Slashdot Mirror
Cheating Via the Internet at College
Electron Barrage writes, "An anonymous professor writes that last year about half of the seniors at his US university were suspected of cheating, mostly due to the Internet and community sites such as Wikipedia. He guesses that perhaps 25%-30% were actually guilty, a huge increase from earlier levels. According to this professor, it's nearly impossible for the universities to keep up with the new forms of cheating enabled by the Net. Will academic institutions learn to deal with this new reality? It sounds a little dubious from this professor's viewpoint." The article mentions the anti-cheating services Turn It In and iThenticate (while decrying their expense), but expresses worry over the new countermeasure represented by Student of Fortune.
Sounds About Right
When I TAed a CS class, we caught about a quarter of them turning in the same assignment, some with 0 byte diffs from the others, some with just renamed variables. I think about 8 of em got serious disciplinary actions taken.
I remember a certain incident here at school in a class of my friend's. Apparently, after the professor started the exams, he would go back to his office and post the answer keys on the course website. Some kids found out, and would have their friends wait until it was published, then send a text message with the answers. The professor found out this was going on, so during one test he published a false answer key and found all the kids who were cheating.
--- At my sig, unleash hell.
Being technically apt, I helped her mark most of the assignments for that course. After the first round of marking, I had an inkling that a group of her students were cheating by handing in duplicated spreadsheets.
Her: How can you tell? .... Bob.
Me: Well, for starters, they have the exact same data.
Her: They did do web searches, so they could have found the same site.
Me: Okay, but look at this. (alt-tabs between the 'sheets). They have the same formatting, font and cell size.
Her: It is the default font...
Me: True, but the formatting isn't. But check this out. You know how when you scroll down, then exit the spreadsheet, it "remembers" where you were when you re-open it?
Her: Yes?
Me: Check this out. (scrolls up to "title" line). See the student name?
Her: Yes. It's Bob.
Me: Right. Because this is Bob's spreadsheet. Now (alt-tab to Mary's, scrolls up) check out the title bar.
Her:
Me: (repeats for three others)
And laziness is very easily spotted. I was able to see the simliar formatting and data. Anyone with a little bit of tech knowledge could spot it. But forgetting to remove the first student's name after the copy-and-paste...
The point is, students who cheat are lazy. And lazy cheating is sloppy cheating. And sloppy cheating is easy to spot. The amount of effort one has to put into cheating "undetectably" would be equal to, if not much greater, than just doing it honestly.
UTF-8: There and Back Again
The sad thing about this is that most professors know that this is happening. And the solution, well, a lot of people aren't going to like it. There's a principled answer (do lots of delightfully unique, practical assignments that can't just be cribbed; include a lot of 'called onto the carpet' type assessment where the students must verbally justify their essay/code/proof/whatever).
Unfortunately, the 'I don't have time or funding for anything special' answer to the problem is to move massive amounts of assessment into in-class, high-pressure exams. So, if you're like me (thrive in these kind of exams, don't mind cram-studying, etc.) you'll love it. But there are many smart people out there - especially, it seems, women - who do comparatively worse under these kinds of high-stakes, high-pressure assessment than they do under comparatively more realistic settings.
As an aside: As someone keen on maintaining the integrity of undergraduate education, I think it would be a great idea to seed sites like Student of Fortune with plausible answers that would slide by some cheating twit, but would instantly be detected by a TA or professor. I bet you could slide some really amusing stuff past these guys...
people's reputation will come less from the judgement of university systems, but rather from people's actual connections and accomplishments.
If you haven't noticed, a large part of going to university is about connections. Dubya would not have become a member of Skull and Bones if he had been taking classes of "www.affordabledegrees.com" instead of going to Yale. Do you think Dubya has had most use of the classes he took or the club membership?
It's all about connections. Nobody becomes great alone.
Tell your friends about xenu.net
As I mentioned in another thread, this doesn't make sense. The problem is that 'handed-in work' and exams don't actually serve the same purpose. Professors don't want students to write papers in order to demonstrate their knowledge; they want students to write papers because that format promotes original thought and the development of new ideas. You can't replace this function with exams.
You make an excellent point here, I must admit. Perhaps my bias stems from the fact that I work in fields in which exams are the norm, and not papers.
