37
$\begingroup$

Suppose someone has asked twenty questions, but has accepted no answer on any of them. Is there not a strong case for preventing this person from asking any more questions? Of course, it may be possible that no answer to any of the questions is 'good enough' to accept. This problem could be overcome by giving people the option (say, on questions over a week old) to select an option saying: 'none of the answers provided suffice'. If they do so on each question (say, over a week old) with no accepted answer, they could be permitted to ask further questions.

What are your thoughts on this? It would add an extra degree of complication to the system, but it would prevent the significant problem of people asking questions, getting the answers they want, and then not fulfilling their responsibility of accepting an answer.

$\endgroup$
  • 28
    $\begingroup$ Some people accept too quickly, which can be equally annoying. I don't know which end of the spectrum I find less acceptable. $\endgroup$ – robjohn Apr 2 '14 at 16:56
  • 1
    $\begingroup$ A valid point. I guess this comes down to what the aim of math.stackexchange is. I always assumed it was so people could get answers to questions. So if someone thinks their question has been answered, then it's up to them to accept the answer. $\endgroup$ – Frank Apr 2 '14 at 19:40
  • 11
    $\begingroup$ @Frank That's basically right, but people does not mean only the person who asked the question. The idea of SE network is that questions and answers can be collected, vetted and organized for use by other people, far beyond the site participants. From this point of view, it is important for the top-presented answer to be a good one. An answer voluntarily chosen by question asker is likely to be good. An answer chosen by a user who does not care, but was forced to pick some answer... not as likely. In such cases, it's better to let the community decide what goes on top, via voting. $\endgroup$ – user127096 Apr 3 '14 at 0:34
  • 1
    $\begingroup$ There is a comment template that can be used to make user aware that they can accept answer. $\endgroup$ – Martin Sleziak Apr 3 '14 at 15:09
  • 4
    $\begingroup$ I think in general the number of upvotes is a better (although also not infallible) indicator of answer quality than which answer is selected. Simply because more people are involved, and because (hopefully) only people who can judge the answer vote on it, while the person who asked may well select a misleading or even wrong answer, thinking it is helpful. Also, even if at the time of selection, it is really the best answer, it is not that uncommon that batter answer follow later, which likely gather upvotes, but unlikely gather a moved accept. $\endgroup$ – celtschk Apr 3 '14 at 21:05
  • 4
    $\begingroup$ I agree with this. Perhaps the 'accept' function should be done away with altogether...? Instead, after a period of a week (say), the answer with the most upvotes could be automatically accepted by the system as the 'accepted answer'. This would make the process more democratic, and it would not then be prone to the faults you identify. $\endgroup$ – Frank Apr 3 '14 at 23:14
  • $\begingroup$ BTW doesn't SE software reminds new users about accepting answers? I mean something like this. And I also think I recall seeing some kind of reminder when I was viewing list of my questions on my profile page. $\endgroup$ – Martin Sleziak Apr 7 '14 at 16:58
  • 2
    $\begingroup$ The new site PhysicsOverflow has adopted a different point of view: The idea of accepting answers was completely unnecessary as the decision gives the false impression that the accepted answer is really the best. $\endgroup$ – Américo Tavares Apr 16 '14 at 16:36
  • $\begingroup$ They are physicists and maybe can't perceive well a mathematical possibility like this: there are questions having only one (good and correct) answer. $\endgroup$ – user26857 Jun 25 '14 at 7:31
  • 4
    $\begingroup$ Let's change "Suppose someone has asked twenty questions, but has accepted no answer on any of them" by someone has asked 58 questions, but has accepted no answer on any of them. $\endgroup$ – user26857 Jun 25 '14 at 7:34
39
$\begingroup$

Various proposals around acceptances are floated on various metas all the time: allowing moderators to mark as accepted, allowing other users to mark as accepted, and now more or less forcing the users to accept answers. I honestly do not understand the obsession. So, there is no green checkmark next to any answer. How is that a significant problem?

The network's founder was not obsessing over green checkmarks himself. He wrote

This is a simple social convention we use to close the loop between the person asking and the person answering... Accepting an answer is not meant to be a definitive and final statement indicating that the question has now been answered perfectly. Heck, we don’t even expect people asking questions to come back and accept an answer most of the time.

Just between you and me, not all question askers are equipped to recognize the best answer to their question anyway — that’s where community voting comes in.

Later in the same post he warns against

contorting a simple social convention into far more than it was meant to be. Expecting every asker to come back and mark an accepted answer is totally unrealistic.

