I have a lot of thoughts on this issue and I've been trying to figure out the best way to arrange and present them. Really, there are too many things for one answer, so I'm going to try for a more straightforward evaluation of your proposal that spends less time on philosophical objections.
You're right to be concerned about the guidance we provide new users. My gut reaction to your proposal was that it's not necessary or viable but there's no reason we should rely on instinct when there's data to be had. I went through every closed or deleted question that was asked between December 1, 2015 and January 31, 2016 and classified them by hand based on their content. Here are the highlights, excluding nuked spam posts.
Of the 128 closed questions (including deleted):
Of the 17 closed homework questions:
- 8 (47%) are no-effort homework dumps
- 9 (54%) only resemble homework (but may not be)
- 1 duplicate
- 2 off-topic
- 1 too broad
- 5 unclear
Now let's look at the proposal.
Arguments for a custom close reason
I've given the proposal a fair amount of thought, read it and re-read it since it was posted and I think it can be reduced to four fundamental arguments:
- Homework questions that don't show effort are a distinct class of questions that can be treated as homogeneous with respect to closure and guidance.
- These questions are a significant problem, representing "a large portion of the questions that are closed."
- Close-voters typically use the general off-topic reason to close these questions. (The implication being that this is also correct.)
- The guidance provided to new users when we close their questions in this way does a poor job of balancing our desire for the site to be high quality with our desire to welcome new users and help them recover from their mistakes.
The conclusion, if you accept these four arguments, is that we should introduce a special-case close reason for no-effort homework questions. (Keep in mind that these are effectively a limited resource, in that we get three "freebie" off-topic reasons and anything outside of that requires attention and effort from staff.)
I'll concede this one for the sake of brevity; we'll assume that all no-effort homework questions can be dealt with in the same way. As a general rule, this is probably close enough to the truth.
There were 623 non-closed, non-deleted questions asked in December and January. That makes 770 non-spam questions, of which only eight were obvious homework dumps. There's a bit of uncertainty in assessing the nine that I flagged as "possible homework" but it's reasonable to expect the real count of questions that fall into this category is somewhere in the 8–17 range.
I'm not sure if that's the magnitude that you felt you were dealing with or if it's less significant than you thought but it doesn't look like a major issue to me based on frequency, even at the upper bound of 17 questions. We have more than enough users with the experience and site privileges to handle one homework dump every few days, on average, by voting and commenting as normal.
3. The Close Reason
There's a pretty clear split here between off-topic and unclear for choice of close reason. That's not at all surprising, given the guidance in our previous policy discussion:
I'd suggest voting to close as "Unclear what you're asking." It's
probably the best close reason as it reflects back "We, the community,
don't understand what you don't understand about the problem you
and that, if you browse through some of the history of homework on the network, it's clear that many active users don't want their site to touch homework problems with a 10-foot pole.
It's noteworthy that those questions that are clearly homework are largely the ones we're closing as off-topic. I should point out that I was involved in the closure of only two of the homework questions in this sample: this one, which I closed as unclear despite neither of the two previous close-voters having chosen that option, and this (deleted) one where I was the sole close-voter (and went with too broad).
This seems like a good time to go over some examples:
- Input Force vs Output Displacement of a System is closed as off-topic. Definitely bad guidance on our part, but not for lack of a better close reason. "Could someone explain how to reason through this?" speaks directly to GlenH7's point that we don't understand what OP doesn't understand.
- Pivot Point Equations is closed as off-topic, even after OP commented that it wasn't homework. It's a simple matter of applied physics and geometry, and again we're doing a disservice by suggesting a question is "not about engineering." Even worse, a comment directs OP to The Engineering.SE policy on check-my-work questions, which doesn't seem to contain any useful guidance for the situation.
- Engineering Mechanic problem is deleted after being closed as off-topic. It's a classic homework dump, showing the problem that was assigned with absolutely no other information. Some of our users left comments like if you tell us what you don't understand the community will help, be more specific in what you do not understand and What is giving you trouble? That's good; it tells the author that what they need to do is clarify. It would have been better if users who voted to close had chosen a reason that matched that message.
I agree with Wasabi's answer when it says How do I ask a good question? is the most appropriate Help Center article for these users to read. Fortunately, the banner displayed on questions closed as unclear already points there.
4. Quality vs. Accessibility
This is a perennial topic of debate on larger network sites: How do we get rid of all the crap? versus Why are we so mean/elitist/exclusive?
Maybe it's a foregone conclusion at this point but I do agree that we're providing some pretty unhelpful guidance when we close homework questions as off-topic. It should also be clear by now that there's a much simpler solution than creating a custom reason: Stop closing them as off-topic. Use unclear instead, or too broad when appropriate (deleted).
That's not to say that this is always an obvious or easy conclusion to come to. Many network sites have discussed the same or similar proposals (e.g., Stack Overflow, Computer Science, Math, Physics—the rest are left as an exercise for the obsessive and/or self-loathing reader).
