Cheating Has Become Normal
Plus, a school district blocks Grammarly? Plus, more from Texas A&M.
Issue 322
Subscribe below to join 4,179 other smart people who get “The Cheat Sheet.” New Issues every Tuesday and Thursday.
If you enjoy “The Cheat Sheet,” please consider joining the 16 amazing people who are chipping in a few bucks via Patreon. Or joining the 42 outstanding citizens who are now paid subscribers. Paid subscriptions start at $8 a month. Thank you!
Chronicle: Cheating Has Become Normal
The headline — Cheating Has Become Normal — is not mine, it’s the headline on a really long article in The Chronicle of Higher Education.
It’s a must-read. Seriously. It’s good and necessary.
My only pick is that it quotes a study supposedly showing that cheating has not increased in the AI era, without pointing out why it’s probably wrong (see Issue 261). But that’s nothing. The piece is gold. Five stars.
It’s too long to go point-by-point here so I will share a few highlights and say again — go read it.
Unique Assignments. An Army of Cheating.
The piece centers around Middlebury College, which we’ve covered (see Issue 204 or Issue 312), but it opens with a professor Clukey at the University of Louisville who says that upon returning for fall she was prepared that her students “might cheat” with AI. She was going to do all the stuff:
I was always, like, I’ll create unique assignments and they will be somewhat plagiarism-proof
But, The Chronicle reports:
“I was just hit,” she said, “by a student army of cheating.” Students cheated on informal discussion-board prompts. They cheated on essays. A few weeks ago, she emailed a student to say that she knew the student had cheated on a minor assignment with AI and if she did it again, she would fail the course. Clukey also noted there were several missed assignments. The student replied to “sincerely apologize,” said she was “committed to getting back on track,” and that she regretted “any disruption [her] absence or incomplete work may have caused in the course.” But her next paper was essentially written by artificial intelligence. Curious, Clukey asked ChatGPT to write an email apologizing to a professor for plagiarism and missed work.
“And what did it do?” she said. “It spit out an email almost exactly like the one I had gotten.”
A student army of cheating.
And we are — I kid you not — just five paragraphs in.
Pervasive. And Honor Codes.
Here’s the start of paragraph six:
Talk to professors in writing-intensive courses, particularly those teaching introductory or general-education classes, and it sounds as if AI abuse has become pervasive.
It may sound as if AI abuse has become pervasive because AI abuse has become pervasive. Occam’s razor and all that.
Citing the reasons for what appears to be (and probably is) a tidal wave of cheating, the article says:
Students are working long hours while taking full course loads. They doubt their ability to perform well. They arrive at college with weak reading and study skills. They don’t value the assignments they’re given. They feel like the only way they can succeed is to be perfect. They believe they will not be punished — or not punished harshly — if caught. And many, it seems, don’t feel particularly guilty about it.
I’m delighted to see the lack of punishment and lack of guilt listed. As I also am to see the point about students not valuing the assignments. Theme alert. I wish that ‘cheating is just easier than doing the work’ had made the list. But still — fantastic. Cheating is not always induced by stress and unreasonable demands.
The piece continues that some schools such as Stanford University and Middlebury are reconsidering their honor codes “because they’re simply not working.”
True, they are not. They do not.
For Middlebury specifically, The Chronicle reports:
An interim report released in May by an honor-code-review committee found that, while campus culture generally affirms the value of academic integrity and that cheating is wrong, “the reality of daily practice suggests that the honor code has ceased to be a meaningful element of learning and living at Middlebury for most students.”
The article says many professors at Middlebury, “want to proctor exams.”
Continuing on Middlebury’s report, it says:
Other reasons students say they cheat include confusion over what their professors define as cheating; the ease of cheating through phones, AI, and unproctored exams; and the pressure to cheat when you see your classmates doing so.
We know that seeing others cheat incentivizes cheating. But it’s good to see “ease of cheating” and “unproctored exams” make the coverage list.
Easy. No Punishment.
The article goes on to cite the amazing 2023 essay from a Middlebury student (see Issue 204), but quotes the student now. She says, in part:
“the sense Middlebury doesn’t care if it’s allowing students to cheat or making it easy for them to,” simply fuels the cheating culture.
I have said this a thousand times and will keep saying it until schools and others start taking academic fraud seriously — when you do not actively and firmly police cheating, you incentivize cheating.
