When Honor Codes Go Bad

I personally like the idea of an honor code. It keeps authorities from being over your shoulder which can be rather annoying. But at the same time, I like it because I rarely do things I believe most people would consider wrong. Isn't part of the problem with the honor system that it only works when people are going to act the same even without it. It doesn't seem like the honor code system stops anyone from doing anything wrong and asks the people that are already following an internal honor system to keep on following it.

Am I missing something in this?

2 comments on honor codes:

1) I find requiring students to narc on each other annoying. Requiring students to take on the negative obligation of not cheating is fine; requiring students to take on the affirmative duty of policing and reporting cheating is a step too far, in my opinion.

2) If an honor code is implemented, it should be somewhat reciprocal. Having an honor code, and then requiring documentation from a doctor prior to missing an exam, for example, I find inconsistent.

Typical honor codes do include some enforcement provisions-- generally, students and faculty can report incidents of suspected cheating, with the cases then going before an honor council or committee of students (and possibly faculty and administrators) who determine guilt or innocence and assign punishment.

For the most part, though, they do rely on people doing the right thing because it's right, and not out of fear of punishment. That's the main point of the systems, and also why they tend to work better at small elite colleges than at large impersonal universities.

I'm at William and Mary, and the physics TA's have had direct dealings with the honor code. We've had kids open up textbooks during closed book tests and kids copy old labs with different qestions and different data. They all went unpunished because we didn't specifically say "This is a closed book test. Do not use a textbook" or "Do your own work." You know, common sense implied rules. Basically, it turns the rotten apples in your college into lawyers. No offense, Mrs. Uncertain Principles.

Whereas my snooty law school ran on an honor code, with no difficulty that I heard of.

(I think "do your own work" ought to go without saying--though I believe Chad has found that it *doesn't*, even without an honor code--but the question of open or closed book does need to be addressed.)

Unfortunately, there are people out there (e.g., Dr. Wells of the DI) who will do just about anything to achieve their goals.

Do you suppose an honor system looks like a "pro" or a "con" to them?

I'm at William and Mary, and the physics TA's have had direct dealings with the honor code. We've had kids open up textbooks during closed book tests and kids copy old labs with different qestions and different data. They all went unpunished because we didn't specifically say "This is a closed book test. Do not use a textbook" or "Do your own work." You know, common sense implied rules. Basically, it turns the rotten apples in your college into lawyers.

This sounds like the sort of anecdote where it will turn out that some crucial context is missing. Even as stated, though, I don't think that's a problem with honor codes in general-- it sounds like a problem with a specific implementation, both on the test design level adn the enforcement level.

Unfortunately, there are people out there (e.g., Dr. Wells of the DI) who will do just about anything to achieve their goals.

Do you suppose an honor system looks like a "pro" or a "con" to them?

I think life is too short to spend worrying about the opinions of people who habitually act in bad faith.

http://donut.caltech.edu/about/history/Jou02/Jou02Honor.pdf

Here's a report about the honor code at Caltech. The "data" is mixed. My personal experience in courses I've TA'd or taught is that on homework sets, the compliance is frankly awful.

I felt that the comprehensive Honor Code was one of the best things about Caltech.

The price paid was occasional annoyance from enforcement by the students themselves. But that's a small price to pay for a sense of being trusted, of being taken seriously as adults, of being shown how science is inherently ethical.

Once I had a take-home exam that came in two sections. The cover sheet clearly specific, let's say, that one should open the first part, read the first question (or problem set) and spend no more than 2 hours. Then, one should open the second part, read the second question (or problem set) and spend no more than another 2 hours, for 4 hours total test-taking.

I told my room mate (2 students per dorm room, until one had enough clout as a Senior to get a single room). I started the exam. After I finished Part I, but before I opened part II, a female I knew from a ritzy school elsewhere in Pasadena, stopped by to see me. She was wearing a fur coat with nothing on underneath. More than 4 hours after I'd started the exam, my room mate came back and saw that I was still filling out my Blue Book. I didn't tell him about what had ensued between me and the young lady, that being rather personal. Nor do I belive in "kiss and tell" memoirs in general. I handed in the bluebook the next day.

I was hauled before the BOC (Board of Control). My having taken more than 4 hours for a 4 hour exam was at issue. Fortunately, I had the actual format of the exam to show as evidence, plus the testimony of the female witness. I was exonerated, and the BOC told my room mate to please check with an individual as to circumstances before jumping to conclusions and reporting a possible honor code violation. That was the ONLY time I came across the enforcement end, and, the facts and the law being on my side, I bear no grudge against anyone involved.

