Must Read: At Least 59% of Arizona State Students Were Cheating an Online Computer Science Class

Plus, another student tells the media they were wrongly accused of using AI to cheat. Plus, a correction.

Issue 294

Subscribe below to join 3,852 other smart people who get “The Cheat Sheet.” New Issues every Tuesday and Thursday.

If you enjoy “The Cheat Sheet,” please consider joining the 18 amazing people who are chipping in a few bucks via Patreon. Or joining the 38 outstanding citizens who are now paid subscribers. Paid subscriptions start at $8 a month. Thank you!

---

An ASU Professor Nabs 59%+ of the Class Cheating, Says Some Incredible Things

I confess. I don’t know where to start with this one.

ASU On L-I-N-E

Arizona State University has invested considerable reputation and actual capital in being on the cutting edge of online education. That’s fine. But part and parcel with online education comes increased opportunities to cheat and, accordingly, increased cheating. I don’t think anyone debates this correlation anymore.

So, cut to an ASU class, CSE 365, an online computer science class described previously as the:

101 of computer security. The course will offer a basic and comprehensive understanding of the problems of information assurance (IA) and the solutions to these problems, especially the security of information on computers and networks.

In the course offered this spring, a professor recently sent a long memo to all their students, titled, “CSE 365: Academic Integrity Violation Notice.”

I am not kidding when I say you have to go read it. Or you can, of course, get the considerable highlights here.

If you stay here — or go and come back — I’ve reversed the order of the memo to get to what happened, before I get to some of the shocking things this professor had to say.

Note also that, as of this writing, I have not been able to confirm who this professor is. The document is unsigned. I have e-mailed the educator who I think wrote it, asking to confirm its authenticity and the details. I am choosing not to identify him until I can have more clarity. Although, based on what I am about to write, I don’t think he will reply.

I also e-mailed the press office at ASU, asking them to confirm this memo. They have not replied. Even without that, there’s enough supporting evidence about it, and these incidents, to believe it is, that they are, real.

The Numbers

The professor writes:

Let’s cut to the point and break down what has occurred. There are 591 students enrolled in CSE 365 this semester. We have identified serious, blatant plagirism in 6 of the course’s 9 modules, with:

• 70 students suspected of cheating on 1 module

• 72 students suspected of cheating on 2 modules

• 98 students suspected of cheating on 3 modules

• 69 students suspected of cheating on 4 modules

• 36 students suspected of cheating on 5 modules

• 6 students suspected of cheating on 6 modules

In total, this represent 351 unique students who are suspected of cheating on at least one module. This constitutes the largest identified academic integrity violation in the history of this course, and even further, the school of engineering, (possibly even the entire university?), by a large, large margin.

So, the math — 351 of 591 is 59.4%. Setting aside the scandal of a class with nearly 600 students, 59% caught cheating is enough to melt anyone’s mind. And it must raise serious questions about the quality of instruction and credentials coming from these programs, because it’s impossible that this one class was a unique, lawless outlier. It may be the largest in the school, as the professor says, but there’s zero chance it’s alone. If this section is clocking in at 60% cheating, what do you guess the other, more average, rates are? 40%? 30%?

And that assumes that the key difference in this case is the degree of cheating instead of acting on it — 60% may be average.

Lest you think the professor was being a stickler or nit-picking, they also wrote that to make these determinations, the teaching team identified learning modules in which submitted, graded code work was:

• 100% IDENTICAL to at least 4 other students in the course (at least 5 in total)

• Manually verified to be found directly online, in cheating PDF(s), in the case of 5 out of the 6 modules

• Manually verified to consistute partial or complete solutions to the module’s assignment

• Manually verified to not be easy to write the same thing by chance (e.g. many students have identical basic assembly to exit(0), this is of course not cheating)

• Often including artifacts from the PDF(s) page headers

• Often including the same (mispelled) comments In other words, this represents blatant, trivial, and egregious cheating, and is unquestionalbly a violation of the academic integrity policy

I give our professor a break for having misspelled “misspelled” and other items. It’s hard to type when you’re angry. I make mistakes even when I’m not angry.

But with those standards for making the integrity accusations, that’s pretty tight.

Reading between the lines, “found directly to be online in cheating PDF(s)” is code for Couse Hero, more than likely. At Course Hero and other sites, you can simply download answers to nearly any academic assignment. Here, for fun, is the Course Hero page for CS365 at Arizona State. Several documents are helpfully marked as “solutions available.”

59, plus

The 59% of students the professor says engaged in blatant plagiarism is not all the cheating. They write:

We have further identified many more files that are highly suspicious (e.g. 2 students having the same file), but do not meet the above criteria. Considering the nature of what has already been discovered, we have decided not to pursue these cases, but we are nevertheless aware of them. Consider this a warning, and some incredible luck, that future students will not have.

The italics on ‘many’ is original, which means that 59% is not nearly all of the students who cheated. I think we could assume that. But if 60% is your baseline of provable cheating, I don’t know, I think that’s a problem.

The professor also said that some students tried to cover their tracks:

Many students had the foresight to delete their solution code after the fact. This was the case for 80% of the identified files. Unfortunately, however, for those students, we have recovered those deleted files.