The Chronicle quotes another Middlebury student about cheating, this student sat on the committee reviewing the school’s cheating and honor code issues. She told The Chronicle:
“It’s a mix of ease, convenience, and, frankly, lack of punishment,” said Wight, who was also the student chair of the campus judicial board.
Ease of cheating and lack of punishment. I’m scratching my head because I think I’ve heard that before. Can’t quite place it.
I am sure it will come to me.
Strenuously.
The Middlebury report recommends:
that Middlebury “publicly and regularly define and affirm the meaning and importance of academic integrity.” As part of that affirmation, it encourages professors to clarify, “in each semester and in every course, the importance of academic integrity, what it means, and why it matters.”
The italics is in the story — not sure if it’s in the report or planned for the new policy.
Either way, let me say that this is just perfect. The honor code is not working so, we’re going to say it more often. Five Hail Mary’s didn’t turn the trick? Let’s do ten.
I’m reminded of that scene from A Few Good Men where the Tom Cruise/Demi Moore/Kevin Pollack defense team objects to something in the climactic trial. The judge overrules their objection but Demi Moore then stands up and says “I strenuously object.” The judge, again, overrules.
That’s Middlebury. But wait - this time we’re going to do the honor code in every class. And, and … it’s also in italics.
Oh, and:
as part of its call for the campus to “publicly and regularly” affirm the value of academic integrity, it recommended reinstating an honor-code signing ceremony during orientation
Sure. Twenty! Twenty Hail Mary’s.
Some Pain-in-the-butt Thing.
I am also eager to share this quote, from a professor at Middlebury:
“What’s interesting to me is how much they love the honor code and would be very sad if it was repealed,” she said. Students like the tradition and the assumption that they will behave with integrity. “But on the other hand, they also acknowledge that there is not a culture of social sanctions against cheating. One student was not embarrassed to tell me at a dinner in front of 10 other students, ‘Of course I’m going to cheat’ in some gen-ed, or what we call distribution requirements, class. ‘You know, some pain-in-the-butt thing that I don’t see as particularly relevant.’
Yes. Of course students love the honor code. It makes them feel like trusted adults.
Problem is, feeling like adults gives the impression that they can pick and choose what they want to do to, what they feel is important and what is not. And when they tell themselves an assignment or course is not necessary, not important, not relevant, or have some other excuse such as the teacher is bad, or that they’re being picked on — they feel empowered to cheat. They’re actually proud of making those choices. It’s not cheating, it’s making an adult choice to not do stupid stuff.
If you pay attention to academic misconduct long enough, you see this rationalization everywhere.
If schools and teachers want teaching moments related to integrity choices, here’s one. When you get a job, you won’t get to decide what work is worth doing and what you think is a “pain-in-the-butt thing that I don’t see as particularly relevant.”
Yes. And no.
I’ve taken myself far afield. Back to The Chronicle piece. It continues:
Little of what’s happening at Middlebury or elsewhere is surprising to people who study academic integrity. Researchers have long documented that many students cheat at some point in their educational career, and that their motivations are situational rather than character based.
True. And not so true at the same time. Middlebury, not surprising. But recent research suggests a very large role of character-based traits in the cheating decision tree (see Issue 149).
Not a big deal. Moving on. From the article:
The reasons they cheat have also remained relatively consistent over the years. The shifts have been in the pressures that lead students to cheat and the ease with which they can do so. Practically speaking, that means that while some colleges used to see higher incidences of cheating in STEM courses with high-stakes exams, grading on the curve, and competition among students to get the best grade, generative AI’s arrival has made it increasingly common for students to cheat on low-stakes assignments, like discussion-board posts and short essays.
Again, yes and no. There’s been very good evidence for a long time that low value assignments such as those described above are prime targets for cheating because the value, and therefore the risks, are low. If an assignment is worth practically nothing, students assume it’s worth practically nothing and see little point in putting in effort. It’s just “some pain-in-the-butt thing.”
Still, not a big deal. Keeping the feet moving.
The Chronicle is fuzzy on whether cheating is worse now than in years past, citing the study I mentioned up top. But, they report:
Still, data is cold comfort for faculty members faced with conspicuous cheating in their classes, and students who witness classmates cheat without consequence.
Experts. What a Concept.
If you’ve read The Cheat Sheet for any time at all, you know that one of my top complaints about news coverage of academic integrity is that reporters don’t speak to actual experts on the subject. This piece, The Chronicle piece, interviews two and cites a third. Amazing. I say again — go read it.
