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.
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.
> 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.
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:
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.
IMO, the most economically valuable lesson an above average student learns at a top collage is how to cheat effectively. Four years is no where near enough time to provide a good education, but plenty for a gentle wake-up call to how the world actually works.
Edit: Now here is someone that was actually paying attention: I believe it was called Finkleberg’s 101 game, at least when I took the class. Maybe he pulled the same trick with a different problem several times. I remember thinking “this is such a googleable problem” because of the peculiar name so I actually assumed that a google search for the name would be a honeypot trap. My assumption proved correct.
PS: There are plenty of great resources at a top institution to learn, but you can only explore a tiny fraction of them.
Would you think that one of the most economically valuable things universities do is kicking out cheats? Because they are the ones who will (most likely) back-stab their way to a CEO position, then run down the company while still paying themselves out-sized bonuses.
> 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.
as long as the instructor is able to defend the pedagogical purpose of the limitations he demands
I just had a thought over my morning coffee: I would agree that the instructor should be able to defend such choices to the head of the department. The instructor may choose to defend the choice to the student, but the instructor should only be required to defend the choice to the student if the student is also required to be able to defend why they should be allowed to be there. I fear the most precious thing we've lost in our modern society is a student's willingness to bring themselves to the process, and so if I (as an instructor) must defend why I've set boundaries on an exercise then then you (as the student) must be able to defend that you really want to learn what I have to teach.
Alas, higher education has been reduced to a transaction for a service, and "the customer is always right".
few situations warrant such limitations, and I agree with him about that.
I have the opposite view. Most of the homework that I've ever seen (as a Math & Comp Sci double major) was about the benefit of the exercise, and very rarely about strengthening research abilities. But, then, I was in school when Tim Berners-Lee was gestating the world wide web concept at CERN.
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.
> IMO it's dumb -- in real life you have access to Google.
If one is taking a class merely to learn how to Google for the answers, then one should not be in a university. One should be in a trade school which only teaches you enough of your fields of study to be able to figure out the right terms to Google for.
When you take university classes, the goal is to learn to be the people that provide the solutions that the trade schoolers will Google for.
I suppose it depends on the university, department and teacher, but we were allowed to use calculators on math exams. Of course, they weren't helpful at all - if you knew how to solve the problem the calculations were so simple that you didn't need a calculator, if you didn't, you couldn't solve the problem no matter how good your calculator was.
For similar reasons, we were also allowed to bring in any books, notes etc. to exams. If you didn't have a good understanding of the subject, they weren't helpful at all as there was no time to start figuring out things, but if knew stuff, you didn't need the books and lecture notes at all or only for reference.
If you have access to Google, you have access Wolfram Alpha. However, the point of math courses is not to demonstrate that you know how to punch buttons, but that you know the methods and algorithms behind those tools.
Calculator bans aren't nearly as common now as they were ten or fifteen years ago. Pretty much all the major exams allow them, and questions are often written with this in mind.
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.
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.
However, there will be someone in the class who knows that "pick up pennies" is just the game of Nim, and will google it. Renaming makes it harder so fewer people will be able to cheat, but it doesn't make it impossible. (That said, it wouldn't be all that hard for the professor to create his own game which is similar to Nim in some respects, but different in others -- I believe this would be the best solution -- but unfortunately most professors would not put the time in to do this.)
This problem multiplies tenfold when you try to do it with math. (Integrate _____ -> WolframAlpha. Prove ______ -> Google.)
There's a big gap between "not all that hard" and "something you have time to create from scratch for every problem on every exam in all of your classes every term".
In general, creating good, well-balanced problems isn't easy! And many of the "classic" examples are classic precisely because they incorporate just the right blend of concepts in an unusually simple package. That's not easy to duplicate, if it's possible at all. (I do my best to write exam problems that way, but it can be a tortuous process.)
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.
Graded homeworks might be used as checkpoints verifying that students are learning key concepts, prior to the exams that are a larger component of the grade. Without these graded checkpoints, lazy students given only voluntary homeworks might reach the exam with little preparation
(lazy student speaking from personal experience. being graded on the homework 'encouraged' learning the concepts!)
lazy students given only voluntary homeworks might reach the exam with little preparation
And that is bad, why exactly? You're unprepared, you fail, you learn a lesson, you retake. Repeat until done. I too was a lazy student and found that failing a few classes and having to resit the exam was a fine way to motivate you to study.
Agreed, but that all but eliminates graded homework for the first several years of college (not that that's a bad thing, just that I don't think colleges would approve).
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.
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.
The solution to this is to reserve a time and space (maybe with a TA) when/where the students are expected to do the work, instead of doing them at leisure.
Few people would say it's wrong to not be able to google during an exam, but most would agree that not googling during exercises is insane.
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.
>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).
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.
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.)
