How to Catch a Cheater

http://nerdlife.net/how-to-catch-a-cheater/

by bkrausz • 16 years ago

138 98 16 years ago

dot• 16 years ago

run a proxy server company? this would be perfect material for a viral campaign.

protomyth2• 16 years ago

I was actually wondering how he would get the access records for the school's PCs? It would seem that it would be a big bluff as most Sys Admins would give such records to a prof.

enjo3• 16 years ago

At least at the two schools I'm fairly familiar with, when investigating a cheating allegation those are routinely handed over.

aplusbi• 16 years ago

That sounds terrible. If I saw something like that, I'd probably do the problem, then look up more information on it. I'd probably start with wikipedia but move onto google (well, ddg these days) if I didn't find anything.

The sort of policy that does not allow one to do basic research is pretty anti-academic if you ask me.

docgnome• 16 years ago

I agree. In a similar vein, I took a bunch of electronics classes and wasn't allowed to use my TI-92+ on the grounds that other students didn't have them. I though that was the biggest load of crap especially for a program that was geared towards workforce training. My employer is hardly going to say "I need you to do X, and oh by the way, if you try to use all the resources at your disposal, you're fired."

bkrausz5• 16 years ago

The early homeworks in this class (Great Theoretical Ideas in Computer Science) were basically puzzles, where if you've seen the style of problem before you more or less knew the answer. The goal was to expose us to these types of problems and teach us how to think them through properly.

It was definitely one of the most fun courses I've taken.

spicyj• 16 years ago

15-251 isn't a freshman-year course, is it?

flogic• 16 years ago

That's one of those critical to include details. Otherwise the story makes the prof sound like a douche.

petercooper• 16 years ago

The policy's wording also bothers me. There's a massive difference between searching for "the answer" and doing some research.

wwortiz3• 16 years ago

> We were explicitly told that searching online for answers to homeworks was forbidden.

I might be in the minority but I am of the opinion that if you are grading students on something that can be easily obtained through a google search you probably shouldn't be grading them on that.

michael_dorfman• 16 years ago

What is your justification? Anything that can be found on Google isn't worth learning?

Just to take a trivial example: take a look at Project Euler. There's no question in my mind that working through those exercises is a valuable learning experience, and (in the context of a programming course) would be worth grading someone on-- even though the answers are readily available to anyone who searches for them.

wwortiz3• 16 years ago

>What is your justification? Anything that can be found on Google isn't worth learning?

How in the world you got that out of my comment is beyond me.

People cheat with resources available it is a sad fact but it is true. You are being unfair if you decide to punish honest students by grading on something that is easily available (I've seen homework grades as 30% off class grades before).

Xk3• 16 years ago

I agree with you for the most part, but nearly all introductory courses must have content that is easily available online. Consider teaching students how to write a recursive method. What would their first assignment be? List length? "find length of list" Factorial? "factorial recursive" Fibonacci? "fibonacci recursive" GCD? "program gcd"

There are only so many simple recursive procedures, and I guarantee nearly all of them are online.

That said, I do think that once students have a basic understanding of computer science all assignments should be given so that the answer are not online. And this wouldn't bee to hard, even. It would just take some effort by the professor.

wwortiz3• 16 years ago

Okay sure those are simple examples that are probably available in large quantities online, but why should that homework assignment count towards a grade is more of my point. Professors can assign homework that isn't graded and the ones who actually do the assignment will be the ones that learn.

btilly• 16 years ago

It is trivial to create a simple recursive problem without saying anything that would be a Google-able search term. The problem may be solved, but the students won't know what to search for.

For example have them write a program that solves the game of Nim, but don't tell them the name of the game they are solving. Instead call it "pick up pennies" or something like that. Now Google won't help.

pohl2• 16 years ago

I suspect you're actually in the majority. I bet that, rather, the minority position is that the instructor should create whatever conditions they want to so that they can exercise the students as they see fit - and that the students should extend enough respect to consider why those conditions are being created. This is my view.

