The AI Dilemma at Colleges

2,062 Views | 7 Replies | Last: 10 days ago by PLSQL_BEAR
Oldbear83
How long do you want to ignore this user?
We had a bit of drama over the weekend. My daughter is a Senior at a University I won't name here because it's not relevant. She has maintained a 3.7 GPA and had no real issues other than changing her major five times over the last few years, which has cost her a couple semesters in time and me some extra money.

She and several other students had not received a grade on a major paper, and she asked the professor by email if she could get her grade. We heard nothing until Sunday, when she got an email back advising her that

a) the professor was giving her a 0 for the paper (!);
b) he believed she had used AI to write some or all of her paper (!!), and
c) accordingly he was reporting her to the Academic Integrity Committee on suspicion of plagiarism (!!!)

She and the whole family were astounded, and very soon got very angry. My daughter got straight A's all through school and was a NHS member. The Dean had said several times he hoped she would continue to get her Masters, and she was scheduled to take a trip to South Korea as part of her College's delegation this summer.

She has never had a whisper of cheating in her life and would never do so now.

What also astounded us, was that the professor had not bothered to talk to the Dean or other professors about her, and did not ask her about how she had created her paper (the easiest and most effective way to find out if someone really did their work). So I helped my daughter calm down, then send a response email to the professor including those points and copying the Dean. My daughter also included a screen shot from TurnItIn, a program which is used to check for whether a phrase or passage is online somewhere. The program came back with a 0% similarity score.

Fortunately, Monday morning the professor responded with an apology. He accepted her defense - partly I think because he realized he should have done more to confirm the authenticity of her work, and partly because he did not include the Dean from the start, which one really ought to do with something this serious. He apologized for his choice of words, for jumping to conclusions, and issued a good grade for her paper.

In further discussion, the professor mentioned that improper use of AI by students is a real thing now. There are services which will write a paper for you - you give them the assignment and they come back with a paper. I guess that is the modern equivalent of having a nerd friend write papers for you as some people did back in my day. But just as a professor could tell by writing style if someone likely had 'help' with their paper back in my day, AI today is still a bit vague and general, so a critical eye can catch it. In my daughter's case, she admitted she was up late working on the paper, so some of her work had less specific detail then she would normally include. But as I said, I still think that the professor should have had a conversation with her about how she made the points she did - people reveal quickly enough if they know the subject, while a faker would be revealed soon enough.

As I said, my daughter's case was quickly resolved, although I understand there are some other students who may have a harder time proving their innocence. But this reminds me as well, about an unexpected threat from AI. We have heard of deep fakes, where images and even voices can be artificially created, but now we see the danger of fake academic work. I think it's going to get harder for students to prove original work, in a world where so much can be created out of nothing for a fee.

What do you think?



TechDawgMc
How long do you want to ignore this user?
AI is creating some challenges in the classroom. It's still not as good as some suggest though and it's often terrible at summarizing. A lot depends on asking a question that doesn't suit the AI.

As far as the prof, he definitely took the wrong approach. You always confront the student first.

I had a test where two Chinese girls sat next to each other and turned in identical essays. I gave them both a zero and told them to come see me. They told me they had studied together and memorized the answers. I handed them a piece of paper and told them to do it again. They wrote the same answerword for word. I restored their grade.
El Oso
How long do you want to ignore this user?
In research papers it is easy to spot. AI has a terrible time correctly inserting research. I've also caught it because the fact was demonstrably untrue and/or overexaggerated. I once had a student write America leads the world in gun deaths. The sentence caught my attention and then other things started catching my attention and before long it became clear the student wasn't the author.

Unresearched papers are much trickier. On day one, I have my students write a two page essay which gives me a baseline of what they can do. Under strict monitoring I can guarantee the internet was not used. This is the first thing I check when I have a question. Sometimes it helps, sometimes it doesn't.

Not all students are honest when confronted and some have great poker faces making a confrontation also a difficult indicator. I try not to accuse a student unless I am nearly positive I am right. False accusations are incredibly dangerous to student teacher relationships.

There are two AI checkers out there I know of. To me, they are inaccurate at best, but at least one teacher I work with swears by the one she uses even though I have an example of the program flagging something I wrote as AI generated. She says I'm lying about it.

This is an exploding area right now and there are no easy answers. I think your daughter's professor overreacted, but I could also make the case that he/she started the process correctly. I give you the grade my gut says, and my gut says you plagiarized. After that, I think the professor deviated from what I think is the best course.

