Login
“All I did was go to a website that is designed to facilitate cheating and set up a kind of camera to see who visited it.”
That’s Garret Merriam, associate professor of philosophy at Sacramento State University, who recently caught 40 of the 96 students in his online Introduction to Ethics course cheating on a take-home final exam.
[“Girl with a Pearl Earring” by Johannes Vermeer, 1665, (left) with “The Smiling Girl” by an unknown artist, 1925, (right)]
I decided to ‘poison the well’ by uploading [to Quizlet] a copy of my final with wrong answers. (The final is 70-80 questions, all multiple choice, 5 options each.) Most of these answers were not just wrong, but obviously wrong to anyone who had paid attention in class. My thinking was that anyone who gave a sufficient number of those same answers would be exposing themselves, not only as someone who cheated by looking up the final online, but who didn’t even pay enough attention in class to notice how wrong the answers were.
When the students turned in their finals, and he noticed that many of the students had selected the “obviously wrong” answers from the planted version of the final, he had to decide how to distinguish the cheaters from those who merely made mistakes. He ended up using the following standard: if there was no more than a 1 in 100 chance that the number of matching wrong answers a student gave was a coincidence, he counted them as having cheated, as he explains:
When my students turned in their finals this semester, I compared their answers with the wrong answers from the planted test. A total of 45 questions on this semester’s final were on the planted final. (The exact questions change every semester, depending on a number of factors.) As expected, nearly all students had at least a few wrong answers that matched; statistically speaking this is likely given the number of questions. I ran a binomial analysis and found the likelihood that someone whose answers matched on 19 out of the 45 planted questions had about a 1:100 chance of doing so by coincidence. That was my (admittedly somewhat arbitrary) threshold, and anyone who matched at least that many, I suspected of cheating. (The highest match was 40 out of 45, which has a 1:10-Quintillion chance of being a coincidence.)
To my amazement, that threshold implies that 40 out of 96 students looked at and used the planted final for at least a critical mass of questions.
When he confronted those students about this, most of them admitted they had cheated; the consequences for their grades are still being determined:
I emailed these students telling them what I had done and what I found. About 2/3rds of them confessed right away or denied it at first and quickly changed their tune. The remaining third either haven’t gotten back to me yet or have insisted on their innocence. (I am considering that possibility for one student who is right ‘on the bubble’, but the rest are upwards of 1:1 billion chance, or more.)
I am in discussion with my Chair about exactly what response is appropriate for these students, but a zero on the final is the bare minimum, and an F in the class is likely for some, if not all of those who cheated.
He adds:
As you can probably imagine, this has been exceptionally stressful for me (I’m neither a forensic mathematician, nor a cop, so this work took a lot of time that I would have preferred to have spent grading final essays.)
Professor Merriam wanted to share what happened on Daily Nous to see what other people in philosophy made of the situation and the actions he took. He had discussed it a little on Twitter, and while some people were, he says, “sympathetic and supportive,” others (for example) expressed the view that what he did was itself unethical. He disagrees:
As far as I can tell, their argument seems to boil down to the claim that my actions were deceptive or dishonest. I was accused of ‘entrapment’ and ‘honey-potting.’ More than a few seemed to think that my transgression was as bad or even worse than my students’. They suggested I should have just taken the copy of my test down and left it at that. As far as I can tell most of these people are not teachers of any kind, and none of them seemed to teach philosophy, ethics, or humanities.
These charges don’t make sense to me. I did not encourage or nudge my students to cheat, I did not do anything to make such cheating more likely or easier. Quite the opposite: I tell all my students what will happen if I catch them cheating, and I gave them a comprehensive study guide for the final.
As far as Quizlet goes, all I did was go to the website that is designed to facilitate cheating and set up a kind of camera to see who visited it. I honestly do not see what is objectionable about that. My University has an academic honesty policy that explicitly says that looking at other tests without the instructor’s permission counts as cheating (Although had I know it would be this much of an issue I would have been explicit about that in my syllabus as well, rather than just linking to the policy, an oversight I plan to correct going forward.)