I think it perpetuates the ideas of how to wield authority. The concept of enticing people to cheat in order to be able to catch people at cheating is of a piece with "anti-terror" operations where federal agents set up fake terrorist cells, recruit feckless malcontents, and guide and cajole them until they can get them to do something illegal.
How is doing this preparing the students for the real world (I'm being serious and not condescending or anything)? I don't see the real world being particularly more difficult than normal high school. Example - didn't file or pay taxes on time? No problemo - just take a penalty and finish up your work late.
You've got it backwards: the students who studied honestly had much better retention after two weeks than the kids who wasted time trying to cheat or game the test. It was the teacher's intention to reward honest work by giving the hard-working kids a big advantage over the cheaters, and that's exactly what he accomplished. If only life was always so fair!
Your explanation is for a fantasy-land. Studying honestly in many AP classes comes down to cramming. Here's how Semiapies summed it up in a comment in this thread:
"In high school and particularly AP, those courses are blatant cycles of 1) cram, 2) test, and 3) forget all but the core of what that section was about.
(If you want to say, "You should learn it all for the final" keep in mind nobody does, nor are final exams built around remembering everything from the whole course.)"
--
It's not like the crammers are even bad kids. A lot of them have no choice because if they didn't study that way they wouldn't have enough time to study and will end up getting worse grades.
This would mainly show people that the world screws you over and people who are part of crappy systems will go ahead and screw you over on top of the crappy system. I see little fairness in this situation.
Semiapies' description of AP classes is really bizarre and can only reflect a situation in which the teachers did not give a rat's ass about their students' AP scores. The nice thing about an AP class versus a non-AP class is that the AP test gives teachers an excuse to care whether their students are actually learning or not. Naturally that means the kids end up working a lot harder, but teachers who care at all about the scores at the end of the year (and mine did) will do everything they can to discourage cramming and other fake learning.
Kids who can't pass the class without cramming and who aren't going to retain the material shouldn't be in an AP class in the first place. They're going to run into a concrete wall at the end and waste their money (or the school's money) taking a test they aren't prepared for. Some students can handle it, and that's who the class is for. In my graduating class, there were maybe a dozen kids who got a 5 on at least one AP test, and I think there were two dozen who got a 3 on at least one AP test. That's out of a graduating class of a little over 300. (The entering class was over 400, so you can tell I didn't go to an especially good school.)
Only then, and most courses are not straight progressions.
In high school and particularly AP, those courses are blatant cycles of 1) cram, 2) test, and 3) forget all but the core of what that section was about.
(If you want to say, "You should learn it all for the final" keep in mind nobody does, nor are final exams built around remembering everything from the whole course.)
Good AP courses are completely the opposite, and I had several in high school. We had regular reviews to make sure we refreshed and retained the material until the end of the year.
The great thing about AP classes was that the teachers could really teach, and they could justify themselves by pointing to the AP test if anyone complained about the rigor of the work. The students were game, too, at least the ones who didn't drop the classes. Everybody sincerely tried to understand what they were doing instead of just trying to make the right pencil marks on the test paper, since understanding was the most efficient way to retain something for months at a time. The test looming at the end of the year rendered cramming pointless -- you could cram for one test, but you would have to eventually learn the material anyway in order to take the AP exam, so cramming for a test actually felt like the failure that it was.
By "cramming" most people mean the kind of studying that enables you to remember information for one or two days, which would not result in good AP scores. Are you sure by "cramming" you don't just mean "intense studying?" If you and your classmates got good AP scores, then you must have understood most of the material well enough to retain it, since you would only have had time to cram a small amount before the test.
You misunderstand. Please note that from the start, I've allowed that people do retain some knowledge from cramming-based classes. (For that matter, someone writing notes to sneak into a test probably gained some small benefit from the act of writing the notes.)
But no, there's no mistake, here. Much like other high school classes, the AP classes I took were sequences of sections followed by intense "reviews" the day prior to each test. They were the very model of how schools encourage cram-and-forget education.
I really don't understand how it could be "cram-and-forget" unless the students failed the AP tests. If they did well on the AP tests -- covering an entire year's worth of material in one day -- it was more like "cram-and-remember" and it was apparently a good way to learn the material.
I'm not sure what your difficulty is in understanding this. They cram before the tests in the class, then they cram before the final exam (or AP exam). The "forget" is after the AP test.
This naturally made students who crammed have a tougher time in later classes in college. I wasn't one of them, but I refuse to look down on the people who were behaving exactly as the teachers were training them to behave.
I've taken 5 APs and none of them were cram/test/dump. There's one large test at the end of the season. It would be counterproductive to forget each chapter as you go.
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.
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.
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."
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.
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".
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.
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.
"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.
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.
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.
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.
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?
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."
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.
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))
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)
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).
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).
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.
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.
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.
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."
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.
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.
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.
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.