News:

Welcome to the new (and now only) Fora!

Main Menu

Paraphrasing tools are getting more sophisticated

Started by downer, October 27, 2021, 10:47:32 AM

Previous topic - Next topic

Kron3007

Quote from: ergative on October 27, 2021, 01:23:14 PM
Quote from: Kron3007 on October 27, 2021, 12:42:54 PM
Quote from: downer on October 27, 2021, 11:09:24 AM
A court? What?

My course, my rules.

Not where I teach.  For any academic misconduct situation, we cannot penalize, only report them and it gets dealt with at a higher level. 

I love this since it means I dont have to deal with the fall out, just pas along the evidence and wait for the outcome.  It is also good for the university because it helps identify patterns and trouble students, as well as ensuring punishments are equitable etc.  I'm surprised that you have this much leeway.

I also have this system. It's great. The only thing I don't love is having the student conduct board send me an email a week after grading is done telling me that I need to disregard the plagiarized portion, grade what's unplagiarized, and then return the grade which will then be reduced by X amount as a penalty. I usually agree with the decision (grade reductions for minor first offenses seem fine to me--no need to fail the whole assignment when there's a centralized system keeping track of the reports), but I hate having to grade these weird one-offs after I've finished the pile of papers.

Quote from: downer on October 27, 2021, 01:01:47 PM
Given how things are these days, if we are going to be legalistic, I'm happy for the burden of proof to be on the student that they wrote their work without cheating.

How is that even possible?

Yeah , in my case they usually come back and I have to revise grades as well.  It is a lot of paper work and such, but I guess I still prefer this over having to deal with it on my own and it is a far more equitable approach. 

dismalist

If there was evidence of cheating on a particular exam, I would decide what the penalty measured in percentage points on that question or exam should be. [Normally, no credit for the question. I never went overboard on penalties, for one can't be completely certain of cheating.]  I would then e-mail the student explaining what I had done with his or her grade.

The students were of course free to follow the rules by going to the appropriate panel. In the many, many years I practiced this, no student ever did, including one case for a final exam [and hence course] failure.

A paraphrasing technology would not obviate this strategy. Put the ball in the court of the accused.
That's not even wrong!
--Wolfgang Pauli

apl68

Quote from: downer on October 27, 2021, 01:39:49 PM


One of the many crises higher ed faces is that soon degrees will be worth nothing because it will be so common for students to find ways to massively cheat for any work except exams done in a classroom in front of an invigilator.

Even assuming an element of hyperbole in that statement, that's disturbing to see.
If in this life only we had hope of Christ, we would be the most pathetic of them all.  But now is Christ raised from the dead, the first of those who slept.  First Christ, then afterward those who belong to Christ when he comes.

ciao_yall

Quote from: apl68 on October 27, 2021, 03:22:46 PM
Quote from: downer on October 27, 2021, 01:39:49 PM


One of the many crises higher ed faces is that soon degrees will be worth nothing because it will be so common for students to find ways to massively cheat for any work except exams done in a classroom in front of an invigilator.

Even assuming an element of hyperbole in that statement, that's disturbing to see.

Yeah, just a bit of hyperbole. I have seen a handful of oddly written discussion posts and short assignments that were probably run through a paraphrasing tool.

Still, a student who relies on these has a lot of other educational issues and I can't imagine them being successful in school in the long run.

Parasaurolophus

Like kron, we have a panel, too. It's entirely toothless. It's where plagiarism cases go to die.

There's a lot of plagiarism and contract cheating here.
I know it's a genus.

jerseyjay

Over the past several years, I have been seeing more and more essays that are "paraphrased" in this way. It is usually obvious because the essay reads like something that was translated literally by an intermediate language student with a big dictionary. The words do not flow, and they do not really fit.

I do think they are getting better (in the technical sense). I ran the above paragraph through one (https://quillbot.com/) and this is what it gave me:
"Over the last few years, I've noticed an increase in the number of essays that are "paraphrased" in this manner. It's frequently clear since the essay reads as if it were literally translated by an intermediate language student using a large dictionary. The words don't flow well and don't seem to belong together."

If I saw the above, I wouldn't think twice. (Except for the use of the subjunctive mood, which is far too advanced for many of my students. On the other hand, "literally" morphed from being, well, literal, to the more typical meaningless sense.)

But my rules for paraphrasing are the same for cut-and-paste plagiarism. If I can prove it is plagiarized (i.e., find where it was plagiarized from), I will fail it. If I just think it is plagiarized, it usually has other problems (such as not answering the question, or semi-literate constructions in the case of paraphrasing) that results in a lower grade, but I cannot fail a student just because I think they plagiarized. If a reviewer said this about my book, I would probably sue for libel.

kaysixteen

Setting aside the very problematic notion of 'my court, my rules', which apparently suggests that Prof. Important can fail a student for plagiarism just because Prof. Important decides stu is guilty thereof, how exactly is stu supposed to 'prove he did not cheat'?   Think very carefully, and logically, about your answer.

jerseyjay

Quote from: downer on October 27, 2021, 01:39:49 PM
One of the many crises higher ed faces is that soon degrees will be worth nothing because it will be so common for students to find ways to massively cheat for any work except exams done in a classroom in front of an invigilator.