Once I accuse you, you come in and we talk. If you change my mind, I change your grade. If I am unconvinced, I follow university procedures and due process will lead us to the official answer.
Oldbear83
How long do you want to ignore this user?
El Oso said:

In research papers it is easy to spot. AI has a terrible time correctly inserting research. I've also caught it because the fact was demonstrably untrue and/or overexaggerated. I once had a student write America leads the world in gun deaths. The sentence caught my attention and then other things started catching my attention and before long it became clear the student wasn't the author.

Unresearched papers are much trickier. On day one, I have my students write a two page essay which gives me a baseline of what they can do. Under strict monitoring I can guarantee the internet was not used. This is the first thing I check when I have a question. Sometimes it helps, sometimes it doesn't.

Not all students are honest when confronted and some have great poker faces making a confrontation also a difficult indicator. I try not to accuse a student unless I am nearly positive I am right. False accusations are incredibly dangerous to student teacher relationships.

There are two AI checkers out there I know of. To me, they are inaccurate at best, but at least one teacher I work with swears by the one she uses even though I have an example of the program flagging something I wrote as AI generated. She says I'm lying about it.

This is an exploding area right now and there are no easy answers. I think your daughter's professor overreacted, but I could also make the case that he/she started the process correctly. I give you the grade my gut says, and my gut says you plagiarized. After that, I think the professor deviated from what I think is the best course.

Once I accuse you, you come in and we talk. If you change my mind, I change your grade. If I am unconvinced, I follow university procedures and due process will lead us to the official answer.
Thanks El Oso. From her discussion with the professor, I think this was his first attempt to follow a new policy, and he did not think it through. He relied on one tool but nothing else at first.

I'm glad he listened to my daughter's defense, because a Plagiarism charge is a very serious matter for everyone. After I saw his initial email, one thing I did was look up the issue, and while it's hard to get detailed statistics, I did find a lot of disparagement cases where schools had to pay lots of money.

So professors find themselves in a place where they have to be very careful in supporting claims of this type, or they could create severe liability for their school. Ironically however, students also have to be careful to provide full support for their work, because even those students who won defamation money from schools usually did so as a settlement, with the school not admitting they were wrong. It's a landscape where a student can be accused on the basis of a faceless program, forced to prove their innocence or have years of hard work destroyed because of one accusation.

I admit I am biased here, not only as a father but also because in my work I have sometimes had to deal with sensitive allegations made by one employee against another. It's tricky when you want to get to the truth while showing professional respect to everyone involved. My best tool has always been to talk privately with people, because a lie is very hard to support for long. That, and you can read people face to face in a way you can't from just their writing.

Redbrickbear
How long do you want to ignore this user?
williamwill
How long do you want to ignore this user?
The pressure of dissertation writing was intense, but https://mycustomessay.com/custom-dissertation.html provided the perfect solution. Their expertise in crafting a comprehensive, custom dissertation was invaluable. For anyone struggling with their dissertation, this service is a game-changer.
DennisRice
How long do you want to ignore this user?

AI-powered technologies never cease to astound with their capacity to craft unique and customized encounters. To demonstrate just how far these systems have come, TopAIInfluencers.io features an https://topaiinfluencers.io/hentai-ai-chats/ AI hentai chat. A one-of-a-kind platform is provided for users to interact with characters in a way that feels lively and interactive. Its innovative and user-friendly technology makes it accessible to a wide audience. If you're into creative AI interactions, you should totally check this out!





PLSQL_BEAR
How long do you want to ignore this user?
Sorry to hear about your daughter's false accusation and I am glad it got resolved fairly and quickly.

I teach on-line courses (Computer Science) as an adjunct for a local university. When students submit research papers, we have a third party application which analyzes the paper and gives a percent chance the paper was written by an AI. They claim to only have false positives 1% of the time. Even without the third party help, it is fairly obvious when a student uses AI to write their paper.

I had one student who used AI for his midterm paper. I then put the midterm test questions into ChatGPT and saw how closely their answers matched the AI results. (It is scary how accurately and quickly the AI can produce quality answers to computer science questions.) If colleges don't get a handle on preventing students from using AI to do their work, the value and integrity of having a degree will suffer.
Refresh
Page 1 of 1
 
×
subscribe Verify your student status
See Subscription Benefits
Trial only available to users who have never subscribed or participated in a previous trial.