cheating. Their solution: old-school exams.
No question that, if a pen-and-paper exam is supervised, it’s more difficult to use Chegg or AI-created text to cheat.
Tulane isn’t the first to recognize the benefits of “old-school” testing in the age of digital, profit-driven misconduct. Australia’s top universities adopted a similar policy earlier this year (see Issue 180).
At Tulane, a sophomore told the paper:
At this point, I would say I’m not really using online software anymore to take any of my tests
Reporting further, the Hullabaloo wrote:
Senior Professor of Practice Tim McLean gives physical tests instead of exams on Canvas or another virtual platform. He said he writes new exams every semester to stop students from cheating in his classes, but student plagiarism has always been an issue.
“Even before Covid, students were using phones, using the internet to complete homework assignments,” McLean said, “When there’s a disconnect and you’re not in the classroom setting it does provide some ease to pull up things and cheat.”
I confess, I am not sure what a Professor of Practice teaches. But Professor McLean deserves an award. Writing new exams, recognizing that online settings are more susceptible to cheating, even acknowledging that plagiarism happens - good stuff.
But back to the pen-and-paper, supervised exams, the same student quoted earlier also told The Hullabaloo (still loving it) that he’d prefer digital exams - and not for the ease of cheating. He said:
he thinks it is easier to take a test online. In-person exams can cause test anxiety, he said, and pressure in testing rooms “might have an effect on your ability to do well because you’re nervous for yourself.”
That’s legit. Some students do prefer the ease and comfort of testing remotely, at home, in their bathrobes or what have you. Balancing that accommodation and preference with fair exam security, that’s the challenge.
Keep an Eye on This Story on AI
The public broadcaster in New Zealand ran a story recently on the academic integrity technology “arms race” between AI and the efforts to detect AI.
At the headline level, it’s pretty standard. AI is getting better, detection systems are lagging behind, the threat is serious. We know.
The segment itself is pretty good but the first few minutes are worth watching as a professor demonstrates relatively new AI technology that can modify where a camera shows his eyes are focused - allowing him to read from his screen while the AI makes it seem like he’s looking straight into a camera.
The technology can mask a student checking their notes or using their cell phones while appearing to keep eye contact with the web camera or keyboard, and it presents a new wrinkle for remote exam proctors who rely on being able to see where a student is looking.
The TV segment also mentions schools in NZ that have decided to proactively shut off AI detection systems, for reasons I cannot fathom.
Concerns about inaccuracy, they say. Though, if they are not using it, they cannot possibly know that, number one. And number two, should they exist, inaccuracies are no problem at all when the detection is used correctly. When a school says they don’t want to use AI-detection or plagiarism detection tools because they’re worried about false accusations, it tells me they think the detection systems should decide, not their educators. That’s not how these things are designed to work.
And finally, I will never wrap my head around the idea of an institution of higher education deciding they do not want information - they’d rather not know things. “No, we’re not sure how to handle this information, what it means, so we’re going to block it from the education conversation,” is not an enlightened view.
Anyway, the segment is decent but the first part - the eye gaze manipulation with AI - is absolutely worth keeping an eye on.
I’ll show myself out.