52
$\begingroup$

So I strongly suspect that one of my students is asking all their homework questions here. In particular, all the questions that user8917 has asked about probability are homework questions that I have assigned in my current course. I am using a standard text (Pitman, Probability), so coincidence is possible, but there are many problems in our textbook that I have not assigned -- it seems unlikely that a person studying the book on their own, or even a person in another class using the same book, would choose exactly the same set of problems to be handed in.

My policy on academic dishonesty is simply that students cannot copy their answers from any source, although of course with relatively simple problems this is difficult to impossible to enforce. I of course would not want to forbid my students to use sites such as this one - if they have more general questions, or questions which are not specifically "how do I do this homework problem". But I don't want people solving my students' homework questions for them! I would appreciate the community's advice on what I can do here.

This is in some respect a repeat of this question, regarding a similar anonymous user who consistently asked homework questions; the major difference is that I, the instructor, am the one who noticed this.

$\endgroup$
11
  • 17
    $\begingroup$ Since you have identified these questions, you may be able to compare the solutions that are turned in with the answers posted here, and enforce your policy if someone turns in an answer copied from this site. You may also be able to verify whether or not somone is indeed posting your homework problems by modifying them slightly from the book (change notation, wording) before assignment. $\endgroup$
    – Arturo Magidin
    Apr 19, 2011 at 18:06
  • 20
    $\begingroup$ Something similar happened also to Gerry Myerson. He basically just left a comment on the question telling the student to stop doing that. I think it worked. $\endgroup$
    – Willie Wong
    Apr 19, 2011 at 19:37
  • 1
    $\begingroup$ +1 to @Willie. I would recommend trying to talk to the student or otherwise contact them and telling them to not do this. $\endgroup$
    – Glen Wheeler
    Apr 19, 2011 at 21:16
  • 4
    $\begingroup$ If I knew who the student was, I'd contact them; I suspect they know that what they're doing is wrong, because they asked the question anonymously. I may say something in class tomorrow though... $\endgroup$
    – Michael Lugo
    Apr 19, 2011 at 23:48
  • 4
    $\begingroup$ "I may say something in class tomorrow though..." - that might be the only thing you can do for the time being, since there is still the off chance that it's all a coincidence. You can monitor submissions later and confront students if you find anything sticky. Arturo's suggestion of always using modified questions instead of using questions from the class textbook verbatim is a good one (though admittedly it requires a bit more work on your part ;) ). $\endgroup$
    – J. M. ain't a mathematician
    Apr 20, 2011 at 0:17
  • 4
    $\begingroup$ @Willie, in hindsight, I don't think I did the right thing in the case to which you refer; I think it would have been far better if I had contacted the student in private and not in public. But in my case, the student used his real name. Since Michael's presumed student is not using a real name, the issue of embarrassing the student does not arise. I see nothing wrong with Michael posting something to the effect of "if you are my student, then please [insert here what you want the student to do, or not do]." $\endgroup$
    – Gerry Myerson
    Apr 20, 2011 at 2:03
  • 14
    $\begingroup$ I tend to link this answer of Joel Hamkins to a related question, as I find his idea one that worths consideration as a teacher. meta.math.stackexchange.com/questions/1652/1661#1661 and as Carl Brannen answered here, this is the era of the internet. In all likelihood they can find complete or almost complete solutions online. Let them, grade them on how well they understood those solutions and can reproduce them under different degrees of variation in the details, in ways that require some understanding and not just memorizing. $\endgroup$
    – Asaf Karagila Mod
    Apr 20, 2011 at 8:39
  • 4
    $\begingroup$ How is a mathematics student getting homework solutions from the internet like and unlike a history student copying an essay from the internet? Should existing "plagiarism" regulations be applied to math homework? $\endgroup$
    – GEdgar
    Apr 23, 2011 at 13:42
  • 3
    $\begingroup$ GEdgar, I think it's entirely the same thing -- but quite a bit harder to detect. My understanding is that a lot of teachers in subjects where the grade is based on essay-writing can often Google a sentence that seems out of place in student work and find the source. But I don't think we can do that. $\endgroup$
    – Michael Lugo
    Apr 23, 2011 at 15:25
  • 2
    $\begingroup$ @Michael: user8917 now has the user name mary. Perhaps this helps in the identification. $\endgroup$
    – t.b.
    May 2, 2011 at 17:19
  • 4
    $\begingroup$ You'd think that would help, wouldn't you? But I don't have any student named Mary. $\endgroup$
    – Michael Lugo
    May 2, 2011 at 17:25

5 Answers 5

Reset to default
30
$\begingroup$

The era of the internet is upon us.

