Marginal Revolution: The Law of Below Averages

July
S	M	T	W	Th	F	St
		1	2	3	4	5
6	7	8	9	10	11	12
13	14	15	16	17	18	19
20	21	22	23	24	25	26
27	28	29	30	31

« Dark horse picks for the Great American Novel | Main | The Great American Novel -- my pick »

The Law of Below Averages

I sometimes find evidence of cheating on exams but I rarely take action, I don't have to. Almost invariably the cheaters get abysmally low grades even without penalty. Some people I know get annoyed when students without evident handicap ask for and receive special treatment such as extra time on exams. I comply without rancor as the extra time never seems to help. Over the years I have had a number of students ask for incompletes. None have ever become completes.

I call this the law of below averages.

Addendum: Any student who attempts to take advantage of my lax attitude should first reflect on the Lucas Critique. Comments are open if you have experiences to share.

Posted by Alex Tabarrok on February 10, 2006 at 07:12 AM in Education | Permalink

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/t/trackback/3576/4232522

Listed below are links to weblogs that reference The Law of Below Averages:

» The Law of Below Averages from Mike Linksvayer
I probably only noticed Alex Tabarroks post in my feed reader this morning because of the title similarity to Nathans the law of averages blog. The former has some amusing stories in comments of student cheaters foiled by their own stupid... [Read More]

Tracked on Feb 10, 2006 1:07:31 PM

» Threat Models of cheating from Not Bad For a Cubicle
Alex Tabarrok writes over at Marginal Revolution: I sometimes find evidence of cheating on exams but I rarely take action, I dont have to. Almost invariably the cheaters get abysmally low grades even without penalty. Some people I know get ... [Read More]

Tracked on Feb 16, 2006 12:54:01 PM

Comments

I like the name. It describes my policy toward missing midterm exams. A student e-mails minutes before the midterm to say he has the "stomach flu" or that his elderly grandmother has passed away, can he be excused from the exam? I say sure. But there is no make-up exam. The weight of the missed exam is applied to all remaining exams. (This policy is on the course syllabus and announced first day of class.) In my experience, students
hate this; they hate the "pressure" of high-weight exams. I've seen some truly remarkable recoveries from the stomach flu and other maladies. And the students who don't recover, including ones who may have phony excuses, almost always don't do well on subsequent tests. I don't feel I am being taken advantage of, especially since I view midterm exams as for the students' benefit, not mine.

Posted by: Craig Newmark at Feb 10, 2006 8:06:48 AM

2 cheating stories (lengthy post, but good stories):

I caught a student one time, but didn't do anything about it on that test. I always give the same tests (same questions, same order) to the entire class - on the record, this eliminates any effect the order of the questions may have on student performance (off the record, it's easier to grade one set of tests than two). Anyways, it was a multiple choice test. One person had the exact same answers as another, and I knew where they were sitting in the class. On the next exam, I kept the questions in the same order, but changed the order of the answers on one test. Using slate of hand, I dealt my alleged cheater the test with the different ordered answers off the bottom of the stack of tests. He ended up with something like an 18 (and had nearly identical answers to the same person as he did on the first exam), which balanced out the 80 something he received on the other test. Problem solved, and no paperwork, and no hearings.

Another time I inadvertently thwarted a cheater. I like to give my class the test and answers back so they can look over what they've missed. I post the answers online. One student missed the exam, and was taking a couple of days (it was a summer class, so it was actually about 3 class meetings) to get in touch with me to take a make up. To make a long story short, I posted the answers online before she took the make up (my "make up exams" are identical to the regularly scheduled ones). She took the exam in about 10 minutes, and when I graded it she had gotten something like a 12. What I found out that day in class from one of my other students was that I had accidentally linked the answers to test 1 to the link where the answers to test 2 were supposed to go. I regraded her exam, and she would've gotten a 96 if I used the test 1 answers to grade her exam rather than the test 2 answers. Again, problem solved (although accidentally this time), and no paperwork, no hearings - and after she failed the class, she called me up to ask what her 2nd exam grade was, figuring she received an A. When I told her 12, what was she going to say, how is that possible I used the answers you posted online? Classic.

