Technology vs. Cheating at the University of Virginia
Posted by
michael
on Wed May 09, 2001 03:02 PM
from the grep-strikes-back dept.
from the grep-strikes-back dept.
Isaac-Lew sent in this story about a professor at the University of Virginia who heard rumors that his students were cheating and took action - he wrote a program to search through all the papers, identify common phrases, and flag the cheaters. Now a large chunk of the class is facing possible expulsion for plagiarism.
This discussion has been archived.
No new comments can be posted.
Technology vs. Cheating at the University of Virginia
|
Log In/Create an Account
| Top
| 590 comments
(Spill at 50!) | Index Only
| Search Discussion
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Re:The hard part is telling just who is guilty... (Score:3)
Lets assume that this testing for cheats is done and that everyone knows it (ie, it's mentioned once per semester).
This would mean that (as at the end of the article) very few people would cheat this way.
For each set of "matches":
If paper(s) match against a paper from a previous year or semester, then it's obvious that this current student is cheating.
If the paper(s) match only in the current semester, bring both of the students, and interview them seperately. It would be fairly easy to ask questions that would make it obvious that he or she cheated. Why? Because people cheat to be lazy. If they could provide the answers off the top of their head, they'd not need to cheat.
For the really odd case that both answer questions equally well, then you'd either have to mark it down for both or let them go. Your choice (I'd make it depend on whether they both seemed well versed let 'em go, else get 'em both in trouble).
This process is made easier if one has records of prior cheating or potential cheating.
Ciao!
Re:Watermarks! (Score:3)
Re:Only one thing shocked me (Score:3)
Re:The hard part is telling just who is guilty... (Score:3)
Re:"Group" Projects (Score:3)
Every academic discipline has them now (especially colleges of Education and Business) with the supposed goal of "teaching the students to work in groups".
Speaking as a professor who introduced group projects into his PChem course this semester, I think you miss the real point of them.
I used them so that students could teach each other. I wanted the strong students to help the weak, simply because you don't really understand something until you have to teach it to someone else.
Did the weak students benefit from the stronger ones helping them? Of course. But IMHO the strong students benefit even more: I was asked far better questions by the students who helped others.
Eric
Always was a factor for me... (Score:3)
This isn't uncommon (Score:3)
One of the CS professors at the university I attended was incredibly paranoid about cheaters. He wrote a similar program for scanning different students' submitted source code and flagging those that seemed similar. It's a pretty smart thing to do, if you ask me. Heck, I even know someone who got caught by it. I'm not sure how effective it was in general.
Writing Programs Rather Than Papers (Score:3)
Regardless, though, receiving a poor grade on any type of project is a million times better than copying someone else's work, even if you don't get caught.
I don't cheat and I don't steal, which is common sense in my mind, but unfortunately, not in the minds of many other students.
Re:"Group" Projects (Score:3)
I can tell you that still now, I am amazed on how well it works, many many times it happens that before the LAB session neither my partner(s) nor me have any clue on how to do it and don't feel capable of getting anything; but it happens very often that the pieces of knowledge that we both have add up to the knowledge needed to solve the given problem.
At the end, we get a good solution, and we both learned from each other.
Also I've been working in group projects and I really think it is worth it, usually they require quite a lof of thinking (and not easy) and sharing your thoughts and explaining your points helps in understanding better the problem and reaching a right solution; many times one thinks in one way about a problem, and when sharing the ideas with your group some difficulties in your reasoning are found, or improvements, or maybe a new different approach.
The main importance of group working is that the strong points of all the partners are added together.
Of course that there are drawbacks: the main beeing having bad group partners that are not at the same level than the others, then they cannot add to the group.
Student defense (Score:3)
The faulty dynamics of group projects (Score:3)
As I'm writing this, I'm currently a graduate student working as one of two currently employed by the university who are qualified to mark assignments in what is a popular course. From this perspective, group projects are great, because I only have to mark 1/3 as many assignments. My experience with actually working in a group as a learning experience is the opposite, though.
To date I've been in groups from both points of views. In computer science groups, I've been a very strong member of the group, and in some cases I've been a very weak group member. In both situations, I've hated it.
In university, people traditionally get assessed individually. Whether working in groups or not, everyone's primary aim is to get good marks for themselves. This is completely opposite from what group work implies.