Replace the teacher with a coach in a weightlifting class. One day he wants you to lift some freeweights, but he wants you to lift them in a particular way that does not use your legs. (Let's say he's having you do "military presses".) "Nonsense!" says the modern student, who already knows that it's best to lift heavy things with their legs. "Why shouldn't I be able to use this awesome resource (my legs) that's right here at my disposal!?"

Well maybe - just maybe - the coach wants to target your deltoids. Hmmm.

Enter Professor Luis von Ahn, who wants you to do homework without using google. Perhaps that is because he doesn't want you to target and improve your research abilities. It could be that, rather, he would like you to exercise and sculpt your ability to work problems from first principles. The professor may be aware that practice makes permanent, and homework is intended to get students to strengthen specific things, and that taking the easy way out robs the student of the intended benefit.

That's how I approach being a student, and so I grant that when sensei says wax-on, paint-uh-fence, and sand-uh-flo that he's got some idea of what I should be working on. But that's an old, old attitude that exists only in a handful of Hollywood movies these days. So that's why I'm betting you're actually in the majority.

By the way, the instructor here is an awesome guy. Anybody who hasn't heard of Luis von Ahn should do themselves a favor and watch his presentation on Human Computation:

http://www.youtube.com/watch?v=dtFroEJN1nI

bkrausz5• 16 years ago

Can't upvote this enough.

1) The early homeworks in this class were basically puzzles, where if you’ve seen the style of problem before you more or less knew the answer. The goal was to expose us to these types of problems and teach us how to think them through properly.

2) The class was taught by Luis just as he was beginning to hit mainstream media. The guy has a unique approach to problems that was absolutely inspiring. His lectures were one of the few that I actually made an effort to go to.

jemfinch• 16 years ago

> the instructor should create whatever conditions they want to so that they can exercise the students as they see fit

I don't disagree, as long as the instructor is able to defend the pedagogical purpose of the limitations he demands. The OP is essentially contending that few situations warrant such limitations, and I agree with him about that.

xenophanes2• 16 years ago

Many math teachers ban calculators on math tests. I'm pretty sure this has majority support, not minority.

Google does simple math, so that's basically an equivalent issue.

IMO it's dumb -- in real life you have access to Google. But I mostly wanted to say I think there's clearly majority support for artificial grading conditions, since the no-calculators thing is so popular.

enjo3• 16 years ago

I had a CS professor who expressly prohibited the use of graphical debuggers for much the same reason.

jamii• 16 years ago

For my MSc exams we were given the papers to take home and it was considered perfectly acceptable to look up the answers as they as they were correctly cited. Someone clearly decided that the easiest way to prevent cheating was to set questions that took more than a few seconds googling to answer.

danielford• 16 years ago

That can be really hard in certain situations. My background is in biology. One of the more interesting and fun assignments I was given in college was a series of case studies. Basically, the patient has symptoms X, T, and W, what is wrong with them?

This is a situation where you can't give problems to an undergraduate student that are easy enough to answer while they're simultaneously impossible to Google. I'd love to give my own students the opportunity to experience these types of critical thinking questions. Unfortunately, I know that I'm completely incapable of assigning ungoogleable problems without turning them into some Dr. House shit that can't be answered by a college student.

dkarl8• 16 years ago

My AP (Advanced Placement) Chemistry teacher noticed some cheating on a major chemistry test we took. The cheating was pretty obvious; some kids just decided they were going to fail the test so they might as well have some fun. A week later, after a week of lecturing on a new section of the chemistry book and no mention of our scores, a student asked. No, he hadn't graded the tests. He was sorry, he had been really busy. After that, students started asking him every day if he had graded the tests.

Two weeks after we had taken the test, he said, "I have to come clean. The reason I have not graded your tests is that I lost them and have been trying to find them for the last two weeks. I'm very, very sorry, it was entirely my fault, and as a result" (he produced stack of paper from behind his back) "you all have to retake the test right now. Put your books away, take a test and two sheets of scratch paper, and start working."

