The outcome of the homework tag discussion is not clear so far: there is a slight majority in favor of having a tag, but no agreement as to what the tag means. The meanings I've seen proposed in discussions on other Stack Exchange meta sites and in the CS chat are:

  • The person asking the question is enrolled in a course and the question is asking to solve or help solving a homework problem that was set as part of the course.
  • The question is asking to solve or help solving a learning exercise with artificial constraints, as opposed to a real-world problem.
  • The question is asking to solve or help solving a learning exercise, which may or may not be an interesting theoretical or practical problem.
  • The question was set as a homework problem somewhere.
  • The question could be set as a homework problem.
  • The question is asking for a hint as opposed to a complete solution. ( has been proposed for that meaning).

I may be missing or misconstruing some definition proposals.

If we have a homework tag, what should it mean? When should it be used?

Note: while the debate about the use of the tag is going on, please do not add the tag to someone else's question or request that they do so, and please refer anyone suggesting to use or not to use the tag in a comment to this discussion.

Status at the 9 day mark: we have an answer with +9/-0 score advocating the ban of the tag, and an answer with +0/-0 score advocating a Math.SE-like policy. Unless someone objects in the next few days, it looks like we're going to ban the tag by consensus.

  • 4
    $\begingroup$ We were trying to solve the issue about the homework tag, now we have a homework tag on meta! ;) $\endgroup$ – Kaveh Apr 29 '12 at 20:13
  • $\begingroup$ I don't see this question as a referendum, and taking it as such is, I think, a bit unfair. If we want a referendum, I think you (or someone) should lay it out explicitly and ask for votes. Your six bullets might be the choices, with the addition of a "no homework tag at all" option, or perhaps some other formulation, such as simple: yes/no. $\endgroup$ – David Lewis May 15 '12 at 6:39
  • 1
    $\begingroup$ @DavidLewis The point of this question was to establish the options for a referendum, or, better, find a consensus. So all we have is “no tag” and “follows the Math.SE policy” (which Math.SE is thinking of changing), with a very clear winner. What would be the point of yet another thread at this point? $\endgroup$ – Gilles 'SO- stop being evil' May 15 '12 at 7:47
  • 1
    $\begingroup$ @gilles -- I, for one, did not take it as a referendum. To me it was just a discussion of some particular points and one incident. If you want a valid referendum, it seems to me you need to be very explicit about it -- put "referendum" in the title, say exactly what you are deciding, set clear "candidates" and embody each in a pre-set answer, set clear voting rules such as the end date/time, how to handle comments (like 1/person, say they are "electioneering"), etc etc. It's really not fair to start a discussion and then, ex post facto declare it to have been a referendum. $\endgroup$ – David Lewis May 15 '12 at 11:59
  • $\begingroup$ @gliies -- also, you might consider announcing an actual referendum in a short post in the main question area that points directly to it, and refresh it (somehow) every day until the deadline. Otherwise people who don't visit the meta area or do so infrequently might never know there is a referendum going on or find out too late to vote. Also, in terms of voting, is it one-member-one-vote for anybody visiting the area? A weighting function of reputation? A simple reputation threshold? Etc. $\endgroup$ – David Lewis May 15 '12 at 12:05
  • $\begingroup$ @gilles-- I also suggest consulting the staff in Area 51. It seems their mission is to help sites be healthy, and there are folks there with a lot of experience working with online communities. Can't hurt. $\endgroup$ – David Lewis May 15 '12 at 12:08
  • 1
    $\begingroup$ @gilles -- sorry last one. It seems to me that 11 votes for a post that has a clear 1st line, against 1 vote for a post that doesn't but lays out a personal preference down inside does not constitute a meaningful "vote" in any sense of the word. $\endgroup$ – David Lewis May 15 '12 at 12:11
  • 1
    $\begingroup$ @DavidLewis Again, I am not declaring this a referendum — but the paucity of voices speaking for the homework tag means that most people either are against or don't care. I expected more responses, but it's been two weeks, which is plenty of time. If we do have a vote, anyone with 15 reputation on the site can participate, and it's one vote per user. I don't know yet what we'll end up doing: maybe we'll vote on how to vote about voting rules, or maybe we'll do something more sensible. I have over a year's experience moderating a SE site in beta, by the way. $\endgroup$ – Gilles 'SO- stop being evil' May 15 '12 at 18:52
  • 1
    $\begingroup$ @Gilles -- I understand, and I'm glad you aren't declaring this a referendum. I just don't think it's valid to draw any conclusions about what "most people" think about an issue -- for, against or don't care -- until you specifically ask them for their opinion in nice, easy well-publicized form, like a vote. Otherwise, you just get what a random sample of us discussion/meta junkies think and are willing to spout about, in the context of one particular discussion from a limited set of angles. $\endgroup$ – David Lewis May 15 '12 at 22:30
  • 1
    $\begingroup$ @DavidLewis The problem with a vote is: voting about what? There have been plenty of proposals on what the tag might mean. That's why we discuss before voting (and ideally instead of voting): to see all the angles. In a vote, if you aren't 100% behind one of the proposed options, you're stuck. The three site moderators are currently discussing what meta rule we'll apply to reach a decision, which may mean what options we'll put forth in a vote, or not. $\endgroup$ – Gilles 'SO- stop being evil' May 15 '12 at 22:48
  • 1
    $\begingroup$ @Gilles -- Yep, that's how elections work. You have a number of choices and you pick the one that you favor. If you can't do that, you abstain. The alternative is free-for-all town meeting style debate, which has serious limitations. But to make decisions, even that ultimately needs a proposal on the floor and a yea/nay vote. Good luck with your deliberations! $\endgroup$ – David Lewis May 16 '12 at 2:47
  • $\begingroup$ @DavidLewis: I expect such a vote to be featured which will make it show up in the new Community Bulleting in the right sidebar (on all pages). $\endgroup$ – Raphael Mod May 17 '12 at 16:46

