# Generalizing the homework policy

In our last chat session two weeks ago (sorry for the delay), we had a discussion about updating the homework policy and homework close reason. I'm making this post to summarize that discussion and solicit feedback on how to proceed afterwards.

# The current status

According to the homework policy, questions are considered "homework-like" (or "educational") and fall under the policy whenever

...[the question's] value lies in helping you understand the method by which the question can be solved, rather than getting the answer itself.

However, it's not always clear whether a question fits this description just from its content. And if you track the questions that we actually close these days using the homework-like close reason, quite a few of them are likely not of an educational nature. Instead, we've taken to using the homework-like close reason on questions that simply ask us to calculate something without the original poster making an attempt at it. We do have a policy on showing effort, separate from the homework policy, but it doesn't justify closing those questions. (Also, it was meant to catch questions that the homework policy doesn't.)

It seems that most people believe that questions where the poster asks us to do some simple calculation for them should be off topic. However, it's not clear exactly where we should draw the boundary of what sorts of questions are off topic, and what reason is best to give for considering them off topic. There have been a number of discussions on this in the past, among them:

# The options

We will probably want to create a new policy that replaces the current homework policy, and gives a different criterion for a question that is off topic. This new policy would become the new justification for closing low-effort homework questions (those where it is clear that it's a homework/educational question), so whatever we wind up going with, it should be some criterion that does catch those questions.

There are a few options that I can identify:

1. Questions which ask us to perform calculations are off topic
2. Questions which don't show sufficient effort are off topic (I suppose this is the nice wording for the poster being lazy)
3. A combination of the previous two - kind of like our current homework policy, except that the criteria of "ask a conceptual question" (more or less) and "show effort" would apply to all questions, not just those we consider to be of an educational nature

Those are the most likely contenders, but for completeness, some other options:

1. Questions which ask for a conceptual understanding of something are on topic, and everything else is not (this is kind of like #1, but excludes more questions)
2. All questions which come from homework assignments are off topic (difficult to enforce)

etc. etc. And for comparison, our current policy would be something like

1. Questions which are educational in nature are off topic unless they ask a conceptual question and show effort

# The issues

We have some ideas for what kind of criterion should replace our current homework policy. For each of these ideas, we should consider:

• Are there a significant number of questions which are off topic under our current homework policy which would be on topic under the new policy? If so, should they be on topic?
• Exactly what sorts of questions would become off topic under the new policy which are not covered by current close reasons? Do we want them to be off topic?

Hopefully the answers we come up with will help us focus the new close reason.

At some point in the future, we will also have to discuss the role of the tag in this new policy.

• Are we going to have straight votes on #1 through n (i.e., one answer here per point above, plus whatever else is posed)? – Kyle Kanos Jan 12 '16 at 16:39
• @KyleKanos I don't intend to put anything up for a straight vote at this point. Maybe later. – David Z Jan 12 '16 at 16:52
• I'm not sure we can discuss the topicality of homework questions without really addressing the tag at the same time. In the end, it will come to putting it down in writing, in a shared policy that embodies the consensus of the site, and that needs to include the tag wiki and related documents. And as we've seen in the past, it's surprisingly hard to write consistent policies that include that tag wiki and its excerpt (and hopefully one of the new tag warnings), which are the front lines of telling new users what the deal is. – Emilio Pisanty Jan 13 '16 at 10:55
• Related discussion at electrical engineering: meta.electronics.stackexchange.com/q/5667 – DanielSank Feb 28 '16 at 20:04
• I would note that this post doesn't really ask a question. Under the section labeled The issues there are two questions, but they have trivial answers ("no", and "none") and probably aren't the main intended point of the post. – DanielSank Sep 21 '16 at 8:39
• Just pointing out that those are exactly the main intended point of the posts, but it's those two questions for each proposal. So there are at least 12 questions, and there would be more if anyone came in with a new proposal separate from the ones I listed in the question. – David Z Feb 22 '17 at 22:36

1. Questions which ask us to perform calculations are off topic.

This is too broad. I recognize that it's intended to head off boring copied-from-homework questions like "what's the optimal angle for a 45 mph banked turn if the coefficient of friction is μ = 0.233457821234 also are all those digits important kthxbye". However calculation questions like "what's the average antimatter content of a banana?" or "with my eyes closed at sea level, how often should I expect to see Cherenkov flashes from cosmic ray muons in my vitreous humor?" are interesting entrées into a host of more complicated topics. It'd be the responsibility of the asker to elaborate on which of these other topics is really at issue.