I say let your students gather their learning how they wish. If you want to grade them based on their abilities use in class tests. And by the way, do not let them bring their cell phones or "calculators" into the test room.

$\endgroup$
12
  • 14
    $\begingroup$ As a professor, I think that this comment has the right sentiment. However, the role of graded homework is not only to assess the students. Graded homework also motivates students to study, but more importantly it gives a grade boost to students who work regularly but may have difficulty on the exams. The idea of only using graded exams is attractive (no more grading awful homework papers!), but ultimately not as effective. Unfortunately, I have voted this answer -1, only because I think votes on meta just measure whether other people agree with the answer. No disrespect at all is intended. $\endgroup$
    – Carl Mummert
    Apr 20, 2011 at 1:18
  • 1
    $\begingroup$ @CarlMummert - Can you elaborate on why using graded exams is not as effective? I understand comments are not the place for a full explanation but I am curious about the concrete reasons for this, as I believe "learning" isn't just about working regularly. $\endgroup$
    – Nicole
    Apr 20, 2011 at 1:52
  • 10
    $\begingroup$ @Carl; In terms of how one learns the material well, I agree with your comment completely. As far as using grades to distinguish between the better and worse students, I insist that in the era of the internet, only in-class tests will do this. $\endgroup$
    – Carl Brannen
    Apr 20, 2011 at 2:15
  • 5
    $\begingroup$ And when I was a student, I did all my work without asking other students, and only a couple times did I have to resort to asking the instructor. This made me very strong, but I think I was naturally better at the material than a lot of students. Doing the work on their own was simply not an option. I remember one or two who undoubtedly, if transported 30 years into the future, would be here on MSE getting other people to do their exercises. $\endgroup$
    – Carl Brannen
    Apr 20, 2011 at 2:17
  • 3
    $\begingroup$ @Renesis: of course just spending time doesn't ensure learning, but most people need to put in a significant amount of time to learn technical material. One goal of grading homework is to motivate students, especially those in the middle of the class, to put in enough time to do well on the exams. I am thinking particularly of low-level classes here, where the students have little commitment to math, and are also unaccustomed to having to work hard. Just like weekly "productivity reports" can help lazy office worker be more productive, weekly homework helps some weaker students pass courses. $\endgroup$
    – Carl Mummert
    Apr 20, 2011 at 11:39
  • 8
    $\begingroup$ From the teaching point of view, there is one big downside to your answer: the assignment of homeworks is not just for the sake of students. For myself (and for some other teachers I know), the performance of students on the homework sets also serves as a gauge of the effectiveness of our lectures. If students who show up to all the lectures regularly perform better than those who don't, I know my presence at least count for something. $\endgroup$
    – Willie Wong
    Apr 20, 2011 at 11:39
  • 4
    $\begingroup$ @Carl; It's really a shame, but I think it's arguable that, as far as teaching method goes, the best mathematicians were the ones who learned under the Moore method. I think it's still possible to teach that way, but to do it you have to have the cooperation of the students. $\endgroup$
    – Carl Brannen
    Apr 20, 2011 at 16:27
  • 7
    $\begingroup$ @Willie, it does suck enormously when the students that do not show up in class do better than the others, doesn't it? $\endgroup$
    – Mariano Suárez-Álvarez
    Apr 21, 2011 at 19:54
  • 1
    $\begingroup$ @Willie; I agree with you 100% on this. But I don't see a method of implementing this other than to require that the students live in a region where there is no internet connection to the outside world. Perhaps it will eventually come to pass that the next great generation of mathematicians will be educated in a language so obscure that their efforts to have the world community solve their HW problems will come to naught. -- No. That will only change the behavior of the mediocre students who really don't matter that much anyway. $\endgroup$
    – Carl Brannen
    Apr 21, 2011 at 20:50
  • 1
    $\begingroup$ @Mariano: yes it does. Though most of the time it happens the students are honest to me about it: "I know the stuff already, and just need an easy A to boost my GPA for Med/Law/Grad school application." There are, of course ... outliers. $\endgroup$
    – Willie Wong
    Apr 22, 2011 at 13:55
  • 3
    $\begingroup$ @Carl: if the homeworks are not graded, it is not impossible to implement it. As you yourself observed in the comments to Numth's answer. And I suspect I misunderstood your answer. I read it as "Don't assign homeworks". Whereas you probably meant "Don't base grades on homeworks". $\endgroup$
    – Willie Wong
    Apr 22, 2011 at 13:56
  • 3
    $\begingroup$ @Willie; Yes, that's what I meant, or better, "if you want to assign grades based on merit, don't use the homework." $\endgroup$
    – Carl Brannen
    Apr 23, 2011 at 3:43
26
$\begingroup$

I never assign book problems in my classes. Composing my own problems is a lot more work, of course, but I feel it's necessary. It's too easy these days to find answers to text problems (especially popular texts!) on the web.