We didn't have as many students sign up for the second semester of AP Chemistry. Not many people failed the first semester, and scores in AP classes were given a full letter grade boost compared to regular classes, but students knew he wasn't messing around. As a result, it was one of the best AP classes we had, and lots of kids got good scores on the AP test.

Semiapies15• 16 years ago

I wonder how many perfectly honest students did worse on the impromptu re-take two weeks later, with entirely new material on their minds.

joezydeco2• 16 years ago

If the new material built on the old, shouldn't the old material be a solid lock?

tomjen32• 16 years ago

Agreed it seems that it is always the honest students who gets the short end of the stick.

Thats a good thing though - it prepares the students for the real world.

rms• 16 years ago

In terms of actual learning, rather than the sort of grade optimization, it probably was valuable in terms of spaced repetition. http://en.wikipedia.org/wiki/Spaced_repetition

dkarl8• 16 years ago

Good question -- I honestly don't remember. He only wanted to punish the cheaters, so I imagine he curved the test. He often added a fixed number to everyone's score to bring the best or second-best score up to 100. (It's not really "curving," but that's what we called it.)

Robin_Message• 16 years ago

I think that's a great thing — not only have you caught and disadvantaged the cheaters, but you've also screwed the crammers as well.

You either know the material, or you don't know the material. The timing, location or preparation for the test is irrelevant. The test is irrelevant.

jeangenie• 16 years ago

my university's general chemistry series had a problem with kids loading their TI-83s with "notes" to remember what oxidized and what reduced (and much, much more).

it was crude and pretty obvious but extremely effective and no one in my class was ever caught.

lazyant• 16 years ago

It seems this professor is more interested in entrapment and showing how smart he is rather than in his students learning (OK, maybe this was the lesson)

rick8883• 16 years ago

..or catching cheaters. The students that cheated obviously don't care about learning or education, so why should he?

cousin_it• 16 years ago

Why not just put an incorrect solution online?

ultrasaurus• 16 years ago

Agreed, that's what I thought the punchline would be. A sample solution that only works with the sample data provided. (something along the lines of x*y = x+y, with the sample data being (0,0) and (2,2))

ecaradec• 16 years ago

I remember professors doing that kind of things when I was a student then declaring that students didn't studied well, didn't followed their rules, cheated, or anything else.

I don't think it help students to progress it any way. It just feel like you are fighting against a bureaucraty which has no rules, or which has rules that will only be revealed until it's too late. What they are creating is traps, and traps don't teach anything because they are random.

Please teachers, read Richard Feynman stuff, he had such a joy to teach his students and stop bullying your students.

sachitgupta2• 16 years ago

Wow - this seemed strangely familiar, until I realized I was in the class! What sucks is that a lot of people found the solution and showed it to other classmates, but only the 1 person who found it got into trouble. (No, I wasn't one of the students who found it).

Semiapies15• 16 years ago

This is a neat little sting, but looked at honestly, it's entrapment.

"Commonly referred to as The Glorblar Problem"

That phrasing is a cue that the student is expected to either look something up or to refer back to notes or memory of lectures or discussion of the Glorblar Problem in class. Why else mention it, if it were an honest homework problem?

We know they didn't hear about the Glorblar Problem in class - it doesn't exist. For the same reason, we know that hitting the books, either back at the dorms or in the library, wouldn't have turned up anything.

Remember that this is homework, not a test. Yes, presumably the problem is doable without any reference material. The students don't necessarily know that, though. They do know that they've been prodded in that question to at least check their thinking or work against external information....which isn't available to them through the channels they've been permitted to use.

So, the professor deliberately tempted people to look up something online, then amused himself by busting the people who did. Cute method, but it proves nothing other than "schmuck bait works".

Semiapies15• 16 years ago

Interesting - either someone thinks it's good for an authority to encourage people to do things against the rules in order to catch and punish them, or else someone never picked up on suggestive question phrasing in school.

rick8883• 16 years ago

"So, the professor deliberately tempted people to look up something online, then amused himself by busting the people who did. Cute method, but it proves nothing other than "schmuck bait works"."

