On the syllabus of an elective college course, I have stated that a student who plagiarizes once will receive a zero for the assignment. A student who plagiarizes twice will receive zeros for the assignments and may withdraw or take an F for the course. So, a student in an elective course submitt...
she had submitted the journal after the meeting in which I'd agreed to let her rewrite her essay -- should this say "before"? I imagine the "excuse" was that she submitted the second one before realizing that she'd been caught on the first.
Why did you allow the student to submit a new essay? It seems the rules are clear, and you somehow let the student get away with breaking them. The only thing this teaches the student is that people are pushovers, and that complaining works. How will that prepare the student for a real job?
@JackAidley If the AI generated text is good I'd argue it's demonstrating competence and readiness for future jobs, while all the students who write it themselves are behind the curve, much like slide-rule or sextant users. You have to admire their prowess with obsolete tech but they'll be next to useless in the workplace.
@JackAidley presenting work you did not do as your own is plagiarism. Whether you are plagiarizing a human or a machine is irrelevant, if you pretend you wrote it but didn't, then you are committing plagiarism.
@terdon she did do 'work', less work as expected but she did apply and inquire the tooling herself and (hopefully) vetted the response to fit the assignment.
@Sim nevertheless, she presented work as hers which was not hers. That's plagiarism. The assignment wasn't "edit someone else's work", it was "write something" and she did not do that.
@Peter-ReinstateMonica, wait, are you suggesting slide rules are obsolete?!? In all seriousness, I do plan on teaching my children (now 2 and 5) how to use them, as I believe it's a good lesson in both math history and how logarithms work. I've already demonstrated how I can use it to multiply simple numbers to my five-year-old, since he's capable of doing easy multiplication, though he's not yet proficient in logarithms. ;)
@terdon Does that apply to tools that spell check, as well? Should you cite those tools? I ask, because I do use ChatGPT to help me write professionally. For example, I'll ask it open-ended questions sometimes to help me generate ideas (I always verify everything it says, of course), and I'll use it to help me rewrite something less awkwardly or more simply. I don't cut-and-paste from what it provides me, but I'll sometimes use whole sentences pretty much unaltered, which would definitely be plagiarism if I were copying from another human author. NB: I have not done so on submitted papers.
@BenHocking I admittedly had my doubts when I wrote it; proficient slide rule users can be faster than calculator users and make fewer errors, at least certain classes of errors, because they are aware of the orders of magnitude involved. Still, not many people use them any longer. Like with manual matrix solving (and navigating by the stars!), it is useful to have done it a few times and understand it, no doubt.
@terdon To be clear, of course there's a difference of scale between having a spell checker (and grammar checker) suggest small changes to your writing vs having ChatGPT (or similar) give you whole sentences, paragraphs, or longer, but where one draws the line is not obvious. Is using ChatGPT at all cheating? Is it only cheating if you're copying more than 3+ consecutive words? Is it acceptable if you cite it? The answers are not obvious, so should be part of professors' policies going forward, I would suggest.
@BenHocking you said professionally, this is about a student. There is no suggestion that they did anything other than submit chatGPT-produced content as their own work, and that is plagiarism by definition. What you describe is a different beast entirely.
@terdon In this particular case, if the whole assignment (or even a sizeable portion of it) was written by an AI, then I would agree it's cheating. Whether it's* plagiarism is trickier to answer, but not necessarily 100% necessary to answer, unless the professor's school has strict requirements to be precise. Assuming it is plagiarism, though, wouldn't that still apply to using a whole sentence from an AI? Would it also apply to a simpler tool that suggested multi-word (e.g., 4 words) replacements? *Apropos of my prior comment, my built-in grammar checker wants me to change this to "its". :P
You shouldn't be operating in a vacuum. Find your school's academic dishonesty policy, and follow it. I don't know your school's policy, but just about everything in your original post falls outside of what faculty are allowed to do.
@ScottSeidman While I agree with your first sentence, what he describes in his original post (or original to when I saw it) is definitely within scope of what faculty are allowed to do at many, if not most, universities in the US. At UVA, I think his policy is probably against the honor code, as no second chances are allowed as I understand it, so I'm not sure if you think his policy is too lenient or too strict.
@DanielHatton that's an incorrect assumption. The AI doesn't keep a copy of the training data, thus it cannot plagiarize that. It's more complicated than that but the language model only keeps word frequency information inside its neuron weights. If the model could regenerate its training data, it would be an impressive compression algorithm on top of doing its original functions as the size of the trained model is orders of magnitude smaller than the totality of the training data.
@BenHocking Frankly, I'm fairly sure you're not right. Most campuses specify some form of review to sanction, and take it out of the profs hands to some extent or other. I've spot-checked UVA, Penn State, UCLA, Johns Hopkins, and my own school, U of R -- the requirement for some form of reporting and review is present in all of them, in every school I checked. If you can find one that doesn't require that, let me know. It's not the harshness or laxity of the sanction I was critical of. It's the prof taking action like failing a student for dishonesty without review I have issues with.
FWIW, U's are reticent to do anything a student may sue over without showing an official policy in place, and demonstrably following those policies and procedures. For my own case, the offer to withdraw would also be off the table. Students here may not withdraw once accused.
@ScottSeidman You are correct that for an actual honor code violation, UVA requires official submission to the honor council (or something like that). However, I know from experience as a TA*, that professors have been free in practice to give zeros without official submission. I'll admit that I do not know exactly whether such acts are officially condoned, but IME such policies are the norm, when such policies are announced. *I was also an adjunct professor at UVA for a year, but fortunately I did not have to implement such a process.
@ScottSeidman, but +1 to your prior comment because admittedly, I was basing what I wrote off what I experienced and what I've heard others say without actually checking policies.
@BenHocking -- schools have a real need to follow such stuff. a prof dealing with honesty violations has no idea if a student had a clean record, or has been accused fifty times. In the first case, pedagogy demands that the student has an opportunity to learn from the mistake, and in the second, a school should be thinking of denying that student a degree. If the profs handle things in a vacuum, there's just no way to know.
@ScottSeidman I don't disagree, but UVA's honor policy is that there is no second chance, should they be found guilty of cheating (or other honor violations). As I understand it (this has been reported in the UVA magazine), there are concerns that the honor council is biased, giving certain classes of students more leeway than others, but officially (i.e., not in practice) the rules are fairly clear cut. (FWIW, the history behind the honor system at UVA is quite interesting.)
Bias is certainly a factor, and the numbers in that article are frightening. That said, the latest honor code incarnation at UVA looks like there are second chances: uvamagazine.org/articles/…. "The Honor Committee gains broad discretion in fashioning penalties, starting with rehabilitative education and having offenders make amends, escalating to temporary or permanent removal from the University community."
Discussion on question by Dan: What s…
Discussion on question by Dan: What s…