Posted by: AZ at Feb 10, 2006 8:16:07 AM

My favorite cheating story: some years ago, when I was a math TA, another math TA came to me to ask how to handle this situation: one guy had copied from another on a calculus exam. Thing is, the cheater had picked his target poorly: he was the 2nd-worst student in the class. And, of course, the cheater didn't have enough time to copy everything.

Basically, ignoring the cheating and grading the exam like all others, the cheater's exam got an F and the target of the cheater got a D. I told my friend that ist seemed the problem took care of itself.

Posted by: meep at Feb 10, 2006 9:27:59 AM

i'm a real life below average student, although i don't cheat and wouldn't try to ask for a test extention. so while i'd like to think that there is an additional advantage to your approach, (the chance that the student, not having the professor to blame, might take a step toward responsibility), i'm not sure its true for this n=1. i think the problmes that make one a poor student often require a shrink, not just a professor.

Posted by: anon at Feb 10, 2006 9:37:36 AM

I've found that this policy generally helps your student evaluations as well.

Students are convinced that those marginal points matter a great deal. So I give in on most arguments. The thing is, most of our grading is on an average rather than a marginal scale.

I have actually kept track of those concessions - they almost never make a difference. And I've done controlled experiments with my evaluations as to whether being a "hard ass" makes them worse or not. It does.

Posted by: David Tufte at Feb 10, 2006 9:54:16 AM

I too have found that most of the time when people ask for extensions, or
the chance to use extra time for learning disabilities the ones who really
need it take advantage of it well. The ones who are trying to scam still
write a lousy paper the night before it's due, they just turn it in a couple
of days later.

Posted by: dave at Feb 10, 2006 9:54:28 AM

Two stories:

I caught a grad student recycling work from a previous class. The evidence: comments from the previous professor on the back of a map. While that in and of itself isn't really cheating, I noticed that the student never bothered to incorporate the professor's edits into his draft. Side-by-side readings indicated that the only difference between the two papers was the inclusion of some statistical tables.

Another "stupid criminal" story involves a colleague. A student needed to make up an exam that he missed. My colleague granted the make-up and in the meantime the student sent out an email to his entire class requesting information that was on the exam. Unfortunately he recycled an email that my colleague used but forgot to take the instructor's name off the mailing list.

Posted by: Rich at Feb 10, 2006 9:56:59 AM

I had a linear algebra professor who let students take the exams home, but you weren't allowed to use any notes, or your book. When we came to class after the exam, he told us, "I can tell some of you have cheated. If you drop now, I won't go to the dean." About half the class dropped. I still wonder if he really knew if anyone had cheated.

Posted by: Michael Stack at Feb 10, 2006 10:08:02 AM

It's relatively easy to catch people who copy whole tests, or spot cheating in people who take make-up exams. It's marginal cheating that happens more often, and is more difficult to spot and stop. I would imagine that people who are willing to do it, (1) consistently get about a third of a grade higher than they otherwise would, and(2) don't consider themselves cheaters.

Posted by: John at Feb 10, 2006 10:11:19 AM

I remember that we had an honor code in high school and for AP Chemistry, the professor had to start giving closed book, take home exams to allow enough instructional time to finish the course before the AP.

I recall him making the first closed book, take home test impossibly hard. The surprise was when he handed the graded exams out, he handed out everyone's but mine. He turned to me in front the whole class and said, "What's wrong with you X, everyone else cheated and got a better score than you?"

Problem solved, he disregarded the test result and showed them that he knew when they were cheating.

Posted by: Erik at Feb 10, 2006 10:38:22 AM

I caught two cheaters once on a take-home test. It was supposed to be an in-class test, but a snowstorm intervened and I ended up having to give it as a take-home, with the directions that students were to take it individually at home without books, notes, etc.