Still, I do want to argue a different position than yours. Now, I agree with you that exams do not test original thought, the development of new ideas, and research skills. However, I claim that the majority of papers do not do this much either. Many undergrad papers are basically 'book-report'-type things (albeit with several books and more difficult subject matter than grade school book reports). For this type of paper, an exam is a reasonable substitute (the only thing it might not test is long-term writing skills, i.e. editing and so forth, and not short-term writing skills).
For papers of a higher level, that is those that do focus on original thought, I would say the following. First, cheating is less of a problem with such things; they appear mostly in graduate-level courses, with less students, and more direct professor-student interaction. So, perhaps you are right in this case, and papers could continue to be used as the grading technique. However, if cheating were still an issue, there is another option, apart from exams and papers, suitable for classes with few students: a grade based on class interaction and/or a one-on-one interview-type exam. I recall taking classes where part of the grade was determined in this way, it seemed surprisingly fair, actually.
Students learn by doing, and in most academic fields, doing means research and writing. Many college courses need fewer tests, not more.
Again, an excellent point, which you have mostly convinced me of. I would only add what I just said above, that class participation and interviews could also be used, not just papers. A paper accompanied by an interview seems like a particularly useful method; I would think that discussion of the research and thought process that went into the creation of a paper would do much against the possibility of cheating. Of course, this would make sense in small classes only, as I said above.
Reminds me of when I was cheating (although not on univerrity) I made a cheating-not and wrote it smaller and smaller and that a few times. I perfected that cheatingnote so often that by the time I needed it, I didn't anymore. So the joke was on me, instead of making a cheating-note, I was actually learning and probably spend more time on it this way then when I would have 'learned' it the regular way.
i've done that since a history teacher in high school allowed us a single side of a 3x5 (or whatever the standard size) index card for notes for an entire semester. you'd write as small as you could on a notebook sheet of paper, reduce and figure out what was mos timportant, redo, etc. until you remembered all the fairly trivial things - as you decided "if I can remember it, I won't need to write it down". eventually you get your note card, and then you rewrite it so its legible. by the time you're done, all your notecard has are short abbreviations with dates, numbers, and other pertinent data in CSV, and you've essentially memorized which event belonged to which section of the index card, allowing you to figure out what you had to do.
i've used that method of studying ever since. it's very effective.
~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
I think learning can go on in the university but it's rare. Of all my teachers, of all my classes less then 2% were about learning something. The rest were mere memorization. I remember one pysics teacher who held open book tests, you could use anything you wanted, books, calculators, computers whatever. He could do that becuase he would ask questions that required thinking really hard and not memorization of formulas. I wish there were more teachers like him but alas that's not true.
These days an undergraduate is lucky if he takes five classes from a real professor in four years. People cheat because they can. They can cheat because the teachers are lazy and ask stupid questions you can look up in a second.
evil is as evil does
Ahh, an exam cynic. A product of the US (or North American) higher educational system, no doubt. :)
Patronizing remarks aside, what I'm saying is that I understand your position, but it's coming from a limited perspective. Go study in Europe a bit. I did. (Sweden)
Yes. True/false and multiple-choice exams are worthless. Yet very common in the US, and perhaps only there.
My experiences from Sweden were the following: First, exams are far more common, and far larger in size.
Second, true/false exams are never ever used. I never had one and never heard of anyone else having one either.
Third, multiple-choice exams exist, but seem to be quite rare. I only had one in four years, and that one was still augmented by two essay questions and the fact that they deducted points for wrong answers on the multiple-choice part. (The clear message being: Don't even think about guessing your way)
I think there are two very simple reasons these tests are used so much in the US, and neither of them are because they're a good method. First is because they're established and part of the academic culture. That kind of stuff is hard to change. Students don't want tougher exams, and Profs don't want to appear harsher than the others. Second, the Professors are lazy. Why use a test that involves more work from them and will generate complaints from the students, when you can stick with the same-old and everyone's happy?