After some reading and thinking, I think the tag has no right to exist.

Tags are supposed to categorize and make searching easier. They are used to find questions that hopefully answer questions users might have (before they ask them and have their question closed as a duplicate), or to allow easier searching for questions that a user may be able to answer (I know quite a bit about parsing, so that tag is interesting for me).

The tag does neither. It does not narrow down where a user's question might be answered (what is homework for one person may not be homework for another, or your homework may be a special case of a more general question). It does not help help anyone answer questions (I can't answer homework questions about database theory, or at least nothing nontrivial).

Furthermore, a tag should not be needed to understand the question. If I ask a question about 'variables', and if the only way to know I'm talking about 3SAT instead of programming languages is through the tags I use, then my question needs to be edited to include what I'm talking about.

The same thing should apply to the tag. If a question is homework or if the user wishes to only get a hint, he needs to say so in his question, and not through his tag use.

To invoke SE policy, the tag is a meta tag: see here.

A separate issue is therefore whether we allow homework questions (rather than the tag), and how we treat them. I think our current (implicit) policy works very well: if the user makes it clear the question is homework, or we guess it is, we provide a hint, along with the full answer in a spoiler tag.

This gives us the best of both worlds: users that only want a hint only get a hint, yet the answer is still there if they get stuck, or if a future visitor is more interested in the answer than how to get to the answer. This also prevents issues if we guess wrong and a question is not homework: the answer will still be there.