I first noticed something was wrong when there were two papers that mis-identified the compound interest formula as "the area under the curve." A quick check showed that the two had cheated. One student would work down the space allotted for the problem until reaching the bottom of the space, then break to a new column of work. On the other student's paper, the steps were all identical and the column breaks were at precisely the same step, even though the work didn't reach the bottom of the space. The two tests had been scored within a point of each other (there was one problem on which the students did not cheat).

I was all for giving each of the students half the test's points. Seeing as how they'd completed one test between the two of them, it seemed only fair to give each half a score. But the assistant department head wouldn't let me, because I hadn't actually seen the students cheat. So I made them take their finals separately, and the student who'd copied did not pass the class.

The worst part about it (other than the sense of thwarted justice) was attending the graduation where the failing student was supposed to have graduated, because I had a family member graduating that day. The student's name was in the program and everything, but she couldn't graduate because she had just failed my (required) class.

Posted by: Wacky Hermit at Feb 10, 2006 11:49:35 AM

I caught a pair of twins cheating on a programming assignment. When confronted, they tried to say that they "were geneticly predisposed to write the same code". It didn't explain their differing grades on the rest of the course material.

Posted by: anon at Feb 10, 2006 12:06:12 PM

I really wish that professors would crack down on cheating. I got B’s with a normal amount of effort. Why shouldn’t I cheat and get a B+? When you get out into the ‘real world’ Dealing with the lazy people, the ones that cheated, is a pain. If college is a signaling tool, it would be nice if it signaled that graduates were honest/ethical. I understand it’s a pain for you to bust cheaters. But isn’t it part of your job?

Posted by: Garble at Feb 10, 2006 12:47:40 PM

So cheaters who get caught don't do well anyway. But I always wonder about the students who are good enough cheaters to not get caught do. I certainly know people who have cheated a lot, passed classes they would otherwise have failed and then done fine. All my teachers of course say that they are excellent at detecting dishonesty and point to all the comically bad examples that they have detected. But I always want to shake them and say "How do you know!? If they're good enought that you wouldn't detect them, you'd never know you missed them." Furthermore, the fact that almost all of the examples they give are comically bad kind of implies that you only have to have a low minimum competence at cheating to go undetected.

Posted by: Michael Loewinger at Feb 10, 2006 1:20:29 PM

Posted by: Michael Loewinger at Feb 10, 2006 1:21:37 PM

I'm a student but I have a fairly absurd story:

I was in a freshman philosophy course and one of the exam dates fell the Friday before a break. Three other students and I asked to take the exam during office hours on Wednesday so that we could simply have more time Friday to drive to our destinations; he agreed to the idea. When we took the test he simply sat us in a conference room with the test and left. The entire time I had to listen to one of the other students ask myself and the rest for answers to the exam. I don't think she did well, she never showed up for class again.

The class was an overview of ethical theories.

Posted by: Steven Schreiber at Feb 10, 2006 1:41:06 PM

I've mostly quit worrying about cheating. I give take-home exams, with rigid length limits, and with questions that force people not simply to look in the text for the answers. Open book, open notes. The grade distributions look almost exactly like they did when I gave in-class, closed book/note tests.

Posted by: Donald A. Coffin at Feb 10, 2006 2:36:03 PM

At a party a few years ago, I met a very drunk lawyer who told me she had gotten into either Chicago, Stanford or Yale law school (I can't remember which one of the three) through brazen cheating.

She'd done very well on her LSATs but had mediocre grades. But she'd managed to earn a couple of A's from big-shot professors who wrote her glowing rec's. She dealt with her bad grades by simply claiming to have the GPA she needed to be a serious applicant -- and then not asking her college to forward an official transcript to the law school admissions office.

Since her scores and rec's were consistent with someone who had top grades, the school accepted her self-reported GPA and admitted her. She told me she had counted on the admissions officers not seeing a transcript in her file as an administrative oversight. or maybe the previous reader had just lost part of her application. but that it would be no big deal because anyone with her scores likely had the grades she claimed.

i always think of her when i hear of students cheating. most cheaters are dumbasses. as much as it pain me to admit it, this woman's scam is impressive.