The real world has teams everywhere. Realistically, it takes years for a really good team to form, where everyone's strengths and weaknesses are used efficiently and people work together. In the real world though, people aren't paying to be fairly assessed. In contrast, they're paid to work with other people. And there's a reasonable chance that if they're dragging other people along, they can leave the job or at least will eventually get reassigned - without effectively losing anything.
In an student team though, you're effectively thrown into a group and given about a week to work out each other's strengths and weaknesses. Then you're required to fight to the death about the best way to get the job done instead of being told by a team leader of some sort who takes responsiility (since everyone's geared towards their own individual assessment). Once a path's chosen at the expense of everyone else's ideals, there's not much option to change it down the track.
When I've been a strong member in a group, the weak members are often just completely left behind. Right now I'm working in a group of four. Person 1 has been sick for the last five weeks (the entire project so far), person 2 has no clue whatsoever about how to do anything, and most of everything's been done by person 3 and myself.
I'll ignore the sick person for now. The second person is a very nice guy, but he's just not grasping the subject at all, for as much as he's trying. He's repeatedly asking how things work, and no matter how much I explain, he simply doesn't get it, and in the process anything that he does related to contributions is likely to drag the mark down or break everyone else's code if it's not completely overhauled and rewritten beyond his understanding before it's used. Effectively, he's a liability. The group mark gets unfairly distributed to him, and our marks get dragged down because of him.
Having said that, I can sympathise with him completely because I've been in the same sort of position with other subjects in other courses. A couple of times I've ended up in groups where the other members are completely ahead of me, or think about problems in completely different ways. It's a really awkward position to be in, knowing that you're piggybacking on what might be a good mark, and not being able to contribute anything useful.
In these situations, strong students don't benefit at all, because of the typical assessment system. They end up with a proportionally unfair workload, doing their bit and redoing other people's bit so their grade won't suffer. Weak students don't benefit either - they just end up in a sea of not having a clue. If anything they might end up doing the drudgery work like writeups. Even then, it's really hard to find useful drudgery work.
Usually, the only way groups can work effectively when people can actually learn from each other, is when they're evenly matched - and that's a very unusual situation.
When I'm in a weak-student position, I've benefitted a lot more from working with other relatively weak students who are working through and figuring out the same problems that I am. The mark might not be as good, but it's more representative for everyone concerned and I feel much better about it as well as understanding more.
Asking good students is perfectly okay within reason, but it's unrealistic to expect them to work as tutors for weak students at the expense of their own work. From a strong student perspective, it gets really tedious answering the types of questions all the time, and often it doesn't help anyway, because students aren't trained teachers.
With respect to the idea that being able to work in groups is a good thing, I have trouble understanding what use it is to teach this in an academic environment. That is unless or until the assessment system is completely overhauled.
There's almost nothing that can be learned in 12 weeks (give or take) of infighting about the best way to do something. This is easy enough to pick up in a real job, and in many respects I've found it much easier to do in a real job because everyone (me included) is prepared to tolerate other people's ideas when it won't mean the plummetting of a good grade. Such group dynamics exercises would be better left to psychology and sociology subjects.
===
F for Reasoning (Score:3)
As Eric wrote in his post [slashdot.org], from a different perspective, anything we get from a computer we tend to treat as absolute fact. It is all to easy to find some connection that implies plagiarism.
There's a great statistic on birthdays. How many people do you think you'd need to have in a room before the odds were in favour of two of them having the same birthday? About 365/2=182? Actually about 30.
In a new year's science lecture on the BBC, a lecturer asked the left half of a room of 1024(ish) people to think heads, the other half tails. He flipped a coin and discounted the half that got it wrong. He carried on subdividing until he got to one who got it right ten times.
The problem is that most people don't realise how common some probabilities really are. In the first group of 30 (about a class size), "two of them clearly copied each other's birthdays!" In the second group, "no one can guess a coin correctly ten times in a row, he clearly went forward in time and copied what the coin was going to do - or the coin was rigged and he was told the answers!"
These are amusing, semi-trivial examples but they demonstrate the point that putting all of your convictions behind apparently conclusive numbers is flawed. Six word sequences can only be an indicator of cheating, not conclusive proof. All a six word phrase may really be showing you is that two students come from the same area and share the same turns of speach or that they were both equally influenced by something that was presented in a lecture.