The rest of 5 years I spent as an undergraduate, I was 100% helped by, invigorated by, impressed by, and supported by the Honor Code.

The Scientific Method itself relies upon a kind of honor code. Open Source software relies upon a kind of honor code. Writers' organizations such as SFWA, MWA, NWU, WWA, RWA and the like a kind of honor code. So does the blogosphere.

Yes, there are rare cases of fraud in science, plagiarism by professional writers, and (sad to say) numerous trolls and spammers in the blogosphere. Exceptions. All organisms and communities have parasites.

Honor is good. Honor Codes, based on my experience, are good for honorable people.

About 4 years ago, I was a Chemistry prof at Clemson, which is a pretty big school without an honor code. I had a nasty tendency to give take-home final exams, and I required that each student hand in with the exam a signed statement declaring that he or she had done all of the work without assistance from any individual other than me, and that an F in the class would result if I were to learn otherwise. One semester I failed a graduate student who copied answers off of his roommate, and another semester I failed over a third of my undergraduate class who foolishly got together to collaborate on a 20-point problem on the 200-point final exam. I've wondered if a campus honor code would have made a substantial difference.

Vanderbilt's Honor Code is a sham, a shambles, an embarassment, and a joke.

I will write much more on this later when I have time.

I'm not opposed to honor codes in general. Indeed, as a student at Harvey Mudd, I fully believed in the honor code there.

At Caltech, as a TA, I had an experience that left me deeply dubious of the honor code.

At Vanderbilt, I have absolutely no faith in the students' respect for the honor code in general, or in the systems that are set up to enforce the honor code.

I'll blog about this either later today or tomorrow.

-Rob

I went to a university with an honor code that covered both academic honesty as well as personal conduct and appearanced. I never had any personal dealing with the honor code office one way or the other.

Since I've never attended a school without an honor code (and never really gave it much thought) I find myself wondering: What difference is it really going to make? I don't think there's anyone out there who really thinks that cheating is permitted. If someone is dishonest enough to cheat, is making them promise not to cheat really going to stop them? Also, what does the average honor code cover besides cheating?

I like honor codes in principle, for basically the same reason that JvP gives above -- it seems right to me to treat the students as adults and trust them to not cheat. Also, from a teacher's perspective, I don't want the students thinking of me as the police for their academic honesty. However, my sole experience with an honor-code system was not good.

This was during my abortive career as a chemistry grad student at Stanford. The system got several things right, the most important of which IMO was that professors were not allowed to set exam-taking conditions that would create temptation to cheat. For instance, take-home exams had to be open book and untimed. (The situation with JvP and the young lady would thus not occur.) (I should mention that I partially like that rule because I think open-book untimed exams are pedagogically better.) However, the procedure for punishing violations was so baroque that professors didn't want to bother with it, and we were not allowed to take any action on discovered cheating without going through the procedure. Thus on several occasions I as a TA was forced to ignore cheating, which I think is absolutely the wrong thing -- even if the offenders are going to fail the class anyway, they shouldn't be left with the impression that they got away with the cheating.

Stanford is a large impersonal university with institutional delusions of being a small friendly college, so may represent the worst case for an honor code.

I've been to two schools with honor codes, one was terrific, one was terrible.

I did my undergrad at Reed, which has had an honor code essentially since it's founding. The culture there was very conductive to this working, the students and faculty worked closely together. The students also thought of themselves as scholars, so the attitude of "don't cheat" was very strong. Timed, take-home closed-book tests were common, and essentially never a problem.

Stanford, on the other hand, was a disaster. I TA'd there as a grad. student and had much the same experience has Zack in #13. The culture around the honor code there was less that it was about "students should't cheat" as it was about "professors shouldn't try to catch students teaching". This, combined with the fact that it was a pain in the a** to do anything about cheating if you found out, meant that enforcement was essentially non-existent.

Having gone through for a time a school with an honor code system as a student, I only have two observations really:

1. The important thing about an honor code isn't even the pledge itself, it's the fact that the school is making a visible commitment to taking academic dishonesty truly seriously. Most colleges seem to police-- and punish-- underage drinking, for example, more severely than cheating. This sends a certain sort of message-- especially since nobody honors the rules about underage drinking. Though the honor code does to an extent work because the students can see they're being given a certain degree of trust, the most important thing was the strict level of enforcement the rules had-- the pledge was basically unimportant except as a constant reminder that the rules were going to be enforced in a draconian manner.

And no, the point wasn't even that people were following the code out of fear of being caught-- the crucial thing was the message sent. The students were very intensely aware that the school cared about this stuff, above and beyond practically any other concern. Eventually it reached the point where to some extent (even if in some cases only out of fear) this meant the students cared, too.

2. I sincerely believe that the worth of an honor code varies inversely with the size of the academic institutions involved. When a teacher has a couple tens of students, they can bother to be personally involved and take seriously and look into every potential violation of the rules. When they have a few hundred, this possibility is hopeless and the system ceases to have any meaning. I see no way how an honor code system of the type I saw could have possibly worked even a little bit at a college of any significant size. I don't know what the school size (or, probably critically, class size) is at Chad's place.

At Stanford it seems to work well, I don't know of any major cheating scandals (although apparently one of my introductory classes had been plagued by it in years past, so we had a proctor for that one. Which made it interesting when some naked guys came running in and she insulted their penises).

I had an honor code in HS but not in college.

Well, from some anecdotes that I've heard, honor codes don't discourage cheaters, and only manage to convince faculty that cheating doesn't happen. I think there was a study at Virginia on this a while back.

One point that I think that I saw on PZ's blog last year is that imposing harsh penalties on cheaters discourages other students from reporting. They don't want to be the person who ruined another's academic career.

The (undergrad) honor code at Caltech is a joke (personal knowledge from four years of TAing physics.) It's routinely violated with enough plausible deniability to ensure non-prosecution. The only people who benefit from it are the select few who are both compulsively honest yet still bright enough to do better than people who cheat. Oh, and the professors and TAs who don't have to be bothered to proctor .

I imagine a code would work better in less competitive environments (pass/fail graduate courses, for example, or nonstandard institutions like Reed) but at Caltech all you have to do is poll a few TAs and you'll know the prevailing opinion.

I don't know anything about the social side, except to say that I don't know what's different than any non-honor-code school's code of conduct.

Re: # 8 | TJ

Thank you for that PDF. It was interesting in terms of history, data, and interpretation. I was also thrilled to see quotations from two people whom I very much admired: Leonidas J. Guibas and Phil Neches.

"Professor Guibas heads the Geometric Computation group in the Computer Science Department of Stanford University and is a member of the Computer Graphics and Artificial Intelligence Laboratories. He works on algorithms for sensing, modeling, reasoning, rendering, and acting on the physical world. Professor Guibas' interests span computational geometry, geometric modeling, computer graphics, computer vision, sensor networks, robotics, and discrete algorithms --- all areas in which he has published and lectured extensively...."

Guibas was awarded his Ph.D. from Stanford University in 1976. His dissertation: "The Analysis of Hashing Algorithms" and his
advisor: Donald Knuth.

Phil Neches was founder, chief scientist, and vice president of Teradata Corp. (1979-1988), which applied parallel processing to relational databases, and which was acquired by NCR and AT&T in 1992.

Almost exactly my age, he's achieved somewhat more than I in the corporate world: "He is chairman of Foundation Ventures LLC, an investment bank serving information technology and life science companies. Previously, Neches was vice president and chief technology officer of AT&T's Multimedia Products and Services Group and senior vice president and chief scientist at NCR.... He is a Trustee of the California Institute of Technology, where he earned his BS, MS and PhD in computer science.

Two data points, but they support my belief in an honor system run by honorable people for honorable people, and to the benfit of science (and engineering and business).

Since I've never attended a school without an honor code (and never really gave it much thought) I find myself wondering: What difference is it really going to make? I don't think there's anyone out there who really thinks that cheating is permitted. If someone is dishonest enough to cheat, is making them promise not to cheat really going to stop them? Also, what does the average honor code cover besides cheating?

There's a phrase that's thrown around when people discuss this stuff, the "20-60-20 Rule." The claim is that 20% of students will cheat no matter what you do, and 20% of students will never cheat no matter what, while the other 60% might go either way-- they'll cheat if they think they're likely to get away with it, and they won't if they think they're likely to get into trouble.

The idea of honor code systems is to try to affect this psychology. Not so much through having students actaully turn each other in for cheating, but by creating a campus culture in which cheating is Simply Not Done. It becomes a sort of institutional point of pride for a lot of honor code schools, and students take it very seriously.