Though he disagrees with his critics, he “open to the possibility that I might be wrong”
Maybe (as the saying goes) I am the asshole here. But I would take that possibility a lot more seriously if that were the judgment of my immediate peers (philosophers at least, if not specifically ethicists), and even more so still if those peers could articulate an argument beyond simplistic accusations of dishonesty or ‘entrapment.’
So, I thought I would reach out to you and see if you could share this with Daily Nous readers and ask them: Am I the unethical one here?
That’s one question. But it might be more useful to consider more generally: (a) feasible cheat-deterring strategies for professors teaching large classes, (b) what professors should do when they catch their students cheating (when this is not settled by university policy), and (c) the extent to which professors should concern themselves with whether their students are cheating.
The post “Am I the unethical one?” A Philosophy Professor & His Cheating Students first appeared on Daily Nous.
AutomatED, a guide for professors about AI and related technology run by philosophy PhD Graham Clay (mentioned in the Heap of Links last month), is running a challenge to professors to submit assignments that they believe are immune to effective cheating by use of large language models.
Clay, who has explored the the AI-cheating problem in some articles at AutomatED, believes that most professors don’t grasp its severity. He recounts some feedback he received from a professor who had read about the problem:
They told me that their solution is to create assignments where students work on successive/iterative drafts, improving each one on the basis of novel instructor feedback.
Iterative drafts seem like a nice solution, at least for those fields where the core assignments are written work like papers. After all, working one-on-one with students in a tutorial setting to build relationships and give them personalized feedback is a proven way to spark strong growth.
The problem, though, is that if the student writes the first draft at home — or, more generally, unsupervised on their computer — then they could use AI tools to plagiarize it. And they could use AI tools to plagiarize the later drafts, too.
When I asserted to my internet interlocutor that they would have to make the drafting process AI-immune, they responded as follows…: Using AI to create iterative drafts would be “a lot of extra work for the students, so I don’t think it’s very likely. And even if they do that, at least they would need to learn to input the suggested changes and concepts like genre, style, organisation, and levels of revision.”…
In my view, this is a perfect example of a professor not grasping the depth of the AI plagiarism problem.
The student just needs to tell the AI tool that their first draft — which they provide to the AI tool, whether the tool created the draft or not — was met with response X from the professor.
In other words, they can give the AI tool all of the information an honest student would have, were they to be working on their second draft. The AI tool can take their description of X, along with their first draft, and create a new draft based on the first that is sensitive to X.
Not much work is required of the student, and they certainly do not need to learn how to input the suggested changes or about the relevant concepts. After all, the AI tools have been trained on countless resources concerning these very concepts and how to create text responsive to them.
This exchange indicates to me that the professor simply has not engaged with recent iterations of generative AI tools with any seriousness.
The challenge asks professors to submit assignments, from which AutomatED will select five to be completed both by LLMs like ChatGPT and by humans. The assignments will be anonymized and then graded by the professor. Check out the details here.
Mitchell (right) at the 2007 FAMB convention with former Twin Galaxies referee Todd Rogers and what appears to be a Donkey Kong cabinet with a modified joystick. (credit: David Race)
Over the years, King of Kong star Billy Mitchell has seen his world-record Donkey Kong scores stripped, partially reinstated, and endlessly litigated, both in actual court and the court of public opinion. Through it all, Mitchell has insisted that every one of his records was set on unmodified Donkey Kong arcade hardware, despite some convincing technical evidence to the contrary.
Now, new photos from a 2007 performance by Mitchell seem to show obvious modifications to the machine used to earn at least one of those scores, a fascinating new piece of evidence in the long, contentious battle over Mitchell's place in Donkey Kong score-chasing history.
The photos in question were taken at the Florida Association of Mortgage Brokers (FAMB) Convention, which hosted Mitchell as part of its "80s Arcade Night" promotion in July 2007. Mitchell claims to have achieved a score of 1,050,200 points at that event, a performance that was recognized by adjudicator Twin Galaxies as a world record at the time (but which by now would barely crack the top 30).
FreshRSS 