I don't mean to maintain that statistical analysis is impossible, simply that it is all too easy to put too much weight behind it. Add that to the very valid point that in two identical papers, you may only have one cheat and one victim [slashdot.org], expelling based on the system seems very flawed.
Cheating is so very wrong (Score:3)
Think a program like this will send a wake-up call to those students who have forgotten what the community of trust is all about.
Technology has made some of the easy ways out very seductive and blurred the lines between what's acceptable and what's not. Cheating is on a gray scale. Things come rolling into your computer, and you feel ownership of them even if you don't own them.
And that's what I think.
Murphy's Law of Copiers
Re:Nifty (Score:3)
--
A strange sentiment from Prof. David Gies... (Score:3)
Maybe I missed something. I don't know how performing a pattern match on Term Papers in order to identify cheaters relates to the "community of trust"i>
--CTH
--
Re:Nifty (Score:3)
Good for you. Speaking as an astrologer, I knwo the value in this. I use a calculator for my astrology more often than a computer program because I have found that I understand how it works much better and can theneasily spot data entry errors with regard to computer programs.
All because I can think through the laborious calculations. (about 5 min. with calculator, 3 hrs by hand, less than 1 sec. with a computer). In this case the abilit yto set up the tables by hand is a tremendous asset. I wonder if I could do this if I had not learned to do it without a computer....
However, I am more prone to stupid errors by hand than I am with a calculator. The essence is not that I do not understand the problems (I can help those who are strugling to improve their grades drastically) but rather than little bits get flipped somewhere along the line, so to speak. So I learn best with a calculator OR forced to write my mathematics programs from scratch.
There are two sides to the coin in this case, I think.
Re:This looks like a Good Thing (Score:3)
That's irony.
--
"May the forces of evil become confused on the way to your house"
Re:This isn't uncommon (Score:4)
For about 3 years, I TA-ed an intro-level CS class that tought some rudimentary Pascal programming. It was a computer literacy course, so the bar wasn't high, and it wasn't for majors.
In sections of 50+ kids, I regularly found people who had copped each other's (sometimes non-working!!!!) programs, right down to variable names, etc. How lazy could you get! And this, despite the fact that if they had cheated with someone in one of the other 4 TA's sections they would never have been caught. I never had to diff anything -- you could just tell.
When I found someone doing this, I would hand the printouts back with the following written on them:
See for your grade (on Foo's program)
See for your grade (on Bar's program)
They got the point.
PHYS 106 a Joke (Score:4)
As a UVa grad, let me point out that the class in question here is generally regarded as a complete "gut" class. The large majority of students taking it are either athletes or people who just need a basic physics class to get their degree. It doesn't surprise me at all to hear that a large number of students got buster in that class.
Oh, and the honor system is regarded by many to be much of a joke, too. This really sucks for the students that got busted, but if they're going to cheat that blatantly in what's essentially a "gimme" class, they deserve every little bit that's coming to them. And it's always nice to see the honor code coming under scrutiny instead of simply being exhaulted as the greatest thing since sliced bread. :)
-jdm, (I'm not a bitter grad ... why do you ask?) :)
Re:A strange sentiment from Prof. David Gies... (Score:4)
Well, it wouldn't relate to a community of blind, unquestioning (and arguably in this case, naive) trust, but how it relates to a community of earned trust seems fairly obvious to me.
Watermarks! (Score:4)
Sounds like a job for watermark technology.
"As you can see, prof, if you take any paragraph of my paper and checksum it and rad-50 decode it, you get the word SLOPPY. That's why I had to use the strange word 'strategery' in the 5th sentence; it was the only way I could make the checksum come out right. Let's see the bad kid who sits next to me, who isn't named Sloppy, explain why his paper also has that mathmatical feature."
---
Teamwork = Cheating? (Score:4)
Many Hackers have a bent towards solitary work, and often reinvent the wheel more than they need to in the first place. We don't need the educational system encouraging this bad behavior.
The world of the Internet and open source development is finally providing a way that hackers from around the world can share their work and learn teamwork. This is a good thing.
While I don't know that the professor that was the subject of this article is really a good example of what I'm talking about, his actions are sure to spur on others to crack the whip and take things too far.