My favorite types of problems, in fact, are those where I find wrong answers on the web. I truly relish these. Over time I have found wrong proofs published in solution packs from professors in top-notch places, UC San Diego, Sanford, MIT, etc. Perhaps they're doing it on purpose, and if so... I am grateful to them.

In my quest to use original problems, Math.SE (and similar sites) creates a problem: not only does asking here spur 20-30 PhDs to vigorously compete for rep, racing to answer my students' homeworks, but it also permanently renders the problems useless. So I'm forced to find new problems every term.

I was hesitant to suggest my approach as "an answer" to your question, since adopting it requires a ton of extra work. But it's what I do, despite the fact that online forums teeming with experts is making this approach less effective with each passing semester.

$\endgroup$
8
  • 1
    $\begingroup$ What's worse, you won't be able to simply "tweak" your prior problems since it too may already be solved by a more general answer, see the abstracted common questions $\endgroup$
    – Bill Dubuque
    Apr 21, 2011 at 3:56
  • 29
    $\begingroup$ Most of the people who vigorously compete to answer questions are not PhDs, and most of the PhDs couldn't care less about reputation. $\endgroup$
    – Qiaochu Yuan
    Apr 21, 2011 at 6:32
  • $\begingroup$ If your questions are all original, wouldn't it make it easier to tell when the students are asking them on a forum? Seeing how you don't have false positives to worry about. $\endgroup$
    – Willie Wong
    Apr 21, 2011 at 11:04
  • 1
    $\begingroup$ I agree that this is another good option, composing your own homework problems. I don't do it simply because it is a lot more work. $\endgroup$
    – GeoffDS
    Apr 21, 2011 at 12:52
  • $\begingroup$ @Willie: Yes, I can tell when my homework questions are being asked on forums (assuming google indexes these forums fast enough). But user9017 isn't really going to reveal who he/she is, usually. It just confirms that it's being done, which we knew already. $\endgroup$
    – Fixee
    Apr 21, 2011 at 15:03
  • 4
    $\begingroup$ This is an awesome answer. I suspect that there is a method of designing problems so as to make them so obnoxious, i.e. special case with arbitrary numerical constants computed to high accuracy, etc., that the stack exchange community will refuse to assist with them. $\endgroup$
    – Carl Brannen
    Apr 21, 2011 at 20:53
  • $\begingroup$ Carl: in Philip Stark's "SticiGui" materials for the online teaching of statistics, the software automatically generates problems that are conceptually the same but differ in the numbers. I can see this working well for heavily numerical problems; it seems like it would work less well for more theoretical things, though... $\endgroup$
    – Michael Lugo
    Apr 21, 2011 at 22:17
  • $\begingroup$ Another advantage to setting your own h/w is that the students get used to your question style. I find that in intro classes, students are easily thrown by the difference in tone between textbook problems and the ones the TA or I set on the exams. $\endgroup$
    – Sam Lisi
    Apr 25, 2011 at 12:06
16
$\begingroup$

The era of the internet is upon us can be thought of in another way. You may very well have students not asking questions on here because they already have a copy of the full instructors' solutions manual, because these days you can download these as PDFs on various file sharing sites. At least, I know many of my Calc 1 students had this a year ago. I caught 9/40 cheating on one homework assignment, mainly because of one question where they all had a very similar answer and used something we had not talked much about. I gave them all 0s, told the class I knew about and told them not to do it. And, still I had 2 of these same people do it again and 1 other student do it who was not caught on this assignment. This situation is worse than asking on here because it takes no effort at all. You don't have to ask any questions. You can do your entire assignment in 10 or 15 minutes of copying out of the manual. One problem though is these students don't learn anything so most will do terrible on tests.

Since then, I assign homework but never collect it. I give weekly quizzes. This encourages them to do the homework weekly and not wait until right before the test to do a month's worth of homework. I also usually don't make them all that hard so it still has the benefit of giving easy points to the students, as homework does. I even do 12 of them and drop the 2 lowest.

