Many questions on this site are homework questions. Some students would like to hide the fact that they used this site for solving their homework, and so after getting the answer, they immediately delete the question. They can do so as long as the answer has not been upvoted, without the answerer even being notified.

I see at least two possible ways to fix this:

  • Prevent low-reputation users from deleting answered posts, even if no answer has been upvoted.
  • Notify answerers whose answer has been deleted in these circumstances.

As long as none of these are implemented, the site will remain an ideal place for cheating.

This is not the first time this topic is being raised: I asked this last year, and Nathaniel last month.

Everybody agrees that this is wrong, but nobody seems interested in changing the behavior of the system.

  • $\begingroup$ I obviously agree with both those ways. I thought something like "cannot delete before 24h after an answer has been posted", but preventing deletion with a reputation threshold works too. $\endgroup$ – Nathaniel May 17 at 18:48
  • 2
    $\begingroup$ This was also recently raised at meta.mathoverflow.net/questions/5044/… . Perhaps there is a chance to change the behaviour if more sites ask for it. $\endgroup$ – Emil Jeřábek Jun 1 at 6:54

I think the title is misleading. SE doesn't promote cheating; true is that it doesn't (and can't) prevent it. There is no tech solution to this human problem.

Here are two simple ways everybody can take to make the site less attractive for cheaters, neither of them new.

  1. Don't answer obvious homework dumps.

    I've actually been saying that since ... forever. Oh well.

  2. Upvote answers; users can't delete their own questions if answrs have upvotes.

Note also our homework policy. It's old, but IIRC still active. We've also collected lots of advice on how to write good homework questions. Feel free to liberally comment with links to both, in the hope of nudging users to asking real, non-cheating question.

And, finally, if you notice bad-faith self-deletions, flag for moderator attention. We can and will restore such questions, and take additional action should the user turn out to be obstinate.

  • $\begingroup$ I just found your(?) post, thanks for the good work, I think I will try to use it often. I am just afraid that its length may repulse homework posters/cheaters (I mean, they are trying to get to the answer without effort…) $\endgroup$ – Nathaniel May 17 at 21:36
  • 1
    $\begingroup$ @Nathaniel Thanks, I hope it'll be useful! We also came up with a reference comment you can just copy-paste underneath those questions that are not worth any more effort than that. $\endgroup$ – Raphael Mod May 17 at 22:36
  • 2
    $\begingroup$ @Nathaniel In my book, repulsing help vampires is a good thing. :) $\endgroup$ – Raphael Mod May 17 at 22:36

I suspect it's very unlikely that Stack Exchange is going to change the behavior of the software for just this site. Realistically, the revenue from our site is nowhere near large enough to make it a development priority. My impression is the general trend has been that other projects (e.g., Teams, Careers, etc.) have been a higher priority for software development resources than Q&A sites; and among the Q&A sites, Stack Overflow, the premier question-and-answer Stack Exchange site, has by far a higher priority for software development resources than us smaller SE sites. And if we wait for SE to make some change to their software to address this problem, we'll be waiting forever.

So I recommend that we focus on what we can change about our behavior and our policies. Some of the plausible options: As a community we could change our own behavior (e.g., don't answer homework questions, or vote to close them, or something). Or, we could change our policy (e.g., a stricter policy on such questions, enforced by the community or by flags to the moderators, or a stricter policy on answering such questions, or I don't know what). In any case, the levers that we can change seem like our own behavior and our own policies.

My sense is that as long as such questions get answers, people will continue to post them, and we'll continue to see cheating. Personally, I don't see keeping the questions undeleted as a terribly effective disincentive to cheating; my guess is we'll continue to see cheating even if the questions remain undeleted, as long as they get answered. That's just my sense; but I think what matters more is what the community's overall perspective and position on these matters is.

I think Raphael has been commenting on this pretty consistently over time.

  • $\begingroup$ Oh, thanks for that pointer to the stricter Mathematics policy! That's gold to pull that particular tooth of "but math.SE is much more friendly for my lazy posts"! $\endgroup$ – Raphael Mod May 18 at 21:32

Your post is pretty weird. The timeline is the following:

  1. Someone posted HW problem without any effort.
  2. Someone answered the problem (let's be honest, it happens a lot).
  3. (optional) The question was deleted.

You somehow concentrate on item 3, which I find a bit ridiculous. Cheating happens at item 1. And cheating succeeds at item 2. Item 3 is completely optional.

There is no feasible way to prevent item 1. But it's pretty easy to prevent item 2 by downvoting and closing the post (as I understand, the post is closed with 5 close votes and is hidden when it has a score of -4) (BTW, I believe that "no effort" is a legitimate closure reason). People don't know that they need to show any effort? First, it should be made clear (people must see this info before they post a question). Second, the post can be reopened after the edit (also, downvoters may follow the post to see if any changes happened).

And just a bit of personal reaction: I won't contribute to this site (or any other site with the same problem) anymore as long as cheating remains at this level. And I don't see it changing: the only thing that people answering HW problems observe is their rising reputation. I guess another approach is to punish such answers (and I'm perfectly OK with this; I have no idea why the common policy is "don't downvote good answers for cheating questions", since such answers directly participate in academic misconduct), but there are more mild solutions outlined above. I did downvoting/closing as described above, but I was clearly a minority.

  • $\begingroup$ If we stop responding to homework questions, we can close this site. The vast majority of posts are homework questions. $\endgroup$ – Yuval Filmus May 19 at 3:48
  • $\begingroup$ Well, then you are welcome to continue answering them. I'm out. $\endgroup$ – user136749 May 19 at 5:28
  • 1
    $\begingroup$ @YuvalFilmus Alternative hypothesis: if we shut down homework questions, interesting questions get more attention, and an active community of experts can form. $\endgroup$ – Raphael Mod May 19 at 19:53
  • $\begingroup$ "I have no idea why the common policy is "don't downvote good answers for cheating questions" -- our homework policy rests on the strong belief that we can't tell (or even prove) what's cheating and what's not. $\endgroup$ – Raphael Mod May 19 at 19:54
  • 1
    $\begingroup$ There already exists such a community — Theoretical Computer Science. The number of genuinely interesting questions here is quite small. The vast majority is homework questions. $\endgroup$ – Yuval Filmus May 19 at 19:56
  • $\begingroup$ @Raphael, I agree that's it's not always clear, but there are enough questions of the form: "this is the CS101-level problem, [I don't know how to approach it], please solve it". Even when it's not cheating (in the strict sense that it's not a HW), I believe that answering such questions must be looked down upon. In this sense, I like that CS.SE has a list reference questions,. $\endgroup$ – user136749 May 19 at 20:04
  • 2
    $\begingroup$ @YuvalFilmus, as I understand, that site is mostly for research-level questions. I believe there is quite a gap between HWs and research-level questions. $\endgroup$ – user136749 May 19 at 20:06
  • 2
    $\begingroup$ Also, there's a lot of interesting CS that's not TCS. $\endgroup$ – Raphael Mod May 19 at 21:13

You must log in to answer this question.

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