Re:Nifty (Score:4)
I'd rather have the students think about the answer than sit there pushing buttons on the magic box and taking whatever it gives as the truth.
Eric
Re:The hard part is telling just who is guilty... (Score:4)
Some professors aren't so careful, and will accuse students of cheating on a whim. I was so accused after submitting a final paper for a liberal arts class I was taking. The professor thought it was "too good" for me to have written it, and said that I must have copied from some other source. In fact, the entire work was 100% my own, using my own language. I didn't even do any direct research, just wrote a bunch of BS off the top of my head. After discussing the issue with the professor, and he relented and gave me an A-.
I want to stop students from cheating (and artifically raising the grading standard) as much as anyone, but not at the expense of trust between the student and the professor. Thats why I support systems that log papers submitted and run heuristics checks on them, but students should also be made aware that such systems are in use. I think this will be the necessary disincentive to force students to not cheat.
Ultimately I think the problem is exacerbated by massive classes (like this 500 student lecture) where the sole requirements for grading are usually a paper or two plus a final exam. If the particular professor who accused me had known me personally, or been at all familiar with the previous papers I had submitted, he wouldn't have been so quick to pass judgement. Huge classes also promote cheating because students know they are far less likely to be caught in such an evironment.
But thats just my 2 bits.
Spyky
Cheating might not be the cause of that (Score:4)
This particular prof was acting on a report that there was rampant cheating, and he was more or less looking to confirm. That makes sense.
However, in other fields where it's more text based (like "read these 4 books" instead of "study chapter 3 on partial differentials"), the papers could be excessively similar because they all draw phrases from the same sources.
Of course, you could argue plagiarism if students are pulling quotes and not citing, but a realistic instructor would realize that the students obviously draw from the assigned texts, and kind of take them as an "implied bibliography."
Which doesn't make it right to take other people's words and pass them off as your own. But it's so damn common that it passes for decent paper writing at 5/6 of the institutions in this country. While that's depressing, I don't think the kids need to be busted for cheating as much as get some remedial paper writing classes.
Again, these arugments may not apply to this particular case; these students might indeed deserve expulsion. However, I don't know that the approach is widely applicable.
---
Re:Nifty (Score:4)
Bottom line is a computer can't think, only calculate extremely quickly and accurately.
Another note, to discourage calculator use, give partial credit. The calculator program user will have no work so if they have a typo in thier calculations or whatever, they lose 100% of the value of the question where as someone that showed all thier work and got one little step wrong could get nearly full credit. The other way is to write intelligent test questions that require you to think to understand the solution but have easy setup and calculations so that even if the calculator helps you its not on the important stuff. Word problems rock for this purpose.
Though I must say I pulled off a LOT of B's by heavily using partial credit. Sketch down the first few steps of solving the problem that I could remember and get 75% to 90% credit on the problem even though I had no idea how to actually complete the solution.
Destroyer of Lives (Score:4)
Re:Good (Score:4)
I love these guys, and you should too.
They're the ones who come around the corner every half-hour asking me to explain pointer arithmetic or how a driver interface works.
I'm the star, they're the droids. Pay is commensurate. If this was an egalitarian industry with no pyramid of skill distribution, we'd all be making low-five-figure salaries, and thinking it was as right as the mid-six we're making now, because our peers would be, too. The broader the competition, the better your superiority stands out. It's better to be one in a million than one in a thousand. You get my drift.
It will take a few years after you graduate to sort you to your spot in the hierarchy. But you know how the playing field is laid out. Use that to your advantage.
--Blair
"U. of Macchiavelli, '84"
Re:The hard part is telling just who is guilty... (Score:4)
A potential solution to this would be to simply not punish the ones whose paper got copied -- only the one who plagerized. Sure, some people will get away with "aiding and abetting", but better to a let a few guilty go free than to punish someone who was truly innocent.
Re:Good (Score:5)
Oh-oh! (Score:5)
--
Re:Seriously. (Score:5)
Re:Group projects (Score:5)
I beg to differ. Like it or not, in the real world, you have to deal with other people, and sometimes, other people are dolts. This doesn't change the fact that you have to work with them.
In surveys of employers, `communications skills' are almost universally listed as the most desirable characteristic of new graduates. Actual technical proficiency usually slips in at number 4 or 5 on the list. Group projects are intended to give practical experience at communicating in and with a group of other people.
