Is Cheating a Lucrative Business for College Students?

[Astronuc]

Florida college professor gives cheating business students an ultimatum - confess, retake the test and attend an ethics class, or don't graduate.
http://news.yahoo.com/video/business-15749628/students-busted-for-cheating-22954742

Everyone doesn't.

 

[Grep]

Wow, that's disgusting.

What really astounds me is the kid justifying it with something like "everybody cheats, it's just a witch hunt". Seems like he has some serious issues. I can truly say I've never, ever cheated on tests. And I figure the business world is better off without people like that. I swear you'd have to be some kind of sociopath to say something like that.

I feel really sorry for the professor. I also feel really sorry for the ~ 400 students that have to retake the exam because 1/3 of the class are cheaters. The prof is being really nice in allowing them to take the test again and take an ethics class, IMO.

 

[BobG]

The prof is being really nice in allowing them to take the test again and take an ethics class, IMO.

Identifying an abnormally high average to confirm other suspicions (the answer key left in his bin) is a strong indication that many students cheated. Identifying which students is another thing entirely. The options given are a realistic assessment that they would have a very hard time identifying which students cheated.

Of course, since about 150 people accepted his offer and admitted cheating, it's almost certain that a significant number of the admitted cheaters will reveal enough details about the cheating that most of those that didn't admit cheating will get caught, too.

A real life 'Prisoner's Dilemma', except with 200 participants instead of only two. The number of conspirators tilts the odds a bit heavily.

Ironically, if the student claiming all students cheat were correct, the professor wouldn't have seen much difference in the scores for the test in question.

 

[rootX]

This professor is just making big drama. I don't know how solutions were passed to the students but was it his fault? He can easily make a really hard final exam and bring average down to the normal level.

I remember once at my sister's university, a professor posted solutions online by mistake before the exam. Everyone did very good on the exam.

For final, professor engineered a really hard exam that put everyone back to their normal marks.

 

[BobG]

I wonder how many didn't cheat, but confessed anyway. If a student did much better on the exam than they usually do, they probably were feeling a little nervous.

The worst case of admitting to cheating when they hadn't was to lose an abnormally high score, take an ethics class, plus suffer moral indignation at confessing to something they didn't do.

The worst case of being wrongly found guilty of cheating was not to graduate, plus the moral indignation of being found guilty of a crime they didn't commit.

Avoiding a really bad worst case scenario is often times better than achieving the best scenario (one good test score that wouldn't change their overall average significantly). The choice could come down to how the student perceives the school will catch the cheaters. If there were a feeling that an abnormally high test score alone could be all the evidence the school needed, then the student would certainly be tempted to admit to something they didn't actually do.

 

[D H]

Whoo boy. "Every cheats".

We used statistical techniques to detect cheaters 30 years ago when I TAed an intro computer science class. Our solution was simple: We graded the assignment or test as-is and then divided the score evenly amongst the cheaters. Four people cheated on an assignment: Their mediocre 80 turned into a 20. Two people cheated on an exam: Their fantastic 96 turned into a 48. Ouch!

 

[waht]

How about changing homework assignments, and test questions every semester. That's what many of my profs did.

 

[CRGreathouse]

We used statistical techniques to detect cheaters 30 years ago when I TAed an intro computer science class. Our solution was simple: We graded the assignment or test as-is and then divided the score evenly amongst the cheaters. Four people cheated on an assignment: Their mediocre 80 turned into a 20. Two people cheated on an exam: Their fantastic 96 turned into a 48. Ouch!

And if one person cheats?

 

[D H]

If 200 people cheated on the final they would have each received at most half a point out of 100. That would pretty much be a guaranteed F, by about 69 1/2 points, and that is assuming that those 200 students somehow managed to get a perfect score. One of the weird things we discovered about the cheaters was that their original, unaltered scores tended to be rather low. Dividing those scores by some integer>1 made them low, low, low. Cheaters back then were not only unethical, they were, well, stupid.Oh! Missed what you were asking. Catching one person cheating was not something our statistical techniques could handle. Those problems were handled by the professor.