I'm not sure where I'd put Williams in the range of honor code schools. It was taken fairly seriously, and promoted pretty heavily by the College, but it's not like we sat around discussing it all that often.

Well, from some anecdotes that I've heard, honor codes don't discourage cheaters, and only manage to convince faculty that cheating doesn't happen. I think there was a study at Virginia on this a while back.

The Virginia thing you're probably thinking of is the Lou Bloomfield incident, where he fed a few years worth of papers through a text comparator, and then charged 158 students with plagiarism. I suspect the same thing would probably happen for just about any class that doesn't use some automated check like turnitin.com.

I don't know anything about the social side, except to say that I don't know what's different than any non-honor-code school's code of conduct.

In practice, they're probably different primarily in the judicial apparatus used to prosecute violators. Honor code schools generally handle things through some committee composed at least partly of students, and accordingly take a slightly different approach to punishment. Skidmore is the local example of this sort of thing.

In principle, it ought to be a less legalistic way of dealing with things, but I'm not sure how well that would work.

The Stanford and Caltech anecdotes are interesting, thank you. I'm not sure the problems described are really problems with an honor code per se, though-- it's perfectly possible to have a baroque and unwieldy justice system built on a more traditional enforecement model.

Does anyone reading this know anything about schools that had honor codes, but gave them up?

The biggest problem with Caltech's honor code is that the enforcement body is the Orwellian-named Board of Control. They really ought to think about changing that.

FWIW, I have no hard data whatsoever, but my sense from two years at Caltech (about 30 years ago) is that the honor code was taken seriously but was routinely evaded in fairly modest ways. "Modest" meaning taking 2.5 hours for a 2-hour test, say, or looking up a formula or two in a closed-book test. Like I said, I have no personal evidence of this, but the patterns of behavior seemed very hard to reconcile with 100% compliance.

My undergraduate institution had an honor code, and I don't think it had any effect one way or another. I had plenty of take home exams, but as far as I can remember they were all open-book and without time limits other than the due date. Different professors will always place different amounts of trust in their students, and some students will take advantage of it, but I don't think an honor code - which is really just part of the administrative mumbo-jumbo every university has - is going to matter. I don't even know if my graduate institution had one.

This is a random anecdote that seems appropriate for this thread, but it probably won't indicate much. I took a difficult undergrad English class that was taught as if it was grad. level, and on a compressed schedule over the summer. I don't think any of us were doing well. On the day of the final exam, the teacher left us with the exam and walked out. I wasn't at a university with an honor code, and this was the first time I had an exam unproctored (right before my graduation, too). My first thoughts were "what? how will she know if we cheat? is she allowing us to cheat since the class is so hard?" but nobody cheated. It was disorienting for me, but I've always found it fascinating that the teacher gave us an unproctored final and that nobody cheated.

I know of no instances at Bradley where anyone cheated on something like an exam, that was applicable to you're talking about. Meaning, I don't know of any students who collaborated on take-home exams, for instance. (I don't know if it was university policy, but our takehome exams were always open book, indeed, open library, and not times except by a due date.)

There was a case where several students broke into an office and stole a copy of an exam before the test date, but that's really going above and beyond the duty for larceny.

However, there was a knot of corruption and iniquity regarding homework at the undergrad level. And that knot was called "The Engineering Fraternity," whose past homework files were legendary. Helpful note to professors: Do not be under the illusion that you can rotate homwork assignment sets on a three, or even a five year basis, and think you have solved your problem. You cannot. Your local engineering fraternity will have files of homeworks and likely professorial quality solutions for every class, going back to the fraternity chapter's foundation.

Do not fight this fight.

You will lose.

Since we had an engineering fraternity, cheating on homework was rampant. There was cheating by old homework files, and there was cheating by copying off the smart members' homework when he was done. (Non-fraternity members did that, too, but the infrastructure of the fraternity put that in a class by itself. Daley would have been awed by the efficiency of that machine.)

Predictably, this served them well until about Junior Year, when the heavy duty lab classes started and they had to apply this stuff in real time.

And the labs were all TA'd by angry, non-frat little thugs like MEEE!