The way I see it, the people that are actually doing their homework wouldn't even be tempted by it..because they are actually doing their homework.

"Remember that this is homework, not a test. Yes, presumably the problem is doable without any reference material. The students don't necessarily know that, though. They do know that they've been prodded in that question to at least check their thinking or work against external information....which isn't available to them through the channels they've been permitted to use."

Checking your work against the correct answer and Googling the answer (and just writing it down) are two different things.

Semiapies15• 16 years ago

Thank you - this is the most vapid thing I've seen today not written by a TSA official.

Especially:

"Checking your work against the correct answer and Googling the answer (and just writing it down) are two different things."

And you can tell the difference based on a server log...exactly how?

zeraholladay• 16 years ago

I don't know if it even qualifies as "schmuck bait." The Prof's actions are ethically objectionable (and personally disgusting to me) since he actively solicited unethical behavior from gullible students. Frankly, it's not a challenge to bring out the worst qualities in people, especially scared undergrads.

vkalladath• 16 years ago

Looks like you guys didn't read the whole story. The professor specifically asked the class not to search online. The goal of the homework was to get the students to teach them to think through the puzzle.

Semiapies15• 16 years ago

Looks like you didn't actually read what I wrote - well done.

ETA: Or, charitably, you didn't recognize that the big UPDATES at the top of the blogger's post indicate that since my original post, he thought to clarify the context and indicate that he'd actually misremembered the phrasing of the question.

Or maybe both?

I wouldn't consider the actual professor's question from the linked exam to be entrapment, as it happens. That in mind, I'll save my disgust for the people who thought the scenario of entrapment originally described was so terribly clever.

torme• 16 years ago

This seems to severely punish those who would even attempt to follow the rules. I've had instances where I can't find the answer to a problem and will spend hours hunting it down. If I had a professor give me a problem with that constraint, I'd probably scour a book for hours. What are you other options? Put "I don't know."?

Seems like this just forces everyone to the same result: those who break the rules will and save themselves time, those that try and follow the rules will probably break after hours, and end up googling anyways.

Maybe I'm missing something but this seems totally absurd.

daimyoyo• 16 years ago

The true lesson here is simple. If you're going to do something online you shouldn't, get behind a proxy(or 7).

phalien• 16 years ago

Or you could just watch Lie To Me and learn to interpret microexpressions :)

http://en.wikipedia.org/wiki/Microexpressions

xenophanes2• 16 years ago

He should have put the wrong answer on the website. Anyone who just copies it will get a 0 for being wrong. Anyone who actually thinks about it and notices the difference shouldn't fail, imo, since that's appropriate, thoughtful use of the internet.

WillyF• 16 years ago

During my Freshman year of college (2002), I was taking Econ 101. It was a class of probably 500 students, and our tests were multiple choice. By the time I got back to my dorm room, the answers to the test were always up on the course website (I'm not sure when they actually were posted).

After the first two tests, the professor must have noticed that something wasn't right. I finished the third test quickly, and was pretty confident that I had done extremely well. We used scantron sheets for our answers, so we got to take the paper with the questions home after the test. I had marked all of my answers on both sheets, so I loaded the course website and started checking my answers.

My heart sank as I went through the answers. Somehow I had managed to get every single question wrong. I figured that there must have been something wrong with the answers, and when I matched the answers (not the letters), I realized that I had actually gotten every question right.

The prof swapped the letters, so the kids whose friends were texting them the "answers" got screwed. It was pretty funny, even if it increased my blood pressure for a little bit.

lotharbot• 16 years ago

For freshman year math courses, we wrote two versions of each test. They'd be nearly identical, but with a few of the numbers and variable names modified. We'd alternate passing them out so that people sitting next to each other had different tests. We'd also go around the room and write down who was sitting where. This made it really easy to catch the "look at your neighbor's paper" sort of cheating. (It's really hard to text answers to a calculus test, and taking a photo of your page would be really conspicuous.)

Other departments had their own anti-cheating methods. Kids coming straight out of high school, who were used to less sophisticated cheat-detection methods, tended to get burned pretty bad their first semester.

nohat• 16 years ago

So he was going to subpoena their isps? Only the people who used lab computers needed to worry.

protomyth2• 16 years ago

The sad part is you could make quite a case for being slanted against poor students who don't own their own computer. Not saying I would agree with it, but it wouldn't surprise me.

bkrausz5• 16 years ago

If you're in CMU's network your computer is registered with them, so unless you live off-campus (which Freshmen aren't allowed to do) you can easily look up which student registered which machine (plus it's usually obvious from the hostname).

kbatten• 16 years ago

I guess the lesson is that real life is not at all the same as university life.

Semiapies15• 16 years ago

Here's a funny thought - what if someone did the assignment, put it away, then wondered, "Huh, so why is the Glorblar Problem notable?"

Dirty, dishonest intellectual curiosity - you have to catch those people, right?

hdctambien• 16 years ago

I think some people are missing a few critical points of this exercise:

1) the homework was supposed to work out your thinking muscles. Getting the "right" answer wasn't the point of the assignment as much as trying to discover how to obtain an answer. (much better described in these comments by pohl)

2) On crying "entrapment": See Orpheus and Eurydice (http://en.wikipedia.org/wiki/Orpheus#Death_of_Eurydice). There were rules clearly put in place and whether they made sense or not, if broken carried with them penalties that the perpetrators were well aware of. If that doesn't smell like the "real world" I don't know what does. With that said...

3) Universities are NOT the real world. They barely resemble the real world. The point of university has never been to "prepare you for the real world."

Semiapies15• 16 years ago

As for 1, I think you're missing that a lot of people responded before the blogger updated the post (twice) in order to explain those points, including that he'd misremembered how the question was phrased.

As for 2, heaven forfend anyone apply ethics to the actions of those in authority.

As for 3, well, you've just cancelled out your second point.

gnubardt• 16 years ago

It really bothers me when people cheat in engineering classes, or really anything where your competence (or lack of) could affect someone's wellbeing.

PostOnce• 16 years ago

If the point of college is to learn rather than to simply get a degree, then after you do you work, you should check it, to see if you got it right, and if you failed, try again.

So, he doesn't want students to check their work, eh? After that first assignment, I'd cheat by every method I could possibly think of at every opportunity, because at this point, he has made it about grades, rather than about education.

The only time you can't go back and re-work and check the internet for help is during a test. You're SUPPOSED to look things up with homework, homework is for learning before a test. After all, the whole point of homework is to prepare yourself for a test, if you don't do the homework, if you just copy the answers, you won't be ready for the test. So what exactly is he achieving by not letting students get the maximum value out of homework?

enjo3• 16 years ago

General cheating story:

My wife is a professor (ya that again:>). As an accounting professor they do a lot of their homework/projects in excel. Excel, of course, saves the information of the document creator in it's meta-data which is trivial to look-up. She'll routinely get 5 or 6 of the exact same file (with the same origination data) for the same project. Overall, she'll catch 10%-20% of her class cheating in a given semester without even really trying.

pg7• 16 years ago

You guys in US do it all the wrong way. I remember how examiners fought cheating when I was a student at Warsaw University, Poland (1990 - 1995). They simply allowed cheating. So, at the beginning of every exam an examiner announced something like: "Ladies and gentlemen. You are allowed to cheat. You can browse your books and notes, no questions asked. But I must warn you that exam tasks are designed such a way, that they require thinking - and thinking is not compatible with cheating, because thinking is fast and cheating is quite slow. So, you are allowed to cheat, but if you will it is very likely that you will not complete minimum number of tasks required to pass the exam. I wish you good luck."

jchonphoenix• 16 years ago

My year the puzzle was called "Giramacristo's puzzle."

And Luis ended up telling everyone he was going to give them 0's and they'd take a huge hit in their grade, but he ended up not actually factoring that into the grades.

How to Catch a Cheater

Related Stories