  • 1
    $\begingroup$ I can imagine that some professors may want to search the questions related to homework to understand where many students have trouble understanding. For this purpose, the [homework] tag may come in handy. But this is probably a rare use case. In addition, I agree that saying “This is homework and I need a hint” explicitly is much better than the obscure [homework] tag. $\endgroup$ – Tsuyoshi Ito Apr 30 '12 at 12:04
  • 1
    $\begingroup$ +1, although I might go a little further: why should the community encourage giving incomplete answers for homework questions, or using spoiler tags? It's not in the community's charter to police morality. If users don't want the whole answer, they won't ask for it. If the user wants an answer to a part of the problem, they can ask a more specific question. $\endgroup$ – Patrick87 Apr 30 '12 at 14:46
  • 5
    $\begingroup$ @Patrick87: Because we want to help people, and giving a detailed solution is not always the best strategy. Provoke thought first, given solution second -- I like the spoilter tag thing. $\endgroup$ – Raphael Mod Apr 30 '12 at 17:01
  • $\begingroup$ @Raphael It's not our job to be TAs here; it's to answer questions. I think people can be better served by getting better answers to the questions they ask. If you prefer using spoiler tags, that's your prerogative; but I don't think we should make it a policy to expect their use on homework questions. Certainly, complete answers without spoiler tags should not be penalized in any way, right? $\endgroup$ – Patrick87 Apr 30 '12 at 17:12
  • $\begingroup$ "not penalising" and "encouraging the alternative" are two different things, and I think we should do both, if only by example. $\endgroup$ – Raphael Mod Apr 30 '12 at 17:46
  • 3
    $\begingroup$ I'm not a proponent of forcing people to use spoiler tags in any way, but merely of convincing as many as possible to do so (by setting an example). The reason for this is that I think people are helped more by not only by giving them an answer, but also helping them understand how to get to that answer. Furthermore, it does not hinder our ability to give great answers (all the information of the answer is still there), so I see only positive sides. $\endgroup$ – Alex ten Brink Apr 30 '12 at 21:34
  • $\begingroup$ @Raphael It sounds to me like what you're suggesting is social suasion by means of peer pressure. I could see that turning ugly, although that possibility may be far-fetched. For instance, perhaps this gentle urging by the powers that be could cause some people who post answers with spoiler tags and/or hints to revenge downvote people who don't? Anyway, all that's beside the point. To me, the real point is that it's not within our charter to try to educate people. It's to inform through answering questions. I fail to see a need to indulge students' request for special treatment here. $\endgroup$ – Patrick87 May 1 '12 at 14:57
  • $\begingroup$ If they want a partial answer, let them ask a question whose full answer is the partial answer they seek, and let it be given a full answer. No need to hide it or hint at it. This is, of course, all just my opinion. $\endgroup$ – Patrick87 May 1 '12 at 14:59
  • 1
    $\begingroup$ This is not about our charter or the askers' (perceived) needs. This is purely about what I think is a good answer. I (and others who agree) can posts answers in that way (i.e. lead people to the answers themselves, if applicable, and provide details/complete solutions behind spoiler tags) and if they are good, they get upvoted and can therewith set an example. No need to overthink this; everybody does what they think best, the system takes care of the rest. There is one exception, though: if the OP asks for hints, we should take action (editing, downvoting) if answerers do not respect that. $\endgroup$ – Raphael Mod May 1 '12 at 19:15
  • 3
    $\begingroup$ @Patrick87: I'm a proponent of convincing by example only (so not though other means). The reason I want to treat homework differently is twofold: I think students who just want the correct answers are missing the point, are not learning much from the answer and are deluding themselves by taking the easy way out - I think we should try to save them from this. Secondly, using answers from this site to answer homework (that is graded) is fraud: do we really want to encourage and assist these illegal acts? $\endgroup$ – Alex ten Brink May 2 '12 at 11:11
  • $\begingroup$ @Raphael I categorically disagree that we should honor penalize answers that don't honor requests for hints. You're right that ultimately it's every individual user's decision how to treat any given answer... but good answers should not be discouraged, even if people request incomplete answers. Unless I misunderstood your meaning, what you suggest is exactly what I was afraid might happen in the worst case. $\endgroup$ – Patrick87 May 2 '12 at 12:26
  • $\begingroup$ @AlextenBrink As long as retaliation against people not sharing your beliefs doesn't get out of hand, it's not a problem. Everyone is, of course, allowed to have their opinions here and act on them as they see fit. If you happen to convince some people along the way, no problem. I, for one, remain unconvinced of the need for us to worry about either of the points you mention. They're problems we shouldn't waste time trying to solve. $\endgroup$ – Patrick87 May 2 '12 at 12:31
  • 2
    $\begingroup$ @Patrick87: apparently I need to make my stance abundantly clear: I personally think it advantageous for students to answer homework questions in this way, so I will use this format. If asked (only), I will advise and encourage other users to do the same. I'm not a proponent of making a policy out of this or 'retaliating' against people using different formats. You will not see me make any comments on answers in a different format with the suggestion to use my format instead. I hope others will agree with me and follow my example, but nothing more. $\endgroup$ – Alex ten Brink May 2 '12 at 14:08
  • $\begingroup$ @Patrick87 As I understand the definition of "good answer" on SE, it is an answer that answers the question in a good way. I can write good stuff that does not answer the question, and that should be downvoted. So if somebody asks for hints and another person provides a full-fledged answer that maybe, in the worst case, even hides the central points, then this does not answer the question as stated and can reasonable be down- or at leats not upvoted. That has nothing to with what I consider best, but what answers the question. $\endgroup$ – Raphael Mod May 2 '12 at 19:49
  • $\begingroup$ If the OP does not specify wether hints or full answers are expected, I would of course not actively penalise full answers. $\endgroup$ – Raphael Mod May 2 '12 at 19:49

I presume this discussion was at least partially occasioned by my recent behavior. I do apologize for what I did -- here is an explanation, not an excuse.

When I first came on SE it was on the mathematics site. I thought I observed pretty consistent behavior there by those answering -- if it's tagged as homework, just give hints; if it's not so tagged but "smells" like homework, many people asked the OP what was up. IOW, the policy seemed consistent, widely practiced and made sense to me, so I adopted it.

When I later started answering here and in Theoretical CS, I did not stop to determine if the same understanding of the homework tag was well accepted. I did note that quite a few folks were giving full answers to questions that "smelled" like homework to me, but I still tried not to do that myself, on the assumption that those folks just hadn't gotten the message.