Don’t cheat in a CS class. Especially one on computer security. Come on.

Further:

Additionally, we have snapshots of all student activity in this course. That includes files you have already deleted, or might delete after reading this. If anyone is found to to be cheating in the future, we will discover it. This includes sharing your solutions with future students, or sharing them online. If you think you have a clever way to avoid detection, you’re probably wrong. Anyone identified as helping future students cheat will be pursued to the fullest extent of the academic integrity policy.

That’s a glancing reference to a file-sharing service such as Course Hero, where students share their solutions with others.

But — if you think you have a clever way to avoid detection, you’re probably wrong. Dude. Absolute fire.

Don’t cheat in a CS class.

Teachers Set the Tone - the Ugly

The cheating, I kind of expect. In fact, bad as this is, I’ve seen way worse.

As often as I try to avoid criticizing teachers for cheating, there is a throughline from the tone and culture a teacher sets, to the absorbed permission to cheat, to the cheating. That’s not always a teacher’s fault. I do understand that, as a friend of mine used to say, teachers did not become teachers to be police.

Even so, in this case — and this is entirely my opinion — the throughline between this teacher’s approach and cheating is a short distance. Working against the presumption of integrity from the outset is that, as mentioned, this class was online. Additionally, as mentioned, it is gargantuan in size. Also, engineering/computer science is among the most-cheated subjects, along with business and nursing. In other words, that’s a trifecta of known problems that should have evoked the need for more care, not less.

Yet, on this stack of problems, the teacher makes it clear that, at least in this memo, they do not care too much about much of it. They write, for example, that giving tests that require studying “removes time from what I perceive to be optimal learning.” And the professor continues:

ASU says that I must give you a grade at the end; truly I don’t care about this part, but I’m happy to be a part of the system. Probably the grade serves as a good motivator for you to do the work, probably the biggest motivator for many of you, and I’m happy to provide that. And so, your grade is 100% based on solving [online programming] challenges.

They continue:

I’m aware that this means that cheating is easier. You never have to enter a classroom, and never have to take a test in which we’re able to directly watch you to make sure you’re not cheating. The hope is that you will nevertheless be motivated to learn, and that you will be honest. Many of you are incredibly busy people, and I understand that you have a lot of things going on. If my course is easy to cheat in, and you have a lot of things going on, lots of obligations, then it’s easy to see why you might lean towards cheating if it’s the path of least resistance. I get it.

I’m sorry - what now?

Cheating is easier because there are no tests, classroom time, or supervision. True. The professor knows this but has “hope that you will nonetheless be motived to learn and that you will be honest.”

Hope, as the saying goes, is not a strategy. And the proof is in how well it worked in this case. It did not. It never does. Especially in an online, unsupervised environment.

Further, the professor says “I get it” about “why you may learn towards cheating,” because “you have a lot of things going on.”

I understand that this message was delivered in a memo about cheating, after the fact, but it’s easy for me to draw the professor’s approach on integrity issues from it. And easy to believe the students understood that the class culture on this issue was one of hope and blind empathy. In addition to the blind assessment structure.

But how I wish we were done here. The memo continues:

Unfortunately, though, my passion is in making sure you learn, and cheating gets in the way of that. You’re bypassing the learning, and going straight for that motivating grade. I truly don’t care about you unfairly getting a grade that you don’t deserve: that doesn’t really impact me. Of course, your peers, the academic institution conferring degrees, and employers trying to interpret what those degrees mean might not be thrilled about that. I do, however, care about you not learning. It is therefore my responsibility to make cheating less-appealing than learning.

I am simply speechless.

A professor in a STEM class at a major state university wrote that they “truly don’t care about you unfairly getting a grade that you don’t deserve; that does not really impact me.”

[stunned silence]

I also hate how the idea that others — peers, the school, employers — may care about cheating, but it somehow is not really this professor’s problem.

It’s a good thing I am not the Dean or the President at ASU or this conversation would be short and direct and result in a new hire.

I told you, this professor was absolutely not going to respond to me now.

I genuinely do not understand this mindset. At all. And I know it’s not an anomaly — a good number of professors feel this way. They do not feel as though cheating is their problem. I’ve heard it. I’ve written about it.

People with this approach somehow believe a magic learning nymph is going to come along somewhere and sort out the good, learning-centered students from the cheaters. Businesses will, they assume, weed out the cheaters when they can’t do the work. Sure, after the bridge falls down. And after those businesses stop hiring people with degrees from Arizona State. But that’s not the teacher’s problem.

You can tell, this infuriates me. However, I feel it is the academy’s mess to clean up. Deans and peers need to confront this — loud-mouth writers don’t have much standing. But if Deans and other professors want their houses to keep standing, they are obligated to deal with the people who cannot be bothered to care about the outcomes.

Obligated.

Teachers Set the Tone - the Good

Mixed in with this professor’s destructive not my problem attitude is some genuine compassion and perspective, which I appreciate and recognize.