My first thought is that you seem to have much faith in the integrity of invigilators.

But more to the point, this seems to be a case of,  students today are so bad but were so much better in my day. Yes, cheating is bad. And yes, electronics make cheating easier (although arguably, easier to catch, also). But cheating scandals are nothing new. When I went to university, there were plagiarists, cheaters, and other such types of things.

Broadly speaking, a university degree's value is not really that its holder can demonstrate specific skills or content (outside of a few professional programs) but that graduates have been socialized in a particular way, and with particular people. For most people (i.e., not academics) the academic function of universities has almost always been secondary to these functions. This is why there have always been cheaters, and why there will probably always be cheaters.

To reduce cheating, it would probably be necessary to reduce the social importance of university, so only a very few people want or need to get a degree. Make it like graduate school in history. Nobody needs a PhD in history except a few small number of people--and even there, there is probably too many history PhDs. Why would anybody cheat in history? Yes, there are people who plagiarize their dissertations and books, but this is a small number.

On the other hand, when I taught in an MBA program (albeit briefly), there was more rampant cheating. Because the students needed to have the letters MBA after their name, and wanted the connections that going to B-school brings, but did not really care about the academic side of the program. Cheating in an MBA program makes more sense, from a utilitarian perspective, than in a PhD program in history.

ergative

Quote from: jerseyjay on October 28, 2021, 04:30:38 AM
Broadly speaking, a university degree's value is not really that its holder can demonstrate specific skills or content (outside of a few professional programs) but that graduates have been socialized in a particular way, and with particular people. For most people (i.e., not academics) the academic function of universities has almost always been secondary to these functions. This is why there have always been cheaters, and why there will probably always be cheaters.

To reduce cheating, it would probably be necessary to reduce the social importance of university, so only a very few people want or need to get a degree. Make it like graduate school in history. Nobody needs a PhD in history except a few small number of people--and even there, there is probably too many history PhDs. Why would anybody cheat in history? Yes, there are people who plagiarize their dissertations and books, but this is a small number.

On the other hand, when I taught in an MBA program (albeit briefly), there was more rampant cheating. Because the students needed to have the letters MBA after their name, and wanted the connections that going to B-school brings, but did not really care about the academic side of the program. Cheating in an MBA program makes more sense, from a utilitarian perspective, than in a PhD program in history.

This sounds about right to me.

downer

Quote from: jerseyjay on October 28, 2021, 04:30:38 AM
Quote from: downer on October 27, 2021, 01:39:49 PM
One of the many crises higher ed faces is that soon degrees will be worth nothing because it will be so common for students to find ways to massively cheat for any work except exams done in a classroom in front of an invigilator.

My first thought is that you seem to have much faith in the integrity of invigilators.

But more to the point, this seems to be a case of,  students today are so bad but were so much better in my day. Yes, cheating is bad. And yes, electronics make cheating easier (although arguably, easier to catch, also). But cheating scandals are nothing new. When I went to university, there were plagiarists, cheaters, and other such types of things.

Broadly speaking, a university degree's value is not really that its holder can demonstrate specific skills or content (outside of a few professional programs) but that graduates have been socialized in a particular way, and with particular people. For most people (i.e., not academics) the academic function of universities has almost always been secondary to these functions. This is why there have always been cheaters, and why there will probably always be cheaters.

To reduce cheating, it would probably be necessary to reduce the social importance of university, so only a very few people want or need to get a degree. Make it like graduate school in history. Nobody needs a PhD in history except a few small number of people--and even there, there is probably too many history PhDs. Why would anybody cheat in history? Yes, there are people who plagiarize their dissertations and books, but this is a small number.

On the other hand, when I taught in an MBA program (albeit briefly), there was more rampant cheating. Because the students needed to have the letters MBA after their name, and wanted the connections that going to B-school brings, but did not really care about the academic side of the program. Cheating in an MBA program makes more sense, from a utilitarian perspective, than in a PhD program in history.

I really like your points.

My own undergrad degree was based completely on written invigilated exams. It is an approach that has its problems but I guess I am feeling a bit nostalgic for it.

There is a balance teachers need to achieve between being "cheating cops" and not worrying about cheating at all, just letting students submit what they like. Lately I feel like I've oscilating between the two more. At times, I think to myself "I get paid anyway, and I am not going to waste my time with this stuff. I will focus on the students who want to learn."

Higher ed is under more pressure to show that it actually delivers what it promises -- the whole outcomes measurement movement has been pushing for that, though as far as I can tell, it hasn't achieved anything more than creating more deans.