Posted by: auto at Feb 10, 2006 4:11:26 PM

I cheated in the classes i didnt care about. Im really surprised no one else in here cheated? Nobody ran into a situation where teachers gave exams on the same day? I would study hard for the one in my major, and cheated with a pal to make up for the lack of study time in the elective...
I think the whole cheater gets burned factor is kinda overrated... For instance im still not regretting looking over a pals shoulder for the answer to a history of jazz question.

Posted by: anonymous at Feb 10, 2006 4:27:40 PM

Both my wife and I went to Harvey Mudd College, a school with a very serious honor code. It worked pretty well, in general - take home closed book tests.

My wife taught me a interesting tactic as a student - if you do not feel up to an exam, or are under stress, or just did not finish the assignment, ask for an extension without giving a reason. Most professors will grant it, as long as the delay is not too big - moving an exam from friday to monday, for example, or a paper to friday from wednesday. The professor usually does not care that much about some bogus sob story. Their real goal is learning - if a day or two more lets you learn enough to do well, then they won out overall.

She also noted that she never got a higher grade because of the extra study time. What she did get was the ability to space out an unreasonable schedule. (Two exams in key topics the monday after she had to be off campus for personal reasons, for example.) This did result in grades that were representative of what she actually knew, for good or ill.

I followed this in my own classes - any student asking for a day or two usually got it. The only exception was test review - I ran those a week after the exam, and I really did not want to write a second test. Thus, if you asked for a week, I would likely not agree - six days would be the limit, and I would grade your test the night before the review.

I admit, aside from wanting to avoid having to write a second test, I also did not want to tempt a mostly honest student by having the answers 'out there'. Locks keep honest people honest.

I did see a lot of cheating, even at the grad school level, once I went to Berkeley. I am not sure what you do about it - stop the more egregious examples, then figure it will sort itself out. I also was fond, though, of pulling people up to the board to demonstrate knowledge, or lack thereof.

Scott

Posted by: Scott Ellsworth at Feb 10, 2006 5:47:18 PM

I've often wondered what to think about students who get extra time on their exams because of a disability (not blindness or deafness, but ADD or the equivalent). Most of the time, as Alex says, the extra time doesn't appear to help. But sometimes it might. I had one student who used the extra time and got the very highest grade in the class -- two classes, in fact. I always wondered what I'd do if he asked for a recommendation. He never did, though, so I never had to decide.

Posted by: Glen at Feb 10, 2006 6:02:44 PM

Fascinating thread. I don't quite follow the thinking behind this inaugural comment: "I sometimes find evidence of cheating on exams but I rarely take action, I don't have to. Almost invariably the cheaters get abysmally low grades even without penalty."

Why shouldn't there be a penalty anyway, at least a reprimand and talking-to? To suffer merely the same low grades with or without the cheating means that the attempt is in fact cost-free (not counting the student's self-inflicted wound to his own character). Better to make the student bear an explicit cost to help him understand that he oughtn't try to slide by with this kind of stuff.

Posted by: David M. Brown at Feb 11, 2006 1:21:51 AM

When i was a student i had one professor that just gave a ton of work. I always took a incomplete in his class, and would always finish the class with an A the next term.

I know i'm a outlier in many ways as far as schooling goes, but i personally found the extra time to be just what i needed so i could really get into whatever i was researching for various papers for his class.

I also took 19 credits a term (the limit before you have to pay more money at my school), virtually every term.

Posted by: cameron at Feb 11, 2006 4:20:59 PM

One day when I was in grad school (physics) a fellow grad student came to departmental tea very distressed. He had just come from grading the quizzes from his last recitation section. He had very carefully worked out EVERY problem on the quiz on the blackboard, answering any and all questions as he went. He left all of the problems, with their answers on the blackboard. He then administered the quiz (without explictely pointing out the problems were the same as he had just worked out). The class average on the quiz: D. Some people, you just can't reach...