They repeatedly write how an accusation of misconduct is not fatal and that things will be OK. They will be. Frankly, I think it is serious, but the understanding here is nothing but good. For example:

I expect that many of you are feeling a lot of emotions right now. Relax: this is not the end of the world. If you have cheated, you have messed up; own it, move forward, and make better decisions in the future.

The professor says that the consequences are often light for first-time offenders. And that:

More serious issues start to arise once you have multiple violations. Unfortunately, given the scale of this incident, probably some of you will now have multiple violations. Still, relax. You messed up (again), and the consequences will be more severe, and figured out with the guidance of the AIO. Just remember, you will get through this. Please use this as a learning experience and make better decisions in the future. Again, this is not the end of the world; this is your opportunity to grow from this.

AIO is Academic Integrity Office.

But, good. I really, really do like this tone in what is a terrible situation.

Teachers Set the Tone - the Not so Good

Taking this professor’s compassion and soft hand and delivering it in a backhand to the cheekbone, they also write:

If anyone attempts to guilt trip (“this is the end of the world because...”) the instructional staff, the AIO, any other ASU staff, or their peers, we will immediately escalate your case to the Dean’s Office. If anyone attempts to threaten, intimidate, harass, or otherwise retaliate against the instructional staff, the AIO, any other ASU staff, or their peers, we will immediately escalate this to the Dean’s Office and Tempe PD. Seriously, this is not the end of the world. Do not make this situation worse. I promise that you will get through this.

I get it. But that’s — I don’t know — all over the place.

Finally, I’m sharing this, from the same memo:

I will likely be instituting a curve on the final grades. Currently a C (passing for a required course) is a 70%. I will likely be lowering this to 60%. There is some chance that I will lower this further from that. I will also adjust some of the other letter grades (A+ will remain at 100%). Do not ask or push me, or the instructional staff, about this: attempting to influence or change any Academic Evaluation is a violation of the academic integrity policy. I will offer as much leniency as I reasonably can, but this is non-negotiable. If you fail, you fail; and likely at least 100 of you will fail.

I’m OK with at least 100 people failing. They did not demonstrate subject competency or mastery. They are supposed to fail. It is the school’s mandate not to certify them.

But I flag the idea that, because of massive cheating, the overall course rigor is lowered. I don’t get that. And I assume it’s going to help the one-off cheater who has earned a zero on an assignment or two. If that’s true, I don’t get it at all. This sets the tone that you can cheat, and, if you’re caught, and if anyone does anything about it, you can still pass with C, so long as you don’t get caught cheating too often.

Wow. I’d rhetorically ask what kind of program ASU is running. But I think the answer is obvious. Which, I know, is what makes it rhetorical.

Moreover, taking a step back, 60% gets you a C? Maybe less than 60%? Wow. Again.

<end0>

While this debacle happened at ASU, it’s not an ASU problem. We have a cheating problem. It’s a problem with many parents. And when they collide — online courses, no assessment security, big student enrollments, programs with supposed career riches, teachers who don’t care about cheating — what you get is this embarrassment. And embarrassment is the only word I have.

And I don’t just mean for ASU.

Another Student Claims an AI Detector Falsely Accused Them of Cheating

According to this local TV coverage from Baltimore, a high school student — her parent, actually — says that an AI detector falsely accused her of using AI on an assignment, resulting in an academic misconduct citation.

The TV headline is:

'Disciplined by a computer' - Maryland student fights AI cheating accusation

The idea that a computer has disciplined anyone is obviously false. Computers cannot, do not, do that.

The facts, as reported, are that:

GPTZero determined the paper Davis’s daughter submitted had a 90% probability of being AI generated, meaning it was written by artificial intelligence instead of a human.

The student and parent fought the accusation. The school held its ground. The principal agreed. An appeal to a superintendent confirmed the process and finding.

The other side is:

“She said she didn't do it, and I believe her,” Davis told Project Baltimore.

And:

Davis tells Project Baltimore she has to take her daughter’s word, because she cannot prove she didn’t use AI on the assignment.

So, two fast things. One, denying cheating does not mean there was no cheating. Two, it is not usually possible to prove a negative, true. But that there is no evidence at all of work or work product, is its own kind of evidence.

A not-as-fast point here is that GPTZero is terrible. I would not trust it on anything at all. I am baffled as to why any school would use it. I’ve said this since nearly the day it was created. Still, 90% likely is pretty high. Even for GTPZero.

But most importantly, it’s quite likely that a few things aren’t being said here. It’s highly unusual for a teacher or a school to initiate, and later stand by, a cheating allegation if a probability score from an AI detector is the only evidence. Everyone recommends against this. By now, every teacher and school should know better.

I am saying that there’s probably more here than we’re being told. The school, by law, cannot comment, and I suspect the school has more than just GTPZero on this case.

Department of Corrections Department

In the last Issue, I did an entire piece on the rumor that Byju was considering buying cheating company Chegg, and I linked to an article about it.

That article was from 2022.

Whoops.

In my meager defense, the article was pinging around social media last week. Two people texted me about it. But I did not check the date and that’s my fault. Sorry.

I guess the good news is that I said such a deal did not make sense to me instead of opining that a Byju/Chegg deal was likely, or promising, or something. But that’s no matter. I screwed up by using an old rumor and not checking the date.