 

[lisab]

How about changing homework assignments, and test questions every semester. That's what many of my profs did.

Hear, hear!

 

[rootX]

How about changing homework assignments, and test questions every semester. That's what many of my profs did.

Not practical: time intensive, solutions are often incorrect.

 

[D H]

Hear, hear!

But then how are all the frat boys and jocks going to graduate?

 

[JaredJames]

Not practical: time intensive, solutions are often incorrect.

Says a lot about the teacher (or person setting the test) then doesn't it.

My university lecturers change the questions every exam. So to say it isn't practical is non-sense. The choice is either have new questions each exam and reduce the risk of cheating or re-use questions and risk unrealistic test scores.

 

[rootX]

My university lecturers change the questions every exam.

I was not talking about exams but minor assignments: homework assignments, small tests etc. It is rare in upper years to grade homework assignments, neither there are any small quizes. It's only labs, midterms, and final exams.

 

[waht]

Hear, hear!

I hear the professors go over syllabus at the start of every semester, and particularly plagiarism and cheating policy.

 

[waht]

I was not talking about exams but minor assignments: homework assignments, small tests etc. It is rare in upper years to grade homework assignments, neither there are any small quizes. It's only labs, midterms, and final exams.

True, it's impossible to cheat in sciences and engineering. You either know how to solve a problem step-by-step or you don't.

 

[JaredJames]

I was not talking about exams but minor assignments: homework assignments, small tests etc. It is rare in upper years to grade homework assignments, neither there are any small quizes. It's only labs, midterms, and final exams.

OK, perhaps should have clarified. They don't give duplicate questions in exams, in-class tests or assignments.

The only things that remain the same are the lab reports, but they don't count for much of your final score (5%).

I took homework assignments to be assignments. (Not small pieces of work you are given each day/week etc, but actual assignments that count to your final result.)

We don't get weekly homework.

 

[D H]

True, it's impossible to cheat in sciences and engineering. You either know how to solve a problem step-by-step or you don't.

That's a bit naive. Computer science students will submit the same program, changing names and comments to hide the fact that they cheated. Similar kinds of cheating appear in all kinds of classes. Cheaters can be very creative in creating new ways to cheat. Old ways also work well. Fraternities are notorious for keeping copies of exams on file. If the professor doesn't change the exam every semester, all the members have to do is have a cheat sheet that tells them the answers.

Changing exams is hard. The questions need to be appropriate to the material. The exam should not be too hard or too easy. The instructor needs to anticipate all the ways that the question can be answered incorrectly so that partial credit can be given. This final factor makes developing a test an extremely difficult task.

 

[BobG]

Whoo boy. "Every cheats".

We used statistical techniques to detect cheaters 30 years ago when I TAed an intro computer science class. Our solution was simple: We graded the assignment or test as-is and then divided the score evenly amongst the cheaters. Four people cheated on an assignment: Their mediocre 80 turned into a 20. Two people cheated on an exam: Their fantastic 96 turned into a 48. Ouch!

That describes how you handled cheaters, not how you statistically detected them.

I'm curious how you could detect individual cheaters using statistical techniques. That kind of technique would come in handy in predicting when a baseball batter's hot streaks or cold streaks would come to an end, when a given poker hand was guaranteed to win, and in predicting when long strings of 'heads' would occur during coin flips.

The idea seems counter-intuitive.

It isn't completely off the wall since most questions can be evaluated by who misses them. Ideally, a question is a good test question if the high scorers (top third, for example) almost always get the question right, the average scorers (middle third?) get the question right a good percentage of the time, and low scorers are the most likely to miss a particular question. No one writes a perfect test and there will always be a few easy questions everyone gets right and a few hard questions even the good students are likely to miss. An individual missing the easy questions and getting the hard questions right would look pretty suspicious. I still don't think that would hold up as proof that particular student cheated.