$\endgroup$
11
  • $\begingroup$ +1: this issue has been worrying me since I started composing the homework FAQ for this site, and the only workable ideas I have found are either yours (graded quizzes with ungraded homework) or Cambridge style (the one exam to rule them all). For various reasons the former is much more appealing. $\endgroup$
    – Willie Wong
    Apr 21, 2011 at 11:06
  • $\begingroup$ +n! This is similar to what I would prefer to do. I assign the homework, examine it for errors, return it, and give 100% to pretty much anything handed in. The grading motivation to cheap is short-circuited. And in class I make it very clear that the homework, while required, is only there in order to prepare the students for the rigor of the exams. And weekly exams are a great thing, ITT Technical Institute does this in their math classes and I believe it helps a lot. $\endgroup$
    – Carl Brannen
    Apr 21, 2011 at 20:56
  • $\begingroup$ I've had success using this method too. The students are free to solve the homework in whatever way they can (using math.se, each other, the solution manual, the TA) and then are responsible for understanding the solutions enough that they can re-do the work on the quiz. The one point I am still figuring out is what mix of hard/easy questions to have on the quiz, and whether I should give slight variants on questions or take them verbatim from homework. There are arguments in support of each of these. $\endgroup$
    – Sam Lisi
    Apr 25, 2011 at 12:02
  • $\begingroup$ "I assign homework but never collect it." This is just plain awful. Why would anyone take a course where their progress can not be checked, where their errors can not be corrected? That's just totally bizarre. Especially if the course charges a tuition. I think the whole state of American higher education will require a reexamination soon. $\endgroup$
    – Apprentice Queue
    Jan 12, 2012 at 23:43
  • 3
    $\begingroup$ Read my entire response. Do not take one sentence out of context. All you have to do is read the next sentence. I give weekly quizzes and those are what checks students' progress. Weekly. Every week they can see how they're doing. I also have their overall grade for the entire semester available to them ALL THE TIME. I also encourage them over and over to come to office hours if they need help. When I teach a course, my average ratings are usually around 4.5/5 and I regularly have students say I'm the best teacher they have had. But, you know all about my class from one sentence. $\endgroup$
    – GeoffDS
    Jan 13, 2012 at 15:02
  • 1
    $\begingroup$ I also like how you have judged, not just my class, but all of the United States of America, by my one sentence. I also like how you did that without even knowing if I'm from the United States of America. Saying just America is not well liked by those people who live in South America, for example, as if the United States of America is the only America there is. $\endgroup$
    – GeoffDS
    Jan 13, 2012 at 15:05
  • $\begingroup$ @ApprenticeQueue Just to be sure you know the previous two comments exist, I forgot to address you in them. $\endgroup$
    – GeoffDS
    Jan 18, 2012 at 16:04
  • 1
    $\begingroup$ @Graphth it is difficult to test a student with only quizzes because then the problems have to conform to the very tight time constraints. Homework problems allow more time to be spent on them. Unless the final answer is provided in the homework problem, they will have little feedback. My comment on American higher education comes from seeing in so many classes with very smart professors who are awful teachers; there is no incentive for them to do well because those end-of-class evaluations are rarely acted upon and some of positions depend on research performance not teaching. $\endgroup$
    – Apprentice Queue
    Jan 30, 2012 at 6:52
  • 1
    $\begingroup$ @Graphth and the American colleges have allowed this to persist for decades. Now that tuition is unbelievably high, colleges are looking more like a pension system for PhD holders. I see a time when students will demand serious accountability because so much high quality books AND videos are already available for a smaller fee or even for free. $\endgroup$
    – Apprentice Queue
    Jan 30, 2012 at 6:57
  • $\begingroup$ @ApprenticeQueue Alright, now that makes sense. But, for instance, the class I'm teaching now, there are no problems that are too long for a quiz. The longest homework problem will take a good student 5 minutes, or less. Now, if I'm teaching Calc 1 or 2 or something, there are problems that would take longer than that. And, yes, those may be important. But, I can test all the most important concepts in a weekly 20-25 minute quiz still. And, I am not giving credit to cheaters, and having to deal with hours long processes of turning in cheaters. $\endgroup$
    – GeoffDS
    Jan 30, 2012 at 14:44
  • $\begingroup$ And, it's up to the students to put in the work either way. And, yes, there are terrible professors in the USA. And, yes, with the rising tuition, it's sort of dumb. And, I've heard that it's getting to the point to where your overall financial situation (on average over all people) is about the same if you don't get a college degree, because you end up with higher pay but a huge amount of loans. $\endgroup$
    – GeoffDS
    Jan 30, 2012 at 14:50
14
$\begingroup$

I thought that my Calculus professor was rather clever in this regard. He simply wrote a set of question numbers from the textbook at the end of each lecture, and each set were worth the same number of points. He told us upfront that the way the point-system was set up, we could get an A in the class without doing any homework, because the homework was only a couple points tacked on to an examination grade. He hinted that if we understood how to solve the homework problems, we would not be surprised by the examinations.