One expert is Tricia Bertram Gallant, a friend of the newsletter and friend in general. And while I strongly disagree with her over the idea that professors should not focus on catching cheaters, which The Chronicle attributes to her, she is quoted well:
“I have faculty say that, ‘Well, I just want to trust students. We should be able to trust students.’” said Bertram Gallant. “And I say, ‘You can trust that they’re human beings and they’re going to make bad decisions under stress and pressure. That’s what you can trust.’”
The piece also quotes Talia Waltzer, “a postdoctoral scholar in the department of psychology at UC San Diego, [who] has dug into why students cheat.” She says that in her research, students don’t always recognize they are cheating.
We’ve seen that before in studies in which students don’t acknowledge cheating but admit to specific activities that are typically considered misconduct. When asked to explain the disconnect, students usually say, “well, when I copied the answers, that was not cheating.”
Walzer says:
When students were aware that they were cheating, they would sometimes explain that they thought it was OK within the particular situation. “In a lot of these cases,” Waltzer said, “students are talking about assignments that they perceive as not being valuable for their career and learning goals, not advancing their skills, their long-term goals. It might be for a general-education requirement. It might be a class assignment that they would describe as busy work. They don’t see the point in doing this work.”
I feel like we are — this close — to having a theme.
The Chronicle covers another Walzer study on how teachers address integrity in their classes, which I think is really important. So, please go read the article. But if I can, I will get to it in a future Issue of The Cheat Sheet.
Doing it Right. Still Not Working.
Moving along, on the topic of professors addressing cheating and authentic learning motivations with their charges, The Chronicle shares:
It’s not, of course, as if some professors aren’t trying those things. Clukey, the associate professor at the University of Louisville, for example, spent an entire class period talking to her students about why it’s important not to cheat. She told them that even if they don’t like literature, they are learning transferable skills, such as how to write and analyze language. She explained how you could theoretically cheat your way through college, but what would happen in graduate school or the work force if you had never developed any skills? She spoke about the importance of trust in the classroom and the impact poor preparation has on class discussions.
She doesn’t really know if her talk had an impact. Students nodded along, but a few wrote in their anonymous mid-semester evaluations that spending a whole class period on academic integrity was a waste of time. “And of course, when I went to grade their papers,” she said, “a bunch of them were plagiarized.”
I’m glad this is in here because I had a full three paragraphs written about this effort to reprogram the motivations of students and how it was futile. How making content ‘more engaging’ was a dead end. I deleted them as too far off topic and at least 40% too preachy. It’s better that a professor share that it does not work, at least not at any discernable scale. And it’s why — to me — any such effort must be paired with enforcement and penalty. You may want them to want to stay on the path, but there have to be consequences for leaving it.
On cue:
Clukey is also tired of well-meaning advice to make her assignments more interesting and relevant to students’ lives.
No joke.
Turning Them In.
The Chronicle says professor Clukey has started turning students in for cheating. But notes also that:
Still, she knows that few of her colleagues are turning students in. Instead they are just giving zeros.
Few things irk me more than this. Teachers are eager to recognize how self-defeating it is when students cheat, but are blind to their own self-defeating solutions.
In case I am being too obtuse, a zero is no penalty. If you have not done the work, if you choose not to do the work, a zero is what you’d get anyway. Logically, cheating then becomes all potential reward, no actual risk.
From the article again:
Other instructors have also reached their limit with cheating and believe that enforcement of academic-integrity policies, as byzantine as the process can be on many campuses, may be the only way forward.
Yes, with few if any exceptions, the official academic integrity process at any given school is horrible. It’s stacked with disincentives for educators and the consequences are frequently laughably lenient, another disincentive to act.
Continuing:
The honor-code-review committee at Middlebury, for example, acknowledged that few faculty members report students because they find the process “exhausting and time-consuming, as well as demoralizing when students are found not guilty.”
Ding, ding, ding!
Do me a favor. Jump over to Issue 320 for the student who admitted to using AI on her assignment but was “cleared” because the review panel decided “there wasn’t enough evidence.”
Dark Humor Break.