In fact I find that turning my conceptual questions into model calculations gives me better answers than asking in vague terms, and I'm loath to see such questions closed. Furthermore, many of my favorite questions on this site have inspired clever calculations from answerers; I would prefer to encourage these sorts of answers with inviting questions.

I might propose as an alternative:

Questions which attempt to outsource tedious calculations to the community, without any broader context, are off-topic.

I recognize that a key word here ("tedious") implies a value judgement, and there's a bias in writing these guidelines towards "objective", judgement-free criteria. That bias is flawed, which is why we have human moderators and the opportunity to discuss some decisions with them.

• Oh, and: the potassium in a "medium" banana emits a positron once every 75 minutes on average, and I still haven't seriously tried to answer the muon rate one. – rob Jan 12 '16 at 20:02
• Should I be more afraid of my bananas now or the person who comes up with these questions? :-) Having said that, I do support the assertion that some homework questions are fundamentally interesting enough to be considered... and we don't need formal rules for those. I think we are all pretty smart people around here who can tell the difference. – CuriousOne Jan 13 '16 at 0:30
• I agree that a question like "what's the average antimatter content of a banana?" can yield some interesting answers, but if that's really all there is to the question, I think it should be closed. I'd want to expect the asker to have made some effort by searching Google, trying to set up some assumptions that make a calculation possible, and trying to work through that calculation. To put it another way, I think that's a fine premise for a question to be based on, but not a good question itself. So "outsourcing tedious calculations" is IMO not restrictive enough. – David Z Jan 13 '16 at 8:04
• I would not describe the calcuations in most of the low-level homeowork question we receive as "tedious". They're "standard", and that's why we don't want to do them over and over again. – ACuriousMind Jan 13 '16 at 14:34
• @CuriousOne They're my questions, and you should be afraid of potassium deficiency. – rob Jan 13 '16 at 15:42
• @DavidZ Of course my "good question" examples were only titles and/or first sentences, for rhetorical reasons. It's possible to screw up the execution of a great question, which is why there's a reopen process. – rob Jan 13 '16 at 15:45
• @ACuriousMind Perhaps in that case the more relevant word is "outsource". In many good questions, it's the asker who ends up doing the heavy lifting of putting the details on the answer; the answerer's job is to steer the asker in the correct direction. – rob Jan 13 '16 at 15:47
• @rob (2 comments up) fair enough, but in the cases we're looking at, I think what makes the difference between a good vs bad question (or, on-topic vs off-topic, the way it might wind up) is the execution, the way it's fleshed out in the body. Those titles could be for questions that are really good, or for questions that are terrible. So I don't think it's all that useful to just present a title/first sentence/summary; we need to get into how the question is expanded on. (I expect there will be more meta posts on this in the future.) – David Z Jan 13 '16 at 16:00
• @DavidZ Hence "interesting entrées" rather than "definitely excellent questions." – rob Jan 13 '16 at 16:04
• +1, but I disagree with the last sentence. If the rules include a value judgement it only takes five people to take a strong stance on that judgement and the question is closed, with no chance of appeal. The person asking can't predict in advance whether this will happen, which leads to endless frustration. The whole Stack Exchange system is designed to work with clear community rules that can be judged objectively, and it gets pretty broken if communities don't make a sufficient effort to agree on such rules and stick to them. (Physics.SE is one of the least broken in that respect.) – Nathaniel Jan 30 '16 at 5:52
• So perhaps it would be better to say something like questions that attempt to outsource calculations to the community without any broader context, or questions whose only purpose is to outsource calculations to the community. After all, your cherenkov flashes question may indeed involve a tedious calculation. Its value comes not from the calculation itself being particularly interesting, but from the intuition the answer would give us about the relative scale of a cosmic process compared to our everyday world. – Nathaniel Jan 30 '16 at 6:02
• @Nathaniel Good suggestion; I've edited it in. – rob Feb 1 '16 at 23:44
• Here's what I think is the best we can do. We can't see everything people are doing. We could instead tell them something like "Ask yourself, 'Why do I want the marks?' Then say it's probably to get a job. Then you could tell them that they only have time to take so many people into the job and train them. The people who can get higher marks without asking homework questions or looking at homework problems other people asked generally have a better creative thinking ability and are easier to train on the job to have a real understanding and do the job really well. Do you really want to take – Timothy Dec 11 '19 at 0:58
• the job away from somebody who can do it better and then have expectations to do things in the job that are too hard for you to figure out how to do which may cause a problem when you have a nice easy way to make the job get done better by not taking it away from somebody who can do it better and not having other people rely on you doing things that are so hard for you to figure out how to do?" I myself feel like purposely deciding that no matter how smart everybody is, we should make tests harder to limit the number of students who get really high marks is not a great way to school people. – Timothy Dec 11 '19 at 1:03
• It's too competitive and maybe also nobody has both the ability to get really high marks and a more general overall smartness seeing things from multiple perspectives and maybe they're actually weeding out every single person who's smart in the right way. Maybe some of them develop that smartness later but never would have gotten the job if they had it earlier. I think we should set a lower standard of what's good enough and get all the students to agree to cooperate rather than compete to decide who's going to take those jobs. Maybe they really do believe the job will get done best when they – Timothy Dec 11 '19 at 1:08