Posted by: quadrupole at Feb 12, 2006 5:48:41 AM

Many years ago, my first term as a TA was for a principles course for Dan Benjamin (now at Clemson). Students were seated by their quiz sections so the TAs could keep track of them. There were four in-class tests, plus the final. On the third test, I saw of my students very unsubtly looking over at the test of the girl in front of him (it was a tiered hall). I asked the head TA what to do, and he brought the problem to Dan Benjamin. He asked how the cheater was doing so far (basically low C) and how the girl he was cheating off was doing (struggling to get up to a D). So he said, let him cheat. The cheater's grade plummeted on that test, and we had no problems with him the next time. My view is that part of the difficulty is that penalties for cheating are too severe. Consequently, students have big incentives to fight the penalty. The time involved ends up being too much bother.

Posted by: William Sjostrom at Feb 12, 2006 10:07:51 AM

So Glen-- Why don't you believe that Mr. ADD in your class-- the one that had extra time, and got the top grade-- was a bright student with a handicap, who was indeed the top student in the class, but whose ADD or dyslexia or whatever it was meant that he needed extra time? Do you think only stupid people have ADD?

Posted by: Cardinal Fang at Feb 13, 2006 1:10:21 AM

Quadrupole: "The class average on the quiz: D. Some people, you just can't reach..."

Some people, like the entire class? Maybe he and the grad student were just very, very, very poor at getting the material across. (If the grad student was anything like most physics grad students I've had, his handwriting was illegible and his English incomprehensible). At any rate, it sounds like maybe the professor should have curved the grades to at least bring the average to a C or so. Since professors do get to decide on how to assign letter grades, after all.

Posted by: anno-nymous at Feb 13, 2006 7:28:28 PM

Sorry, by "he" I meant "the professor".

Posted by: anno-nymous at Feb 13, 2006 7:29:35 PM

I had a professor who had a policy of allowing students to miss any test without requiring a reason and would simply weight the final more. I only took the finals, and used the previous tests to study from. I got an A and probably would have got one anyways but only studying for one test is less work than studying for three or four.

Posted by: Mike F at Feb 14, 2006 12:34:33 AM

It's hard to get an A, but it's easy to get a C in a class that has non-exam work. You cheat a little on the outside work, cram for the exams, and then the teacher ends up looking at an A or B project and a D exam. The average will easily be a B or a C. I saw it all the time as a computer science teacher -- we have to give projects for some classes, but projects are ridiculously easy to cheat on.

Some ways to cheat rely on having some sort of social network. If you are attractive, you go blink your eyes at the people who know their stuff, and ask for help. (Probably every nerd type in this forum can tell stories of being on the other side of that!) If you are in a fraternity, you ask your brothers.

But there are plenty of possibilities on your own, too. Google is rather helpful for finding similar or even identical work. If you are expected to read a novel, read the Cliff's notes. If you are expected to read a book in French, read the English translation instead. The last two are particularly irritating if you are struggling in a class, because the cheater ends up knowing the material *better* than those who do things the right way.

It's a disturbing situation to me, who grew up thinking that grades roughly meant something. There are ways to fix it. For example, we could stop grading projects. Or, we could redesign projects so that "cheating" is allowed. But barring major changes like these, I can only conclude that grades are close to meaningless. A's are hard to cheat for, yes, but are possible in many classes. B's and C's? You have no idea. About all you can tell is that if they fail, they probably were honest at least.

Posted by: Lex Spoon at Feb 14, 2006 4:11:36 AM

I agree that this post demonstrates that above profs only catch the below average cheaters. It makes sense for them to think that, it must be comforting to think that cheating isn't one of their problems. Half of the top students in the (highly ranked) program I was in were cheaters (Now they are phd or med students at highly respected universities). They knew their stuff, but they also double checked their answers with each other on tests... among other things.

Posted by: someguy at Dec 21, 2006 6:25:33 PM

I should mention that one of the main ways (The component of cheating that had the largest effect on GPA) these students cheated might not really have been cheating, although some honors codes might define it as collusion. Certain students have networks that get them old finals. At my school, the most organized networks were indian students where everyone is everyone else's cousin or best friend or best friend's cousin (CA). Most profs don't provide their classes with finals from previous years (Which would cut off this advantage). Worse, they are really lazy (Okay, research focussed) and re-use old midterm/final questions.

Posted by: someguy at Dec 21, 2006 8:36:02 PM

The best cheating-busting anecdote I've heard:

A professor had prepared and printed the exam papers the day before, and left them in his office. The next morning, he entered his office and realised that someone had been in during the night: things moved around, etc. But the stack of exam papers was still on his desk.

So he took the stack of exam questions to the photocopy room and chopped a half-inch off the bottom of every page. The exams were distributed and written. When they were turned in, the professor shuffled the papers a bit, quickly identified an exam paper that was half an inch longer than the rest, and called the offending student up for a quick chat.

Posted by: Stephen Gordon at Dec 21, 2006 9:29:09 PM

I think you are misleading yourself about the success of cheaters. The surveys show about 80% do! My favorite experience was when I just happened to catch a rubbernecker. She had copied off an A student's Scantron but failed miserably. I examined his exam sheet and Scantron because he had an A. He had marked the wrong answers on the question sheet to thwart the cheater! I only catch a handful of what goes on and I don't feel the need to rationalize it. They are usually very good..

Posted by: edwardseco at Dec 22, 2006 2:27:16 AM

"But there are plenty of possibilities on your own, too. Google is rather helpful for finding similar or even identical work."

I have a friend who was given an assignment in a class and went home to work on it. He googled a couple of the distinctive terms and found that the instructor had stolen the assignment itself (without any attribution) from a professor at another university.

Posted by: Slocum at Dec 22, 2006 7:52:31 AM

Speaking of reading the Cliff Notes instead of the book -- when I was in high school I just didn't read most of the books I was assigned in literature class. I didn't even read the Cliff Notes. I still got top marks in all my lit. classes, even the AP ones. Go figure. My explanation at this point is that they had really low standards, and no one else in the class read the book either.

Posted by: Sameer Parekh at Dec 22, 2006 8:12:42 AM

I'm amazed by all the stories. Where do you get the choices of reactions. The schools I know insist on written in class finals, even where it's silly. A lot of this is the difference in subjects. You may have to worry about a multiple choice test, but not for an advanced Topology exam.

Posted by: Weldon at Dec 22, 2006 8:31:03 AM

I was a marginal cheater. If I could get an easy glance at a neighbor's test and I had no clue on a particular question I would. It helps to have extremely good vision, and it's not too hard to reorg the work so that it's not a direct copy. I did marginally cheat in classes that I otherwise did well in.

I had a math prof who gave beautiful tests... take home, open note, open book, and extremely difficult. The notes and book wouldn't solve half of the problems, and neither could you. Assuming you got the easier ones right, he would determine your grade based on your futile attempts at the hard ones. Collaboration would be very eay to detect.

Posted by: Anon at Dec 22, 2006 9:45:03 PM

Seriously. I knew a number of people in college who cheated regularly and successfully. They typically recieved 1-2 letter grades higher than they would have without cheating, and perhaps more significantly, spent considerably less time on homework/studying.

It's pretty clear that someone who has no clue is going to have a hard time cheating themselves into a pass. But a C student cheating their way to a B+, or for students who are looking to get into competitive graduate programs, a B+ to an A? That happens all the time.

Posted by: TW Andrews at Dec 27, 2006 10:34:33 PM

liqingchao 07年8月21日

google排名
 google排名
 wow gold
wow gold
powerleveling
powerleveling
wow gold
wow gold
powerleveling
powerleveling
wow powerleveling
wow powerleveling
wow power leveling
wow power leveling
wow power level
wow power level
world of warcraft powerleveling
world of warcraft powerleveling
world of warcraft power leveling
world of warcraft power leveling