I was both a student and a TA at a school with an honor code, University of Virginia. When I was a student I thought the honor code was great and was completely unaware of any cheating. Everyone I knew took it very seriously and as a point of pride. Although I didn't have any cheating issues in the classes where I was a TA, some of the other grad students did and found the Honor system frustrating. The people who cheated considered the Honor system a joke and a few of them were able to play it like a fiddle. (I know of one case where three of four students turned in identical exams. So identical that they all misspelled the word force in the same place. Each one claimed that the others copied from his/her (I don't know anything about the identities of all but one of these students because only one student opted for an open trial) test. As a result, at least two of them got away with it.) Most of the professors I talked to consider the enforcement portion of the system to be a big hassle and just avoided turning the case over to the honor committee when they discovered cheating. (The professors who were once students at UVA seemed to be more loyal to the Honor System) I should note that at UVA the honor system is student run and the professors and TAs are free to deal with cheating cases independent of the Honor system. ie. If you think they cheated you are allowed to give them a big fat 0 on the test. You can do this even if you've referred the case to the honor system.

Overall, I still think the system works and engenders a spirit of anti-cheating comradere for a large portion of the student body. But it's definitely not perfect.

A story that people used to tell as true when I was a Caltech undergrad (20+ years ago) was that there had been a freshman physics exam which was semi open-book. You could use the text book, your notes, the integral tables, and "Feynman". A normal student would assume that meant the dreaded Feynman lecture series: wonderful for intuition, useless for problem solving.

However, a particularly innovative student decided that "Feynman" meant the man himself. As the story goes, in the alloted time he wandered up to Feynman's office and knocked on his door. After explaining his interpretation of the instructions to Feynman, Feynman examined the instructions carefully himself and then proceeded to help him work through the exam. That's why, in my era, the instructions always carefully specified the books by Feynman one was allowed to use.

I'll let you all decide whether you think this ever actually happened. All I can say is that it was told as true.

With regard to the actual issue at hand, while I was never on the board of control or brought before it, I knew people on it and people brought before it. It definitely struck fear into the hearts of normal Caltech students. I'm sure there was plenty of cheating and that only some of it was caught. However, that's almost certainly true in a school without an honor code, too.

Without detailing my elsewhere-blogged relationship with Feynman, let me provide a related "urban myth" -- also described as true during my undergrad days at Caltech (1968-1973).

Student: "Will this exam be open book?"

Professor: "Yes, this exam will be open book."

Student: "Will this exam be open homework?"

Professor: "Yes, this exam will be open homework."

Student: "Will this exam be open notes?"

Professor: "Yes, this exam will be open notes."

Student: "Can I bring my slide-rule?"

Professor: "Yes, of course you may bring your slide-rule."

Student: "Can I bring an electronic calculator?"

Professor: "Yes, you may bring an electronic calculator."

Student: "Can I bring the Feynman Lecture Notes in Physics?"

Professor: "Yes, you may bring the Feynman Lecture Notes in Physics."

Student: "Can I bring --- ?"

Professor: "For god's sake, man. Anything you can carry yourself in one load, you can bring into the exam."

So the student came to the exam with a post-doc riding on his shoulders.

We've had kids ... copy old labs with different qestions and different data. They all went unpunished because we didn't specifically say ... "Do your own work."

Three observations:

1) The punishment for giving the correct answer to a different question is a zero for that question, because that answer is wrong. Same goes for having data that were not collected in that lab. Copying is harder to identify, but if errors like the ones above increase your focus on certain students, the punishment for mis-copying a solution (telephone game type errors) is a rubric that assigns a high negative value to that specific algebra or arithmetic error. Often those copying errors manifest themselves as two mistakes that cancel out, and my grading rules always say that two wrongs are twice as wrong, not right.

2) Does your syllabus now say "do your own work" on certain parts of lab reports? See item 3.

3) You are in charge, not the inmates of the asylum. One of our faculty complained about two students who "always" sat at the back and copied (off of an open paper, having swapped exam colors) during exams. Huh? They "want to sit back there"? They will sit where I tell them to sit, because they will not get a test until they do, and I will stand where I want to stand while they work on the test. (But some academics do eschew confrontation and students will exploit their preference for complaining over action.)

Did you see the story from the Duke Business school today?

http://www.newsobserver.com/1366/story/568984.html

In that case, the violation of the honor code was detected by the professor, according to the news story.

As a grad student at U.Va., the honor code had all of the problems that others have noted: while usually observed, that would have been true without a code, and the enforcement process was unusably pointless. Moreover, we were clearly instructed that use of the honor system was the *only* permissible sanction for dishonesty - which _we_ then ignored. The single sanction I thought was fine, though.

[Someone mentioned Reed: yes, great code environment, but another case where I think that the code makes no difference - no-one would cheat anyway.]