So on to essay exams, which you've probably guessed would be the usual type, then. Your first criticism is that they're too short. Well again, there's a solution: Longer exams. Where I was studying the typical length was 4-6 hours. Bathroom breaks weren't a problem. The basic order was: Noone gets to leave during the first hour but latecomers can arrive in that period. After one hour, you're permitted to turn in your exam and leave if you want and bathroom breaks are permitted. Bathroom breaks are allowed one at a time, and the exam monitor keeps a log of which students leave and the time of their departure and return. They also log the seating arrangement and the time the exam was turned in. There are also voluntary 10-minute rest breaks where those who wish can leave with a second monitor. (who otherwise is outside, usually checking the bathrooms).
I never saw that dicipline broken. Cheating did still occur of course, but the worst I ever saw was people stuffing a note in their pocket and checking it in the bathroom. And that's of course impossible to stop. But so is any dedicated cheat. And the amount of information you can inconspicuously fit in your pants is rather small.
But anyway, point is the bathroom thing can be solved, and has been solved.
Your second argument is even sillier. Bad handwriting? That never seemed to stop them over there. But if it's that bad you can simply require the answer to be written in block text. Or set the condition that illegible answers will not be graded. That should do it.
You leave out what is, IMHO, the most superior form of testing though. Namely oral exams. (not necessarily the formal type) There is simply no better way to assess someone's knowledge and understanding than by talking to them, asking questions and follow-up questions to understand their thought process. And it is very much a real-world type of situation. The only drawback is that it's labor-intensive. I saw some creative solutions to that, though. One was to have a written examination to qualify for a passing grade, and an oral exam for higher grades. That filters 'em out quite well. I think that's because people won't take a chance on it unless they feel certain they've got at least some shot at it. First because it's harder to fake it. Second because people are a lot more reluctant to parade their ignorance in person-to-person communication than on a piece of paper. (Of course, some might simply be too nervous. But if you want to talk real-world, then they'll need to learn to deal with that sooner or later, and
Maybe you were in the wrong field of study. In all of my undergraduate major-specific courses, I always had a professor who loved teaching the subject, was passionate about doing related research, and was very knowledgable about the topic at hand. I never found cheating to be prevalent, simply because the answers to the problems did not exist (each professor usually made up the problems we were given). My undergraduate degree is in Aerospace Engineering, and almost all the homework and exams we received involved questions that simply could not be cheated on. I don't know how it would be easy to cheat on a homework involving the "prediction of viscous losses on overlapping rotor-blade interactions." Try finding more than a handful of relevant papers on the topic on Google.
I think that the problem may stem from people's majors being over-generalized by the professors. I know that if people did cheat in my major, they ended up hurting themselves; it's not easy to understand problem-solving methods when you never spent the time to learn them.
I am now in graduate school for Aerospace Engineering, and I find that cheating here is non-existant, if not impossible. How does one cheat on problems that the professor makes up during class, and expects to be solved by the next lecture? And cheating on a thesis? Good luck defending it.
If you steal from one source, that's plagiarism. If you steal from many, well, that's just research.
In Zen and the Art of Motorcycle Maintenance, the narrator advocates doing away with grades and diplomas. Students can show up and do the work, or not, their choice. Do it that way and there's no reason for anyone to cheat. Those that don't do the work will get farther and farther behind, feel lost, finally get discouraged and quit. What happened? They flunked themselves out!
Over the next few years, the school of hard knocks kicks in. If they're intelligent, they get bored at the mundane jobs they're stuck with. They want to do something more interesting. They start educating themselves. Suddenly they're learning because they have a passion for it, not because they want a piece of paper. They're getting a real education, not a fake one.
Does it give employers a way to pre-filter people? No. Would it work for everyone? I don't know, but it worked for me. I started out as a math major, goofed off, fell behind, switched to a nice easy anthropology major. Got an entry-level graphics art job after graduation. After a while, started seeing ways I could automate things if I knew how to write software. Started buying programming books. Ended up getting a job writing software, now making high five-figures, still kinda bored. Now I'm studying computer real computer science books with every spare moment.
I talk to people who got real CS degrees, and find out they know less than I do. They mighta been taught the stuff in school, but they promptly forgot it, because they had no passion for it.
If it's a fake education anyway, if people are going to forget what they just learned because they don't care about anything besides a piece of paper, then it doesn't really matter that people are cheating, it's just a slightly more advanced expression of the same underlying problem. Get rid of the external rewards, and at least what remains will be real.