The problem with small group projects is twofold.
I was once of a similar opinion as you - given a group project (group of 4 or 5), I would usually end up doing the whole damn thing, and everyone else in the group shared in the good mark: a fact that pissed me off no end. Then I did a _real_ group project - in a group of 60. This was a second year uni project. We had a semester to organise a conference, each write a paper for the conference, peer review the paper between ourselves, and present the paper at the conference. We had to raise funds, organise every aspect of the conference from tea and cookies to keynote address. At the end, we published a 300 page book of proceedings, had it printed. I still have some copies sitting on my shelf.
A project this big cannot be completed by a single person. This forces you to organise, and work in groups. Rather than trying to finish everyone elses job (which is not feasible), you learn that you have to convince others to do their job.
The best feature of the large project I did was the peer review at the end. Students were asked to assess every other student. These asessments formed a large part of the final grade. Surprisingly, when given the responsibility, students will identify those who are not pulling their weight.
Group projects, if done properly, can be extremely rewarding. However, if group projects are to succeed, the project needs to be big, the group needs to be big, and the marking scheme needs to be independent.
A group of people working in concert can acheive much more than a single individual - I would not have been able to publish a book of proceedings by myself. In addition, for the remainder of that degree, the entire class had a great sense of comraderie, as we had all been through something gruelling, and we had done it together.
Russ %-)
PS: Any educators who are interested in the project I talked about here; I'm more than happy to advocate student centred learning to those looking to implement it.
This looks like a Good Thing (Score:5)
The article says that it takes a six-word phrase to trigger the initial match. That's quite a bit if you think about it; three- and four-word phrases are going to be relatively common, but beyond that...
It seems to have worked, too:
Good corrective feedback mechanism there.
Seriously. (Score:5)
"Smear'd with gumms of glutenous heat, I touch..." - Comus, John Milton
It's mostly our fault, not theirs (Score:5)
Then he goes out into a workplace that expects him to know what his diploma suggests he should. What now? Well, his strategy is going to be either to catch up fast, or keep looking for work to borrow/steal and pass off as his own. Probably the latter, because if the former was an option he probably wouldn't have had to cheat in the first place.
Is this any harder in the "real world" than it was in school? Nope. The internet is out there for everybody, and it's now just too hard to track everyone's work in a foolproof way. Will he get caught? Maybe eventually - but he's got a pretty good shot at becoming a comfortable PHB too, since so few of us have the energy to verify everything people claim. How hard would it be, for example, to print up a realistic-looking diploma or grad school transcript on a laser printer at Kinko's? If someone handed you one and it looked real, would you call the university to verify that it was real? No, you'd say "wow, MIT!" and hire him/her.
I used to teach GED classes, and I had students who passed who came back and told me that they had essentially closed their eyes and guessed at the multiple test questions, and done this over and over until they got a passing score. So they were out in the world with the equivalent of a high school diploma, who were barely literate and couldn't add 12+13.
We can write nasty things about cheaters, but they do it because we're all too lazy to police/stop them or really verify what their diplomas say they can do. The professor in this article was a very rare exception (he sounded like a cool professor, too). As long as people accept paper credentials as proof of ability (IT certification, anyone?), cheaters will keep doing what they do. Why shouldn't they? It's a much faster way to the top, and most of the time, we don't mind that much.
TomatoMan
Note to self (Score:5)
Re:Good (Score:5)
Not at all. The cheaters, having both low coding skills and morals, but an impressive sounding degree, are doomed to become senior managers and CEOs.
Don't you read Dilbert at all? I assure you it wouldn't be as funny if it wasn't mostly true.
Re:PHYS 106 a Joke (Score:5)
I always thought it was the latter -- either "make a buster" or "had a buster". Buster, of course, being synonmous for "fart" in the midwest. Or at least in my specifically midwestern high school where people used to sit in chem and phsyics class in hard plastic chairs and make a lot of loud, long busters.
Now, lest this post be marked off-topic, I'll say that as a former freshman comp teacher, I found myself spending more time checking search engines for "matching phrases" than I did actually grading and putting comments on student papers.
What's remarkable about cheaters -- freshmen cheaters in particular -- is that they tend to steal from the most obvious sources. I had one student in a film class filch an entire review from Roger Ebert. The only thing she changed was the byline on the review.