In my opinion, in this discussion there is too much focus on the "question" and not enough on the "value of the question plus the answer".

If the goal of the site is to be a "resource" to all serious students of physics, we need to make sure that the question-plus-answer becomes something of lasting value. "I need an answer to this homework question before 8 am tomorrow" is always a bad start to a question - but if the homework question is in itself interesting enough, the solution to that problem may have some lasting value. I have learnt quite a lot from answering such questions myself; for example, answering this question about a ball rolling down a slope clarified some concepts for me - so although the question was closed as "homework-like", I have actually referred back to the analysis a number of times in my answers to other questions (not all of which were homework-like).

As a second example, when I thought about the recent question about the frequency response of vinyl records I discovered some interesting relationships regarding distortion of a vinyl recording as a function of signal amplitude, track velocity, frequency of the signal, and radius of the stylus. That particular question was closed as "engineering", when in factI think that such analysis, while it is "application of science", is a nice demonstration of the power of physics-based reasoning.

I may be an outlier in this - but when I answer questions, I do it for my own benefit and that of the community, and not just for the OP. And that's the perspective I would encourage us all to use when we consider how to apply the close vote. That doesn't mean we should feed the lazy students solutions - but I, for one, am willing (perhaps more than others) to answer questions that have an "interesting" angle to them. Where "interesting" is of course a very personal judgment.

• It seems like you're saying that, if a question sets someone up to post an interesting answer, then that adds value to the question (I agree with that) and should cause us to be more lenient with our rules about homework, showing effort, or whatever we come up with (I very much disagree with that). Sure there are bad questions - where "bad" means closable according to whatever criteria we wind up using - that allow for interesting answers, but I assert that in each such case, the question could be edited into a good question that also admits the same, interesting answer. (cont.) – David Z Jan 14 '16 at 15:54
• Or alternatively, someone else could ask a good question that allows the same answer. You could even do it yourself, using the "add your own answer" function, if you have an interesting answer to a bad question that the OP shows no interest in improving. – David Z Jan 14 '16 at 15:55
• I agree that "bad" questions may need to be edited into "good" questions - but if you think the question is interesting nothing stops you from editing it yourself. Or as you say, ask/answer your own... My point was really about "adding value to the site" . – Floris Jan 14 '16 at 16:05
• Well, there is the rule(-ish) that significant changes should come from, or at least be explicitly approved by, the OP. In some cases, that should stop you. But in other cases, that is a valid course of action. Either way, it doesn't prevent putting the question on hold. Remember that when a question is put on hold, we expect that it will be edited into shape. If a question is "bad", we close it, let it get edited, and then if the edits fix the problems, reopen it. The argument that a question could be edited into a good one goes in favor of closing that question, not against it. – David Z Jan 14 '16 at 16:18
• @DavidZ - we mostly agree. – Floris Jan 14 '16 at 16:19
• I agree that answering mundane-sounding questions can be interesting, but I still think that if the question is really just asking for a calculation it should be edited such that it shows effort and asks a specific question about the part where OP is stuck, else it should be closed. Note that this doesn't change the potential value we can create by answering the question, it just means the OP has to identify a specific conceptual issue first. – DanielSank Jan 26 '16 at 17:47

I'll comment on the options for the new policy you name in order:

1. Questions which ask us to perform calculations are off topic

Generally yes, but questions which show that standard calulation methods in the field are not sufficient to obtain the result should be considered on-topic, because then there is very likely a conceptual argument hidden in some step of this calculation. This is a bit different from "requiring effort", because it requires a certain type of effort. No amount of effort shown should save a question whose answer is just pages of boring index manipulations, or solving some particularly resilient integrals.

1. Questions which don't show sufficient effort are off topic (I suppose this is the nice wording for the poster being lazy)

I once thought the proper course of action for those would be to downvote them. But downvoting them has not the desired effect of deterring people to ask such questions. The downvoted questions get pity upvotes (or upvotes from people as lazy as the asker), and get an answer sooner or later because the answer is, well, easy to find and it's quickly earned reputation.

Thus, because I see no value in us becoming a repository of questions which are already easily answered by typing the relevant keywords into a search engine, I have come around to the view that we should close effortless questions. This, however, is wholly distinct from the homework policy - conceptual question like "Why do rocks fall down?" can also be effortless. In particular, showing effort as such is not sufficient to make a question on-topic if there is some other reason for it to be off-topic.

1. A combination of the previous two - kind of like our current homework policy, except that the criteria of "ask a conceptual question" (more or less) and "show effort" would apply to all questions, not just those we consider to be of an educational nature

I don't think we should "combine" those policies - they are, to me, about entirely orthogonal issues. Questions which just ask us to calculate something are off-topic. Questions which are lazy are also off-topic. I don't see an intrinsic relation between those two types of question that would mean the policy should really be "combined". In any case, I think we can drop the phrase of questions "of an educational nature" entirely, I've never been wholly clear on what that means anyway, and it doesn't reflect any actual practice.

1. Questions which ask for a conceptual understanding of something are on topic, and everything else is not (this is kind of like #1, but excludes more questions)

Kind of a strange phrasing, and probably too prohibitive. In particular, this will lead to endless fights over what exactly "conceptual understanding" means, even more so than the homework policy we already have. Don't do this.

1. All questions which come from homework assignments are off topic (difficult to enforce)

Just...no. For one, because we cannot honestly enforce this, but even if we could, if there's a perfectly interesting conceptual question coming from a homework assignment, I do not see why we shouldn't answer it here.

Altogether, I say we enforce 1. and 2. as policies, with a bit of leniency on 1. if the question shows that the "calculation" most likely involves conceptual arguments. Of course it would be preferable to rephrase such question to not make them look as if they're asking for calculations, but I'm not completely sure this is possible in all cases. Additionally, we drop any mention of "educational nature" in our policies.

• Your idea of #1 and #2 together is precisely the sort of combination I was getting at with option #3. – David Z Jan 12 '16 at 16:53
• @DavidZ: Ah, then I probably just got a bit hung up on the word "combine". – ACuriousMind Jan 12 '16 at 16:55

A somewhat late answer, but it occurs to me that there are circumstances where doing a calculation in an answer may be appropriate.

I sometimes see questions that don't ask do this calculation but rather how is this calculation done? i.e. what are the concepts behind the calculation? For example look at my answer to Does juggling balls reduce the total weight of the juggler and balls? in which I basically just do a calculation. If the OP had asked:

Prove you can't cross the bridge by juggling the balls blah blah

then it would have been closed as homework and closed, and my answer (with 89 upvotes!!) would never have been posted.

It seems to me that I've written quite a few answers like this i.e. answers that are basically just calculations but their value lies in showing how the calculation is done. I imagine most of us would agree this is a fair way to answer, but it feels as though there is a resistance to any answers that are essentially just calculations and I wouldn't want to be prevented from answering in this way in the future.

• I think there is enough wiggle room in the policy to keep questions like this open, even if they do boil down to doing some calculations. Like the one for finding how deep a diver will go in the water from a platform dive, or whether the girls in Frozen could really survive jumping into snow from a tower. That's why I prefer to keep things a bit grayer. – tpg2114 Jan 20 '16 at 22:20
• This is exactly the sort of interaction that I had in mind. – rob Jan 26 '16 at 3:07

I think the most confusing bit of the homework policy is the "and show effort" aspect found in the close reason:

"Homework-like questions should ask about a specific physics concept and show some effort to work through the problem. We want our questions to be useful to the broader community, and to future users. See our meta site for more guidance on how to edit your question to make it better"

Especially when the link in that message opens with,

It's not enough to just show your work and ask where you went wrong. If you just need someone to check your work, you can always seek out a friend, classmate, or teacher.

Even aside from the fact that the community appears to be against "check my work" problems, I feel that including the "show work" bit is both contradictory and confusing. It suggests that if someone simply adds their attempt to their "do my homework for me question," then it'd get reopened and answered; even some high rep users are of the opinion that "showing work" would/should prevent closure.

As I understand it, the point of the "show work" is meant to encourage askers to not pose questions without having thought about the answer. I don't think that the current phrase encourages this position. It would be better if we did one of the following

1. eliminate the phrase so that questions asking us to "do some math for them" are off-topic
2. leave the phrase so that questions asking us to "do some math for them" are on topic
3. replace the phrase with one that more accurately represents what is meant by it

I personally think that #3 is the better option of these, though I cannot think of a good phrase outside of something along the lines of requiring OP showing their rationale--I'm completely open to suggestions.

• Perhaps "Homework-like questions should ask about a specific physics concept and provided a detailed explanation of the underlying assumptions made."? The reason we want them to show their work is to understand where they got stuck and how to remove the roadblock, provided the roadblock isn't just doing the math. For people who are really struggling, they may not be able to identify the concepts involved, but they can state the problem and what assumptions they used and 9/10 times those assumptions are lacking in some way and that's the cause of the problem. – tpg2114 Jan 13 '16 at 4:56
• In a lot of ways, asking for the assumptions is like asking to see their free-body diagram. The problem will almost always be immediately glaring to us but be completely invisible to the inexperienced questioner. – tpg2114 Jan 13 '16 at 4:58
• @tpg2114: Yes, the "show work" bit is intended to encourage posters to write more than just the problem statement and "I got stuck." But I think the issue is people are hung up on the phrase itself as allowing "check my work" type problems when the HW policy summary says otherwise. – Kyle Kanos Jan 13 '16 at 11:03
• Also curious that this has attracted negative attention considering the positive attention the idea has attracted previously. – Kyle Kanos Jan 13 '16 at 11:11
• I like your point #3. How about something like "We encourage you to make a genuine attempt at solving the problem, and to document the approaches you have taken in the question, and where you (think you) got stuck. This should involve an explanation of the physics you know, what you believe applies to the situation and why, and where possible, your attempt at working through the calculation." A bit wordy... – Floris Jan 21 '16 at 0:36
• I disagree. OPs understand what "showing effort" means. The confusing part is "asking about a specific physics concept." To a newbie, asking a thematic question (eg projectile motion, electrostatic fields/potentials, gyroscopic motion) is "asking about a specific physics concept." When their question is closed they protest "But I made the effort!" What needs explanation is the 1st part - ie why making an effort is not enough. – sammy gerbil Aug 10 '16 at 15:58
• @sammy: your next to last sentence is exactly why I wrote the above. It is confusing/contradictory because the close banner suggested (effort shown) == (remain open), but that isn't the case, necessarily. I agree that 'specific concept' is also vague, but that part is completely irrelevant to the point I'm making here. – Kyle Kanos Aug 10 '16 at 16:08
• Yes, I'm sorry, I have misunderstood your point. I think we are saying the same thing. – sammy gerbil Aug 10 '16 at 16:31
1. ...the criteria of "ask a conceptual question" (more or less) and "show effort" would apply to all questions, not just those we consider to be of an educational nature