 

[waht]

That's a bit naive. Computer science students will submit the same program, changing names and comments to hide the fact that they cheated. Similar kinds of cheating appear in all kinds of classes. Cheaters can be very creative in creating new ways to cheat. Old ways also work well. Fraternities are notorious for keeping copies of exams on file. If the professor doesn't change the exam every semester, all the members have to do is have a cheat sheet that tells them the answers.

Changing exams is hard. The questions need to be appropriate to the material. The exam should not be too hard or too easy. The instructor needs to anticipate all the ways that the question can be answered incorrectly so that partial credit can be given. This final factor makes developing a test an extremely difficult task.

Writing programs would count as homework, or some project. But the exam is still on theory, and bits of selected code that could be easily changed, swapped or renamed.

The classic example is a prof that rescheduled an exam for next week because he was still working on the exam questions. a lot of times professors improvise as they go, and some like to play around.

I'm not disputing the tough job that a professor has from preparing a lecture, delivering it and making sure everything is on track. In fact I respect it.

 

[rootX]

It isn't completely off the wall since most questions can be evaluated by who misses them. Ideally, a question is a good test question if the high scorers (top third, for example) almost always get the question right, the average scorers (middle third?) get the question right a good percentage of the time, and low scorers are the most likely to miss a particular question. No one writes a perfect test and there will always be a few easy questions everyone gets right and a few hard questions even the good students are likely to miss. An individual missing the easy questions and getting the hard questions right would look pretty suspicious. I still don't think that would hold up as proof that particular student cheated.

I love bell curve approach!

My most exams are such that you can take book to the exam and still fail the exam if you don't know your material. For engineering, it would take few hours (along with good sleep) to get prepared to pass the course but it is nearly impossible to individually cheat.

 

[Pengwuino]

Not practical: time intensive, solutions are often incorrect.

You change the numbers. The best thing about cheaters are that they're idiots most of the time. You change a few numbers in a physics test and MOST students, cheaters or not, will think you've made the problem completely different.

In regards to what someone mentioned earlier about people confessing even if they didn't cheat, here's the solution: Retake a test! Immediately (within a couple days) of the first one. Given a new test, if a student didn't cheat, their score shouldn't change by much. If someone did cheat, you should see scores plummet. Boom. Cheater.

 

[D H]

As I mentioned earlier, our techniques could not detect individual cheaters. Since it was 30 years ago, here is what we did:

For programming assignments we used statistics from the the compiler to find cheats: heap size, maximum stack size, and number of tokens. We sorted submissions, by hand (TAs are cheap), by heap size then by stack size. We used number of tokens and visual inspection to eliminate false positives (amazingly we had very few false positives). All of the standard tricks such as changing symbol names or inserting spurious comments didn't help the cheaters one bit.

For tests we selected the two questions of the type "write a program segment to ..." and that counted the most as exemplars. In addition to grading the answers we also had to count the number of words in the answers to those exemplars. We sorted by word count (binned sort/subsort). Total score and visual scrutiny helped us weed out the false positives, and once again we had very few false positives.

Easy to beat? Yes. Since this was a bust-out course, word on how to cheat didn't seem to filter down from year to year.

 

[DanP]

Florida college professor gives cheating business students an ultimatum - confess, retake the test and attend an ethics class, or don't graduate.
http://news.yahoo.com/video/business-15749628/students-busted-for-cheating-22954742

Everyone doesn't.

Frankly, you should be stupid to confess for something which cannot be proven. Mind some posts above , heuristics are not proof. And no college professor should be given the arbitrary power to force students to retake a test. One test. Its your fault you are not able to supervise the test. Don't harass anyone with retakes to excuse yours or your staff incompetence. You as a teacher failed the university, failed the students and yourself if you are powerless to stop cheating in ONE exam.

You was cheated by X hundred students ? Your fault. Eat it up, and learn your lessons for next year. Dont complain.

 

[zomgwtf]

Wow, that's disgusting.