Our article quotes a different professor at Middlebury who:
has also given up on using discussion boards, because she realized that she was wasting time responding to AI-generated posts. “I know people who have just decided, ‘My time is really important, I don’t care.’ So they ChatGPT their reply to the student. And so we’ve got robots talking to robots. And that, to me, is the end of higher education right there. That’s the apocalypse.”
Just picture me slowly shaking my head.
Break Over.
And I have no idea how far in to this I am — trust me, my version is still shorter — but I think this may be among the most important parts of the reporting:
Faculty members may be on the front lines of the battle against cheating, but they say it can’t be theirs to fight alone. They want support from administrators to help deal with the reasons why students cheat, and backing from their college if they report students for academic-integrity violations.
It boils my blood when teachers don’t take integrity seriously or refuse to be the adult and enforce what “required” means. But it’s a Xanax compared to what I feel about administrators who craft policies uniquely designed to make finding and penalizing misconduct nearly impossible while swearing how they “take academic integrity very seriously.”
Cheating in academia is an existential threat. And it is not addressable in any significant way unless administrators get serious. Universally, they are not.
And. In Conclusion.
I must also share:
Nor do honor codes hold any special power. The committees at both Stanford and Middlebury, for example, found that the codes had little impact on student behavior.
Maybe they just haven’t tried it if they make the code font a little larger or have the students say it in Latin once a week.
The article ends with a quote from Wight, the Middlebury student who was helping on the honor policy reviews:
One thing did become clear to her, though. “A big part of integrity,” she said, “is if the school doesn’t take it seriously, the students won’t either.”
I mean — nothing more can be said.
More on Honor Codes, This One at Texas A&M
A friend of The Cheat Sheet sent us this post at Reddit:
It’s Reddit, so caveat emptor. And in case you cannot read it, it says:
A mass email was sent out to one of my public health courses saying that more than 50 students violated the Aggie honor code and that my prof is reporting them. I haven’t done anything to break it but I’m just scared because we turn in alot of typed assignments and I have heard horror stories about false AI convictions. I have 2 questions: #1 How soon will I find out if one of the students is me? #2 How easy is it to appeal a violation?
I share it because, like Middlebury and Stanford, Texas A&M has an Honor Code in which the primary plank is trust — the school trusts students not to cheat. In 2020, Texas A&M faced front page news of “large scale” cheating with — guess who? — Chegg. In other words, it’s clearly not working.
I’ll also point out that this author says they have heard “horror stories about false AI convictions.” You may have heard the stories, but they are just that — stories. Actual incidents of “false convictions” in any form of academic misconduct inquiry are as rare as actual headless horsemen.
Grammarly Blocked on District Devices?
You may know I am suspicious of posts on social media as credible sources. My default is heavy skepticism, especially if names — in this case, schools — are not included.
Nonetheless, I’m passing this recent LinkedIn post along as it claims to show notification from a school district to students blocking access to Grammarly on district devices and networks:
It’s suspicious that the person who posted it intentionally cut the name off, making it unverifiable. That’s especially the case since she wrote:
Can you believe it’s almost 2025, and they’re moving backwards by blocking all AI tools?! Our job is to prepare students for their future, and I can assure you this is a skill they’ll need to succeed.
In other words, it’s entirely possible that this was invented just so someone could be outraged.
But, on the off chance that it’s true — good.
As we’ve shared now several times, Grammarly is a full-on AI writer (see Issue 315). Grammarly is advertising “Write My Essay for Me” services. And it’s time that education leaders got serious about drawing lines of accepted use of these tools. If a school has decided that it’s not wise to permit students to access a paid essay writing service on their networks, I am fine with it.
As for the “I can assure you this is a skill they’ll need to succeed” performance, I rhetorically ask — what skill is that, exactly? Paying someone else to do their work?
If that’s the answer, please also answer what company is going to hire someone whose skill set is outsourcing their work? Why not just outsource the work directly? If ChatGPT or Grammarly can do your work, ChatGPT or Grammarly will do your work.
Schools, in my view, should be teaching students how to communicate and write without AI, on their own. I think that will be the employable skill — not chatbot cut and paste.
Whatever. Bottom line is that Grammarly is no longer a valid and ethical education tool. If schools are blocking it, they are right to do so.
Class Note:
I’m seeking another volunteer or two to proofread and gently edit The Cheat Sheet. All that’s needed is a good eye and maybe 20 minutes, twice a week, on Tuesday and Thursday. If you can help, please raise a digital hand. A reply e-mail reaches me.
Thank you.