Maybe we should build student socialization into outcomes measurement.

I'm not sure how many of the students I see really get socialized by college, as opposed to just growing older and a little more mature.
"When fascism comes to America, it will be wrapped in the flag and carrying a cross."—Sinclair Lewis

marshwiggle

Quote from: downer on October 28, 2021, 05:45:55 AM

My own undergrad degree was based completely on written invigilated exams. It is an approach that has its problems but I guess I am feeling a bit nostalgic for it.

The courses I teach rely heavily on hands-on projects. It's pretty much impossible for students to "plagiarize" something like that. I realize I'm fortunate to be in a field where that's an option, but I still think that the places where the most cheating occurs are the places that are most set up for it. For instance, exams that get re-used from year to year are basically begging to get easier as time goes on and more information flows to future generations. The same goes for essay prompts and things of that nature on things that are easily googleable. The fact that something more or less correct is already out there makes the thesaurus software barely relevant.


Quote
There is a balance teachers need to achieve between being "cheating cops" and not worrying about cheating at all, just letting students submit what they like. Lately I feel like I've oscilating between the two more. At times, I think to myself "I get paid anyway, and I am not going to waste my time with this stuff. I will focus on the students who want to learn."


I think this has to be the focus for a lot of reasons; disengaged students, whether they cheat or not, sap the energy of anyone trying to teach them.

Quote
Higher ed is under more pressure to show that it actually delivers what it promises -- the whole outcomes measurement movement has been pushing for that, though as far as I can tell, it hasn't achieved anything more than creating more deans.

Maybe we should build student socialization into outcomes measurement.

I'm not sure how many of the students I see really get socialized by college, as opposed to just growing older and a little more mature.

If there are no "hard skills" that education provides, then it's a ridiculously expensive way to try and provide "socialization".
It takes so little to be above average.

Caracal

Quote from: jerseyjay on October 28, 2021, 04:30:38 AM
Quote from: downer on October 27, 2021, 01:39:49 PM
One of the many crises higher ed faces is that soon degrees will be worth nothing because it will be so common for students to find ways to massively cheat for any work except exams done in a classroom in front of an invigilator.

My first thought is that you seem to have much faith in the integrity of invigilators.

But more to the point, this seems to be a case of,  students today are so bad but were so much better in my day. Yes, cheating is bad. And yes, electronics make cheating easier (although arguably, easier to catch, also). But cheating scandals are nothing new. When I went to university, there were plagiarists, cheaters, and other such types of things.

Broadly speaking, a university degree's value is not really that its holder can demonstrate specific skills or content (outside of a few professional programs) but that graduates have been socialized in a particular way, and with particular people. For most people (i.e., not academics) the academic function of universities has almost always been secondary to these functions. This is why there have always been cheaters, and why there will probably always be cheaters.

To reduce cheating, it would probably be necessary to reduce the social importance of university, so only a very few people want or need to get a degree. Make it like graduate school in history. Nobody needs a PhD in history except a few small number of people--and even there, there is probably too many history PhDs. Why would anybody cheat in history? Yes, there are people who plagiarize their dissertations and books, but this is a small number.

On the other hand, when I taught in an MBA program (albeit briefly), there was more rampant cheating. Because the students needed to have the letters MBA after their name, and wanted the connections that going to B-school brings, but did not really care about the academic side of the program. Cheating in an MBA program makes more sense, from a utilitarian perspective, than in a PhD program in history.

Yeah, put me down in the "nothing new under the sun" camp. The papers I assign are all designed in a way that cheating wouldn't get you a very good grade. I'm always asking for something fairly specific. A student could certainly dig something vaguely adjacent up and put it through a paraphraser, but it won't really fit the assignment, will probably be weird and stilted and will end up getting a C at best. Honestly, most of the time a student who did this could have spent a couple of hours and wrote a bad paper all on their own. If the student is a bit more clever and tweaks things, perhaps they manage a B, but again, now they have spent their time trying to cheat when they really could have just written a not great paper on their own. All they've really done is cheat themselves out of gaining useful skills.


marshwiggle

Quote from: Morden on October 28, 2021, 11:48:27 AM
Don't forget bespoke assignment services: https://www.reddit.com/r/PaperMarket/

As Caracal indicated above, it would be a pretty poor assignment prompt if some random individual who had never taken the course could successfully complete it on demand.
It takes so little to be above average.

downer

Quote from: Morden on October 28, 2021, 11:48:27 AM
Don't forget bespoke assignment services: https://www.reddit.com/r/PaperMarket/
I especially like the one asking for someone asking to write their Personal Statement for their third time around with dental school application.

That board is not very much used.
It seems that https://paper.papersowl.club/ is also not much used. Their main endorsement is from 2019.

I do wonder how much they pay their writers. It could be tempting.
"When fascism comes to America, it will be wrapped in the flag and carrying a cross."—Sinclair Lewis