When confronted -- and I made the confrontation as quick and as business-like as possible -- she threatened to report *me* to the deans for harrassment. I laughed. She stormed out of the empty classrom and, sure enough, the next day I heard from a dean that I'd been "reported."
I explained the situation to the dean. He was floored by it -- floored by the cheating, the flagrant theft, and then floored finally by the formal report filed by the student.
I flunked the student. I was contacted by the parents and reported a second time. (I had *driven* the student to cheating because my teaching style was sub-par, said the parents)
A week or so later, I dutifully trudged down to the dean's office and came face to face with mommy and daddy. They were furious with me. "What was a graduate student doing teaching a class?"
The dean explained that, well, that was how it was usually done. We all agreed -- myself included -- that grad students weren't *always* the best teachers, but for the most part they were more than adequate and -- oddly enough -- sometimes *more* enthusiastic about the subject matter than their professional peers.
That was the end of that arguement but not the end of the case. The parents insisted that their daughter was innocent. I said, well, it's kinda hard to claim innocence when I have proof.
"Proof? What proof?"
"I have your daughter's paper and Ebert's review."
That's not proof, they insisted.
I was confused. I looked at the paper, looked at the review and then wondered aloud: er, what is it then?
It's proof of nothing, they said. That's not my daughter's paper.
The dean and I looked at each other.
Eh? I said.
The dean explained that, yeah, that *was* their daughter's paper.
"Did you see her turn that paper in? Did you see her give you that specific paper?"
I knew where this was going. The dean did, too. But the parents persisted. They wouldn't let this thing rest.
And on and on
The thing was never really resolved. I personally didn't change the student's 'F'. As far as I was concerned, she flunked my class. But I could never get further confirmation from anyone if, in fact, my 'F' stuck. It was all very insidious.
Anyway, my point with all this?
Some students are lazy fuckers with peabrains. Many students are not.
The lazy fuckers deserve to get caught and flunk.
Speaking As An Alum... (Score:5)
...let me just say that anybody who cheats "How Things Work" probably doesn't deserve to be at UVa in the first place. I could not take that course, because I was an EE. The course was considered both redundant and overly simplified for engineering majors.
However, I would be really surprised if even the most hung-over College of Arts and Science people couldn't at least pull a "gentleman's C" in that course. It's reputation was on par with other offerings such as "Cinema as an art form" and "History of Jazz", aka "History of Guts" if you catch my drift.
The other thing that non-Wahoos may not have picked up from the article is that there is a "single sanction" honor code at UVa. If you are convicted of cheating, you are expelled. There is no other punishment for "honor violations". The system has been criticized for inflicting its penalty disproportionatly on minorities. The flip side of that is that affirmative action programs encouraged people to enter UVa when they were not prepared. These are the people who will feel most pressured to cheat.
Of course, that was the way things stood when I graduated eight years ago. I'm sure some aspects of this are different now. OK, probably not, but one can hope.
Only one thing shocked me (Score:5)
What's shocking to me is not that people are handing in papers with long portions taken from other papers, but that the school is doing something about it. Even when I pointed out that a student had handed in a paper with a different name, the student got no formal reprimands.
Universities know where money comes from. I'd be very interested to see any followups to this article.
An old joke (Score:5)
The TA looks on bemusedly the whole time. When the student arrives up front, the TA says "I'm not accepting your exam, we finished fifteen minutes ago."
"Do you know who I am?" says the student.
"No." says the TA.
"Do you know who I am?" says the student, "Do you know who I am?"
"No." says the TA.
"Good" says the student and sticks his paper in the middle of the stack of papers and walks out of the room.
Cheers,
Good (Score:5)
Do your own work, never have a problem.
Re:Seriously. (Score:5)
--------------------------------
The hard part is telling just who is guilty... (Score:5)
The risk is that some of the students are probably innocent, merely being guilty of having their own papers copied without their knowledge. Indeed, we've seen many cases here where the person whose work was copied ends up in a situation where they have to prove their own innocence.
Unfortunately, the technology of online composition and submission of papers (as typically done at most Universities) lacks sufficient security, encryption, and authentication standards.
I just fear that the cost of this action could possibly end the academic careers of too many students guilty of nothing more than failing to see how their work could be copied.