Hacker News Reader

Comment by ses1984 | Hacker News Reader

ses1984

I'm not sure I'm against plagiarism detection tech.

Plagiarism detection tech seems like a solving-the-wrong-problem attempt to address a symptom of how we're trying to mass-produce education to save money.

If we actually want people to learn, plagiarism detection isn't the problem, it's the industrial factory approach in the first place. Get rid of that, and instructors can know and talk with their students well enough to tell what they actually know and understand instead of having to rely on artifacts that can be faked.

But we'd have to be spending more money in the first place, vs spending money on other things (like cops?!).

I don't agree with this at all. Assessment is completely necessary to understand where a student is at, and plagiarism renders assessment ineffective.

I think the point of the person you're replying to is that there are other methods of assessment which would render plagiarism ineffective - such as in-person presentations and conversations. But, teachers don't have the time and resources to do that with all of their students.

Effective writing is an important skill, so I don't think in-person conversations would replace papers, but from a conversation with a student who plagiarized their paper you could probably tell that they didn't really understand what they "wrote."

Industrial mass-produced "assessment" is only necessary in a mostly-anonymous, un-personal system. If you have time to actually talk to a student, you can figure out pretty quickly how well they understand something.

I think of it like stack overflow. If someone borrows ideas and code from there for a code review, I'm fine with that, as long as they understand how the system is doing. Being able to do research and find information is a useful skill! But only if it's applied correctly, with understanding. So I'm going to probe on that. And it might be blindingly obvious to me that they didn't understand what they were doing, or why the pasted code isn't appropriate. And I don't need a "stack overflow detector" to see that.

Yes, if we had unlimited resources and could give each student an individualized exam, oral and written under supervision, then plagiarism detection of tech wouldn't be necessary, but if you're trying to stretch a budget...

In the past we solved that by the instructor discussing the material with the students; it would quickly become apparent who knew something and who didn't. Unfortunately classes have become too big for that.

That injects a ton of bias into your marking though. Under that method, shy students and students with poor verbal skills (for example, ESL) don't stand a chance. I have seen excellent work come out of students that you could barely talk about the material, but could produce excellent assignments (this is in a domain where plagiarism would be very easy to detect, so I'm very confident it's actually their work).

It's better for shy students or students with poor verbal skills to work on that in an academic environment than to be dropped into a work environment where those things are just as important without their education ever bothering to help them with it.

"Interpersonal skills are optional" is not a good side effect of our current educational system!

I was gonna say. The author never elaborates on it, but this is a profoundly silly take. Plagiarism is not just unethical, it also is (in my experience) a pretty good indicator the student didn't learn the subject matter. If you know what you're talking about, it's relatively easy to put it into your own words. Students that I've caught plagiarizing either:

a) Can't actually explain what they were writing about

b) Didn't learn how to correctly cite things, and/or don't understand what plagiarism is

Are plagarism detection tools perfect? Heck no. Lots of false positives, and of course they don't do anything for essay writing services (unless, of course, the service plagiarizes - which happens!).

While plagiarism is unethical, nearly all means to combat it (detection software, esp. Turnitin) is also unethical; two wrongs don't make a right.

Under the ToS of all plagiarism detection software (at least that I've looked at), the student's work is added to the corpus against which all texts are tested against. This is what you would _obviously_ do if you were building a plagiarism detection app, however that means a student is forced to submit work to a private company, which will then profit off of their work.

You might say what the value of a gen ed term paper is to a student, and I'll agree it's not much. But that's not the point. The student is forced to provide that to a third party by the school, which will then profit from it. This isn't a Facebook-like situation where you can just not make a Facebook account. I would guess in many schools, refusing to submit to plagiarism detection software would be taken as evidence of plagiarism.

This is probably a few years out of date, and I see that Turnitin recently shut off their 'upload your own work so you can see if it's plagiarized' product, but for a student to have an active defense against plagiarism _costs money_. Yes, if a student wanted to make sure their paper wasn't accidentally plagiarized (this is common, because their algorithms are shit), that will cost a few dollars. This is unacceptable.

And for anyone who says, 'just transfer sections to a prof that doesn't use the software', there are contracts with entire departments and schools, which then force professors to use the software. Yeah, you could transfer _schools_, but the point stands.

Plagiarism detection software -- which purports to support the intellectual property of the author -- does just the opposite.

Also, I should mention that if someone wanted to cheat, they wouldn't plagiarize; they would simply get someone to write a paper for them. This service is far, far more common that you would believe.

> Also, I should mention that if someone wanted to cheat, they wouldn't plagiarize; they would simply get someone to write a paper for them. This service is far, far more common that you would believe.

When I was in school students did plagiarize a lot. It went out of style as software was introduced. I distinctly remember the panic'ed buzz as students discovered the teacher (or professor) was using detection software. The services you're talking about are common precisely because of plagiarism detection software.

I'm not disagreeing with your points about morality of it though, just the efficacy.

Plagiarism detection is only needed because we are using completion/grading of work to determine whether students graduate, GPAs etc. The article is arguing that we shouldn't be so dependent on the "grade of homework" metric at all, so plagiarism should be less impactful and less incentivized.

I have seen students plagiarize co-op work term reports, which at the institution that I marked at are binary pass/fail with a near 100% pass rate (barring plagiarism). The assignment isn't long, is on a subject the students already know, and you can hand in some pretty terribly written garbage and still pass. But students still try to plagiarize on it - there's a subset of the population that wants the degree, but can't/won't put in the time to do the actual work.

Requiring a plagiarism detection feature is a symptom of a shitty instructor, syllabus, and/or academic institution.

How does a good instructor, syllabus, and/or academic institution not need this? I went to an Ivy League school with hopefully above average teachers and saw a lot of plagiarism.

The problems are societal in nature. I don't beleive either the parent commentor or original author fully appreciate this.

Adversarial teaching environments are a symptom, not the problem. And the fundamental issues don't have a clear solution, though it is clear that the situation is far from optimal.

In principle, you remove the incentive for plagiarism by making the students' reward for their work be what they learn as a result, rather than some kind of certificate.

Oh cool, so we just need to fix a massive social problem that's generated by long-running structural trends within our society and economy. Then we can stop using plagiarism detection software.

... So we shouldn't try to fix the problem?

Of course we should, but we shouldn't be naive about the challenge doing so represents.

They could have set work that is difficult to lift word for word, or examine in any number of ways that do not allow googling

In one class, in order to submit an assignment, we had to accept an agreement that gave the plagiarism detection company the copyright to the essays we had written. I’m sure there is an okay way to do plagiarism detection but I think many companies do it in ways that don’t respect the rights of students.

Bad teachers may punish students for false positives. It's a tool, not a blackbox of academic integrity.

https://www.nature.com/articles/d41586-019-00893-5

I saw multiple instances of my classmates plagiarizing. It's very rampant in colleges. Removing that tech would be a mistake in my opinion.

Reply on news.ycombinator.com