With the case in question, someone posted a sequence of questions over five days, most of which in retrospect, "smelled" at least a bit like homework to me, but which I had been answering in good faith. In fact, someone actually asked the poster on their first question if it was homework, and I chimed in that we should trust them. But as the questions continued to arrive from the same poster, I began to suspect they were indeed homework, and he was hoping to get answers to his assignments. So I asked him after about 3 days, and he replied that it was for a test, not homework. I accepted that for the moment, but as he posted even more questions over a total of five days, I finally got a little snarky, at which point things did blow up. He reiterated his claim that he was studying for a test, so I apologized to him and wished him luck.But he has not reappeared.

So I apologize for getting snarky -- there's never a good excuse for that -- and for transferring my understanding of homework tagging policy to from Math to CS without checking in any way if that was the prevailing norm... and I guess it isn't, at least judging by this thread.

A final note -- IMHO, the way the homework tag is handled in the Math area is good, and I'd like to see the same policy here. But if there is a different understanding or a lack of clear consensus, I will go along with the community, to the extent I can discern what that range of agreed behavior might be.

Actually, one more note. I really should not have gotten snarky even if I was sure the poster was trying to get homework answers to turn in. Such a poster may get away with it for awhile, but it is generally self-limiting in a number of ways. First, if they do it a lot, it will probably raise suspicions here eventually and they will have to stop. Second, a decent teacher should be able to detect when a student's answers suddenly take a leap in accuracy and clarity (assuming we provide those qualities). Lastly, there really is no way to police this and we can't even try to be cops -- we have to have some trust in the honor systems of posters' schools, and if those don't work, it's not our job to put them straight.

All this is complicated for me by the fact that I am a former teacher, though from a time before academic dishonesty became as widespread as I gather it is today. Plus, I was pretty sensitive to dishonesty, and went through a number of serious and sometimes harrowing cases of cheating, whose details I will spare you.

In any case, again I am sorry that I acted so rashly and started a bit of a tempest.

--David Lewis

  • $\begingroup$ Just to be clear, it definitely was not my intention to make you feel like you'd acted inappropriately. I was simply responding to a flag and related comments that were detracting from the question. How the community wants to handle homework questions is still being discussed, but tentatively it looks like I'm in the (albeit vocal) minority, and the policy will probably eventually be very similar to what you're used to on Math... $\endgroup$ – Patrick87 May 3 '12 at 3:18
  • $\begingroup$ Regarding the current case, iirc the user studied for a multiple choice test (a stupid way to test TCS or any skill, but well) and had the solution key. He wanted to understand why he had to cross that way, and that is worth support (imho). The homework discussion is endless, and I think the main argument against a policy as on math.SE is that we can never tell wether a question is homework or motivated by self-study, so why bother? $\endgroup$ – Raphael Mod May 3 '12 at 8:13
  • $\begingroup$ @Patrick87 -- I didn't think you were chastising me. In fact, I appreciate the even-tempered responses all around. As for the actual argument, I agree with you and Raphael that we can never tell. But it's still good to bother, IMHO: (a) for the honest students; (b) to present a good model for ethical practice; (c) so classroom instructors don't get negative about SE. $\endgroup$ – David Lewis May 3 '12 at 13:09
  • 4
    $\begingroup$ This question was not occasioned by your behavior (or only very indirectly, if you're one of many people who posted a comment that implied that we have a policy on the use of the homework tag, which we don't (yet)). Each Stack Exchange site has its own policy; some have banned the tag altogether (e.g. Software Engineering), others have a detailed policy (e.g. Mathematics), others are still fighting it out (Stack Overflow...). This site is young, and we're debating the policy now. Thanks for contributing. $\endgroup$ – Gilles 'SO- stop being evil' May 3 '12 at 19:17
  • $\begingroup$ @Gilles -- thanks for the clarification. My "vote" is for the Mathematics site's policy. And meanwhile I'll probably silently behave that way here. BTW, how does it get decided -- is there a voting mechanism? $\endgroup$ – David Lewis May 3 '12 at 22:55
  • $\begingroup$ @DavidLewis Hopefully we find a position that's acceptable to everyone, or at least to most. We prefer consensus to voting. If we can't achieve a consensus, we might vote — but we first need to figure out what the options are. In a pinch, moderators can decide, this has happened on other sites when the community was nearly evenly divided on an issue with no middle ground — but this is a last resort. $\endgroup$ – Gilles 'SO- stop being evil' May 3 '12 at 23:03
  • $\begingroup$ In case you haven't noticed, we're trying to collect opinions on a homework policy. So far your view hasn't been represented on this thread; it would help (both your position and the community in general) if you chimed in. $\endgroup$ – Gilles 'SO- stop being evil' Jun 25 '12 at 22:42
  • $\begingroup$ @Gilles -- it wasn't clear how I was supposed to represent my position, so I left a comment to the original post. OK? $\endgroup$ – David Lewis Jun 26 '12 at 3:05

You must log in to answer this question.

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