We don't want to be mean or elitist but we do want to be exclusive. We want to exclude people who don't and won't contribute to our overall mission, whether you call that mission "being a good Q&A site" or "building a searchable index of solutions" or "making the world a better place."
Are there users who come here for help and quickly leave in confusion? I hope so! Listen to Joel Spolsky talking about the culture of the network:
...everything about a community either draws people into that
community or pushes them away and so many people in web design have
been trying to figure out, how do we make a web page that sucks every
single person in the known universe in? And when you're trying to get
answers to difficult questions, that is the opposite problem. You
actually want to drive away as many morons as you possibly can,
hopefully as quickly as possible.
I'm not wild about throwing around "morons" like that but I guess when I found something that helps millions of people solve non-trivial problems I'll use the words I like. The takeaway for me is that we want to calibrate our community moderation features—meaning regular voting, closing, reviewing and editing—in such a way that people who are unwilling or unable to figure out how to succeed at asking a question just don't get to participate. I'm less concerned with who's "most in need of assistance" than I am with who can contribute in a positive way.
Big picture philosophy (abbrev.)
We try to judge content on its merits. When someone comes here to ask a question, the fact that they're working on a homework problem is really just context. It's not useful context, like whether your failed connection was welded or bolted; it's irrelevant context. Nobody needs to know whether a question was asked on a Thursday or by an Estonian or using the official Stack Exchange Mobile App for Windows Phone in order to solve the problem that question describes. That type of context shouldn't be a factor when casting close votes.* Breadth, clarity, topicality and subjectivity are the axes of interest and each of these is determined by content. The context of the problem can be valuable content but the context of the question is usually irrelevant.
Yes, it's possible to force homework onto the topicality axis by adding a custom off-topic close reason, but what does that tell the author about how to ask a good question? Isn't it more likely to send one of the following messages?
- We're here to solve problems but your problems aren't important.
- We're engineers and homework isn't really engineering.
- We're experts and you're too young/inexperienced/stupid to engage on our level.
We use close reasons to succinctly illustrate the major failing(s) of a post so that they can be fixed, but the messages above are really just different flavors of go away. When someone comes to us about a homework problem, and we reject them on the grounds that it's homework, they have no remedy; we're not only not telling them how to fix their question, we're telling them they can't fix it. And even new users realize that the difference between a school assignment and a work or hobby project is not particularly relevant to our model of providing useful solutions to reoccurring problems.
Targeting just a subset of homework doesn't fix this problem because it still frames the issue with the author's context in mind. Our policy is not to close things based on a lack of effort. This is generally in line with policies on other sites on the network.
What's worse is that it's not just the author of the bad question who receives this message; it's also other members of the community who happen to see homework questions being shown the door. Including members of the community who participate in commenting and voting but not in meta, who have a less nuanced view of policy (when they're aware of them at all) and will take comments like "homework is out of scope unless you show effort" and rephrase them as worse comments like "homework is off-topic" and "we don't do homework" and eventually just "I'm voting to close this question as off-topic because homework without effort."
That's what I want to avoid, and why I think close reasons tailored specifically for homework dumps do more harm than good.
Epilogue: But we don't like them!
The best catharsis when you don't like a question is the almighty downvote. See also:
Your votes are yours to use how you want, provided each vote is based on the content of the associated item and not on, say, the identity of its author. If you read the content of a question and think, "Ugh, what a lousy homework dump," that's a perfect reason to downvote.
You take it, thermo! Take it and like it!
You can always change your vote later if the question is edited to where you don't think it deserves the downvote. If you don't care to keep tabs on it until then and are worried about leaving a downvote on a question that becomes good, that's fine; don't downvote it, and give up your catharsis. Your choice.
- Moderate questions based on their content, not their circumstance. Questions should never be targeted for closure based on extrinsic properties like how much effort the asker put in, whether they are a student or whether they are completing a homework assignment.
- Close banners will always be the Soup Nazi. Even worse: they're an implacable robot automaton Soup Nazi. No amount of massaging or finesse is going to turn our close banners into the love child of Mr. Feeny and Salman Khan** and that shouldn't even be our goal, because...
- It's a feature, not a bug! Users abandoning the site who contribute low-quality homework dumps and cannot figure out what makes a good question is entirely desirable.
- Still, we need to lead by example. We're building a community and our comments on active questions are a strong signal for that process. If you think a struggling user deserves a better outcome, the best way to demonstrate that is not with a form letter, but by engaging them personally and directly. Start with comments, and if it gets too long, take it to chat and/or Meta as appropriate.
* I can only think of one strong counter-example on the network—and that was implemented only after "literally years of the community hating and hammering these questions with close votes and downvotes."
** Some early drafts of this answer featured Bob Ross or Robin Williams. Feel free to substitute warm and fuzzy educators and/or pop culture references of your choosing.