For me, this tactic changed the economics of the homework. We weren't rewarded enough for turning in a completed homework assignment for cheating to be worthwhile. The real reward in completing the homework was in the knowledge that it would prepare you for the upcoming examination. The homework was more like a pre-test for our benefit, as opposed to sadistic busywork. It simply didn't make sense to cheat on the homework.

As a result, the professor didn't have to concern himself with cheaters because they would end up failing the exams. If they were able to cheat and still pass, they really didn't need to do the homework. And students who wanted to cut corners could just not turn in the homework at all and face no penalty. Everyone saved a lot of time while reaching the same end. All that is required for this to work is that the tests be sufficiently difficult, which I would think is much easier when you can focus on them instead of making a whole lot of unique homework problems to prevent cheating there.

$\endgroup$
3
  • 5
    $\begingroup$ My discrete math professor told us back then, that the only reason they require homework to be handed in, is to ensure that we think and rework the material from class. If they knew we would do that, they wouldn't bother asking for the assignments to be handed in. But alas, if they wouldn't ask for us to submit homework, we would spend all our time and resources working on other courses' assignments and not on any discrete math problems. $\endgroup$
    – Asaf Karagila Mod
    Sep 8, 2014 at 21:29
  • 1
    $\begingroup$ (My policy for giving out homework to my students is to make it very very interesting. Last year I was told by some students that they spent more times on a 3 credits naive set theory course homework than they did on a 5 credits complex analysis course homework. And I am sure that they didn't spend that much time just because they had to. I might add that it worked, and the average passing grade was 86/100 or so.) $\endgroup$
    – Asaf Karagila Mod
    Sep 8, 2014 at 21:31
  • $\begingroup$ @AsafKaragila That's frequently true and I understand. I admit that this professor's tactic was really more psychological than anything else. He artificially lowered our overall grade by making the tests harder and then gave us the opportunity to make up the difference with homework. Other classes using a weighting system that would ultimately come out the same final grade can "feel" like homework is a punishment or chore. Ultimately, my only point was that since it's nigh unto impossible to prevent cheating on homework, might as well limit it's contribution to the grade. $\endgroup$
    – David Schwartz
    Sep 8, 2014 at 21:57
7
$\begingroup$
  • plagiarizing is bad.
  • online Q&A services provide an easy means for it.
  • it also makes it easy to see when plagiarizing occurs.
  • SE is nice in that it lets questioners (and answerers and moderators) be open about the intension of a question, with the 'homework' tag (but it's not really for finding plagiarizers).
  • I don't see what more SE can do technically to make it easier to deal with.
  • But the restatement that you suspect plagiarizing is occurring reminds others to consider that it occurs (either in how they answer, or how the answer can be used to help with homeworks or help with finding plagiarizers).
$\endgroup$
5
  • 21
    $\begingroup$ SE allowing people to declare the fact that they are asking homework questions is as useful as the fact that evil people are allowed by the universe to walk around with a "I am evil" sign... $\endgroup$
    – Mariano Suárez-Álvarez
    Apr 20, 2011 at 1:11
  • 4
    $\begingroup$ @Mariano: en.wikipedia.org/wiki/Evil_bit $\endgroup$
    – Chris Eagle
    Apr 21, 2011 at 17:10
  • 1
    $\begingroup$ @Chris, I know, I know... :) $\endgroup$
    – Mariano Suárez-Álvarez
    Apr 21, 2011 at 19:52
  • $\begingroup$ @Mariano I don't know the full mechanics of math.SE, but I know on SO those who do declare their posts to be homework are immediately met with more respect (and assistance!) then those who are obviously asking homework questions but have not put the homework tag. That's not to say they have it "easy" per sé (many of them are still greated with the standard "what have you tried so far?" routine, which is, in my opinion, fair.) $\endgroup$
    – corsiKa
    Apr 22, 2011 at 23:18
  • 3
    $\begingroup$ Well, the standard "what have you tried so far?" routine is what any good teacher would do if a student came and asked them for help on a problem. (I like to think I am a good teacher.) $\endgroup$
    – Michael Lugo
    Apr 23, 2011 at 15:26

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .