Is This Your Homework, Larry?

March 19, 2016 5:35 AM

A friend is grading a paper that is obviously plagiarized but Blackboard detects zero matches. Google is no help to us either.

Months ago, a teacher friend of mine caught a student cheating in his college anthropology class and let it slide.

Now, this student has submitted an essay that consists of ninety percent garbage and ten percent professional writing of the highest order. My friend puts this thing into Safe Assign via Blackboard and it comes up with a ZERO percent match.

My friend says that most papers come up with at last a ten percent match, but this essay has a ZERO percent match so I'm wondering if this student found some kind of app that takes an essay and scrambles it up in such a way that makes Safe Assign unable to trace plagiarism. Is this possible?

I'm googling phrases in this essay and I can't find any of them on the web, though they appear to taken from critical essays. What is going on?

posted by Hennimore to Education (21 answers total)

I've run into situations like this before. My two guesses are either that they are plagiarizing from a book that hasn't been digitized (depends on the topic how likely this is) or they paid someone to write the paper for them.
posted by hydropsyche at 5:43 AM on March 19, 2016 [5 favorites]

Sounds like the student hired a professional paper writer -- although not a particularly good one, since you say the paper is a bad mashup.
posted by pie ninja at 5:58 AM on March 19, 2016 [5 favorites]

This happens to me all the time and it usually means the student did a thesaurus replace on every third or fourth word (which always amazes me because that would seem like more effort than writing the damn paper in the first place.)

Or they ran it through Google translate a few times (but that is usually pretty obvious stylistically).

You can usually figure it out by selecting a sentence that doesn't seem very idiomatic and replacing a couple of nouns with more idiomatic equivalents and then googling it.
posted by lollusc at 6:01 AM on March 19, 2016 [10 favorites]

My old trick when I knew someone plagiarized but I couldn't prove it was to ask the student to see me in my office, and then ask very specific questions about their argument, their phrasing, etc. Stuff like, "you say XXX on page 4....that's really interesting. Can you explain what you mean by your use of word Y?" It didn't always lead to a clear cut plagiarism "conviction," but it was awfully fun watching them squirm....
posted by Mrs. Rattery at 6:05 AM on March 19, 2016 [62 favorites]

Has your friend tried googling the suspected plagiarized passages? Also, I'm not sure if it uses a different database of work, but TurnItIn is another plagiarism-detection application that your friend might want to try.
posted by stillmoving at 6:46 AM on March 19, 2016 [3 favorites]

At the .edu where I work, I think that BlackBoard runs TurnItIn on everything that's submitted.
posted by wenestvedt at 7:09 AM on March 19, 2016 [1 favorite]

Your friend should contact the student conduct board and discuss with them but yes, having them come to your office to discuss the paper is usually best.
posted by k8t at 7:15 AM on March 19, 2016 [1 favorite]

Does looking in Google Scholar and Google Books help? Those search results might not show up in your standard googling.
posted by dis_integration at 7:20 AM on March 19, 2016

I caught someone doing that little 'replace every third word with a synonym' once by googling the concepts they were writing about instead of just the sentences - maybe give that a try if you haven't already?
posted by DingoMutt at 7:26 AM on March 19, 2016 [1 favorite]

Turnitin is a separate paid service, but some institutions (like mine) subscribe to both, so you can check to see if that's available. Turnitin is generally the more robust database and algorithm, but either way, they are looking for exact matches of a certain number of words in a row (with Turnitin you can tell it how many words in a row at minimum). If the student used a thesaurus to replace several key words per sentence, it's not going to catch it. I think the most efficient way forward is, rather than wrestling with somewhat opaque proprietary search algorithms is to call the student in and try to get them to express these arguments in their own words. And put a call in to the conduct board.
posted by soren_lorensen at 7:51 AM on March 19, 2016 [3 favorites]

Oh also, I heard talk of some trick used to fool TII that might also apply to Safe Assign that involves white text in the document? So I'd also try to inspect the actual document itself, maybe paste the text into a txt file then repaste into word and submit that to Safe Assign again.
posted by soren_lorensen at 7:53 AM on March 19, 2016

I know plagiarism is a really serious thing, but if the paper is 90% garbage, why not just fail the paper?
posted by xingcat at 8:01 AM on March 19, 2016 [46 favorites]

I am with xingcat. There are two issues here as far as I can see. One, the paper stinks and on its own deserves a failing grade. Two, it appears to be partly plagiarized. I would grade the paper as a standalone paper. Then, I would consider giving the entire class an in class writing assignment of about a page or page and a half. Then you have a baseline writing sample from which to compare. IF the writing style is so remarkably different, you can go from there. If it is similar, maybe the student just has a unique way of writing.
posted by AugustWest at 8:13 AM on March 19, 2016 [8 favorites]

Plagiarism in some schools is penalized much more harshly than a failing grade on the paper. If the school treats such matters as very serious infractions I would continue investigations, but if not just fail the paper.
posted by TheAdamist at 8:40 AM on March 19, 2016 [8 favorites]

Months ago, a teacher friend of mine caught a student cheating in his college anthropology class and let it slide.