Yes. That. As I've argued elsewhere, there is absolutely no reason at all that the criteria for judging a post should depend on whether or not that post comes from a homework assignment or from an educational institution.

In fact, that distinction is absurd; should questions arising from original research done in graduate school be subjected to different criteria than questions arising from my job in an industrial lab?

1. All questions which come from homework assignments are off topic (difficult to enforce)

No. That's not only hard to enforce, it's pants-on-our-head stupid. Whether or not a physics question comes from homework is entirely irrelevant to whether or not that question adds value to the site.

A large fraction of posts coming from homework are bad, but it's not because they're from homework, it's because the posts don't ask anything conceptual or show any effort. To illustrate by a converse example, suppose I were to post this:

What is the transition rate of a two level quantum system coupled by strength g to a thermal environment at temperature T?

That is a question straight out of my real research job; it's not homework, but it's also not a good post. I should show some work and identify the conceptual (or even technical) step where I'm stuck, and then focus my question on that.

Formulating a policy is only half of the problem. The more difficult part is enforcing it.

At present anyone can ask or answer a question. Few users study the site policy before doing either. The result is that while some are busy reviewing new questions to see if they comply with policy, others are busy answering them regardless of policy. By the time a VTC decision has been reached, the question has often accumulated up-votes and answers which themselves are up-voted. It is rather pointless to close the question in this situation. The only thing we have done is prevent anyone posting a better answer.

The situation is just as bad for duplicates. Sometimes the new question receives more answers and up-votes than the original, again making it rather pointless to close the new question. If a duplicate is closed with answers on it, we still have answers to the same question in two (or more) places. Which defeats the purpose of this policy.

I see only two consistent options : either (1) abandon any attempt to control the posting or answering of questions, or (2) place questions 'on hold' as soon as they are posted, and do not open them until they have been approved by reviewers who understand and agree with site policy, rather than those who have their own interpretations of it.

• "At present anyone can ask or answer a question." is by design—it even says as much if you visit the site without signing in—so there is no point in carping about it, the things that might be changed all have to do with policy, achieving rough consensus on it, communication thereof, and how it is enforced. – dmckee --- ex-moderator kitten Aug 11 '16 at 17:51
• @dmckee (1) is a change of policy. (2) requires a new feature. Both are radical suggestions which attempt to address the difficulties of enforcing policy, rather than tinkering with it. What prevents either change from being made? I appreciate that neither option is popular. – sammy gerbil Aug 11 '16 at 19:58
• Popularity is the only thing that prevents option (1). I'll give dollar to donuts that Stack Exchange won't even think twice about option (2): it's been suggested and turned down on the mother meta many times. – dmckee --- ex-moderator kitten Aug 11 '16 at 20:03
• @dmckee : I looked on Meta SE but didn't find any mention of (2). – sammy gerbil Aug 11 '16 at 20:33
• The language is different, but meta.stackexchange.com/questions/185530/… and on meta.StackOverflow something similar: meta.stackoverflow.com/questions/252061/…. – dmckee --- ex-moderator kitten Aug 11 '16 at 20:37