What really astounds me is the kid justifying it with something like "everybody cheats, it's just a witch hunt". Seems like he has some serious issues. I can truly say I've never, ever cheated on tests. And I figure the business world is better off without people like that. I swear you'd have to be some kind of sociopath to say something like that.

I feel really sorry for the professor. I also feel really sorry for the ~ 400 students that have to retake the exam because 1/3 of the class are cheaters. The prof is being really nice in allowing them to take the test again and take an ethics class, IMO.

It makes you a sociopath to say that everybody has cheated before? lol...?


I'd say that the majority of people in life have definitely cheated in school at some point in time. The vast majority even. I know that I did back in math, make a cheat sheet with worked examples I knew would be on the test slide it in my calculator. Or even just using my calculators ability to solve quadratics and cubics and even higher powers to check my final answers is cheating.

Anyways the students got A HOLD of the final answers. That's the profs fault. If I somehow got a hold of answers to an exam I sure as hell would use them. I didn't hear anything stated showing that students stole the answers or something, they just somehow got them. I remember in high school math tests were the same year to year, is it cheating that I could study off my older friends tests?

As well these methods used to catch the cheaters seem kinda iffy. I do not think there's any way to catch a cheater by using statistics. You either catch them in real life or not. Not in theory.

 

[Pengwuino]

It makes you a sociopath to say that everybody has cheated before? lol...?

According to *random source I forgot*, one of the characteristics of sociopaths is that they believe everybody else is just as bad as them but won't admit it.

Anyways the students got A HOLD of the final answers. That's the profs fault. If I somehow got a hold of answers to an exam I sure as hell would use them. I didn't hear anything stated showing that students stole the answers or something, they just somehow got them. I remember in high school math tests were the same year to year, is it cheating that I could study off my older friends tests?

As well these methods used to catch the cheaters seem kinda iffy. I do not think there's any way to catch a cheater by using statistics. You either catch them in real life or not. Not in theory.

This kind of thinking scares the hell out of me. If someone accidentally leaves their keys to their car on the door and someone steals it, the person who stole the car is off the hook? Well I better start installing a 5 ton safe in my house for my valuables because apparently if I don't, it'll be my fault when someone steals them.

Can you imagine using this logic in the real world? An investor looses his retirement because his mutual funds bought a bunch of mortgages that were crap? Well that's too bad, the investor should have gone out and gotten appraisals for every house that had a mortgage purchased to it. Someone threw out a bank statement with important information and had their identity stolen? Well if someone was rummaging through their garbage and ran into it, they should feel obligated to steal the persons identity!

Ridiculous. Stop trying to rationalize things you know were wrong to do.

 

[Grep]

It makes you a sociopath to say that everybody has cheated before? lol...?

I pretty much agree with Pengwuino. Though I'd just add that I in no way imply a serious diagnosis and I'm not a psychologist, etc. But I seriously think there's at least a little bit of something wrong with someone who would talk like that kid. A lot of people have a good sense of ethics. Two thirds did indeed do their own work (leaving aside the validity of the teacher's detection method, which I don't know).

Anyways, I'd kill to be able to go to University and study right now. It's a privilege to be able to go there and learn, and students who cheat are only cheating themselves, IMO. If they learned it, which is pretty much the point of the whole thing, they wouldn't have to cheat. On this, I'm guessing, we can all mostly agree. But it's really the way I feel about it.

 

[Vanadium 50]

There seems to be an assumption that the cheating involved a recycled test. Is there any evidence for this?

The worst case of "bulk cheating" I am aware of involved some students breaking into the professor's office to get a copy of the exam, not test recycling.

 

[Pengwuino]

Here is the lecture where he confronts everyone.

 

[Vanadium 50]

Doing a little reading on the web, it appears that one student handed in the answer key with his test. It also appears that the test is given by computer in a central testing center, and a third party who is somehow involved in this operation was selling the answers: the high-tech version of bribing the secretary who mimeographed the test.