It's not obvious how seriously the school takes cheating. xingcat's solution seems like the simplest solution.
posted by Beti at 9:22 AM on March 19, 2016

I agree that xingcat's solution is a good one. I would probably also have a conversation with the student, under the guise of "I'm concerned about your failing paper, and some issues in that paper" and see what the student has to say.

One thing that makes this more difficult for the instructor that they already caught the student cheating (though I'm not clear on what that means here, exactly - confirmed plagiarism? cheating on a test?) and "let it slide". Since they let it slide -- by which I assume you mean they didn't discuss it with the student -- they can't easily count that as a prior incident. My dept/institution has something like a 3 strikes policy, and if I "let the first one slide" I'd be at square one with any subsequent issue. Either way, the first step at my institution would be a conversation with the student.

It's probably not necessary, btw, to use tech tools to track down a 100% match. There are other ways to catch & pursue plagiarism -- asking the student to sit down and elaborate on the points in their paper being one of them. The big issue is that the work is (probably) not the student's own, and not so much where the paper came from.
posted by dryad at 9:42 AM on March 19, 2016

My position on academic misconduct is that it is important to go through the formal channels because the student conduct board deals with this stuff daily and the staff are better at both detecting and assessing misconduct. In my experience, the discussions that students have with the student conduct board staff are often fruitful. It also puts it in a third party's hands and is less about he-said-she-said. And, perhaps more importantly, the student conduct board probably has a central database where these things are tracked. Repeat offenses are problematic but are undetected without a central system.

So, I would tell your friend to stop hunting and call the student conduct board (or whatever this office is called at his/her school) and talk to them about it.
posted by k8t at 10:28 AM on March 19, 2016 [5 favorites]

Maybe the student asked someone to proofread the paper, and the proofreader realized the whole thing was terrible and gave a few examples of how it could be made more coherent. The student then just put those sentences directly into the paper without changing anything else. That would explain the quality differential and why the well-written parts aren't in any other database.
posted by Blue Jello Elf at 11:00 AM on March 19, 2016 [2 favorites]

Since they let it slide -- by which I assume you mean they didn't discuss it with the student -- they can't easily count that as a prior incident.

That's not really true in the big picture. That means you can't count on that to make this a "you cheated [this specific time], and it's your second offense." But you can change your angle from "you cheated" to "we believe you're a cheater" in general, based on a pattern of behavior. In that case the first occurrence still exists as supporting evidence, even if you didn't take any action at the time.

I like the oral defense idea; if the student can credibly get through that, then maybe who cares, he learned something. An outright cheater couldn't possibly. Alternatively, maybe give the whole class 30 minutes to write an abstract of their paper from memory, the thesis and a rough outline of the major points. That might be an interesting experiment anyway.
posted by ctmf at 2:00 PM on March 19, 2016 [3 favorites]

Longer-term strategy: cheaters gonna cheat. Say nothing, in fact, praise the quality of the suspect parts. Student will be emboldened to do it again and be less careful about it. You'll get him or her eventually.

Drawbacks: borders on entrapment (which doesn't bother me any, ymmv). Also, there might not be enough time left in your term with the student to let it play out, so you end up bagging someone else with the problem later.

Middle ground: bring the student in, say you're not going to press the issue, but that it looks awfully suspicious to you. Ask the student to think about what path they want to take in their academic career and the potential consequences of choosing wrong. Sometimes, just that "oh shit, I thought I was smooth, but somebody noticed" can fix the problem forever. I believe in the goal of "correct the behavior" over "punish wrongdoers" whenever practical, and maybe this is one of those cases.
posted by ctmf at 2:16 PM on March 19, 2016 [2 favorites]

It's probably not necessary, btw, to use tech tools to track down a 100% match. There are other ways to catch & pursue plagiarism -- asking the student to sit down and elaborate on the points in their paper being one of them. The big issue is that the work is (probably) not the student's own, and not so much where the paper came from.

With regard to this and all the other people suggesting calling the student in and discussing it, the OP is presumably asking about how to detect the plagiarism source for a reason. At my university, for example, there are exactly two cases where we can go ahead with a plagiarism accusation. One is when we can prove plagiarism by putting the source and the paper side by side as evidence. The other is if the student confesses. Admittedly an in-person conversation about the paper might lead to a confession, but if the student stands their ground, it doesn't matter how badly they explain their paper, you can't treat it as plagiarism.

Also, the official policy on plagiarism at my university has different pathways depending on what a student says when confronted about it (assuming there is also textual evidence). (For example, if they deny, then it goes to the Dean to deal with; if they admit it, the instructor gets to do all the paperwork themselves, yay). So calling the student in for a conversation could trigger something official, and the instructor needs to check whether or not this will happen before beginning an investigation.

And to be fair, some students, especially where English is a second language, would really struggle to explain concepts orally that they might well be able to write about with plenty of time to think, and look stuff up in a dictionary, and so on.

So I think it's more helpful to the OP to focus on ways they can find the plagiarism source rather than other ideas for how to make the student uncomfortable.
posted by lollusc at 5:37 PM on March 19, 2016 [2 favorites]