$\endgroup$
  • 11
    $\begingroup$ OK, that is interesting. I think the obsession is not the result of OCD as suggested by the founder and you. I think the good reason for trying to increase acceptance rates is so that it's easier for people trying to help others to see which questions require answers and which don't. Which is made easier by questions being accepted. For instance, in the 'Unaswered' section, there are lots of questions with upvoted answers, but no accepted answer. So in every meaningful sense, the questions are answered, but not recognised by the system as being such. $\endgroup$ – Frank Apr 2 '14 at 16:04
  • 4
    $\begingroup$ @Frank A question with an upvoted answer does not appear under Unanswered; this is the subject of the blog post to which I linked. $\endgroup$ – user127096 Apr 2 '14 at 16:31
  • $\begingroup$ OK. So I did an experiment; I refreshed the list of unanswered questions, picked one, and it had an accepted answer. Unfortunately ideology cannot combat empirical evidence. $\endgroup$ – Frank Apr 2 '14 at 16:39
  • $\begingroup$ @cheapeffectivedietpills: I think you might have meant "positively scored" as opposed to "upvoted" in your previous comment. $\endgroup$ – user642796 Apr 2 '14 at 16:59
  • 4
    $\begingroup$ @Frank: I am sure you are aware that just saying "I've done something and something happened" doesn't really constitute providing any empirical evidence. (And I have no idea what "ideology" you could possibly be referring to.) We have no idea what might have caused this event you claim. It is even possible that an answer was accepted in the interim between refreshing the page and clicking a question. But we'll never know. $\endgroup$ – user642796 Apr 2 '14 at 17:07
  • 4
    $\begingroup$ 'I've done something and something happened' doesn't constitute empirical evidence? :S What's just happened is this: You claim it isn't raining outside; I open the curtains and see it is raining outside; you say 'So you opened the curtains and it's raining outside... So what?'. $\endgroup$ – Frank Apr 2 '14 at 17:26
  • 1
    $\begingroup$ And the ideology I refer to is accurately described as the 'if it ain't broke don't fix it' ideology - that is to say, if something works fine, but could be improved, let's not bother to improve it. $\endgroup$ – Frank Apr 2 '14 at 17:29
  • 1
    $\begingroup$ @Frank At present, your report of seeing an accepted answer in the "unanswered category" is status-norepro. As for "don't fix it", it's more like "weigh the cost of complexity when considering new features". You thought about a button saying "none of the answers provided suffice". What signal does it send, how is it recorded and posted? What to do when someone gives a good answer after that button was clicked - will there be an undo for that button? But what if the asker does not bother to click that undo, as they did not bother with accept? Will there be another check for that? Etc. $\endgroup$ – user127096 Apr 2 '14 at 19:20
  • $\begingroup$ @Frank, as a side note on that logic, «I open the curtains and see it is raining outside; you say 'So you opened the curtains and it's raining outside... So what?'». It is not because you opened the curtains that it is raining outside. Empirical evidence usually means there is enough correlation to conclude causation, which is not the case. $\endgroup$ – JMCF125 Apr 2 '14 at 19:32
  • $\begingroup$ @cheapeffectivedietpills OK, I understand and appreciate this argument. Obviously, it was just a suggestion. But in fact that is exactly how I would envisage the system operating. Of course there would be extra considerations to be taken into account (as you have pointed out) and I agree that it wouldn't necessarily be easy. However, I do think there should be some extra incentive to accept answers, or some extra consequence of not doing so. $\endgroup$ – Frank Apr 2 '14 at 19:32
  • $\begingroup$ @JMCF125 I think the fact that I witness rain counts as empirical evidence of the hypothesis 'it is raining'. Unless you discount visual data as empirical evidence, and would expect me to go outside and stand in the rain with a rain catcher and take measurements. $\endgroup$ – Frank Apr 2 '14 at 19:34
  • 3
    $\begingroup$ @Frank: I believe there is a 2 point incentive for Users to accept Answers. $\endgroup$ – hardmath Apr 3 '14 at 0:27
  • 2
    $\begingroup$ I don't have much of an opinion on this proposal of Frank's, but I just wanted to mention one benefit of accepting answers. Suppose a question has 3 or 4 "answers", all quite long and complex, only one of which is correct. The investment of effort required to determine the correct answer (and that one is correct) could be considerable, so it is useful if someone has already done that for you and reported their findings with an accept. Often, the person best suited is the OP. Not always, and not all questions are of this type, but it has helped me in the past. Cheers! $\endgroup$ – user139388 Apr 4 '14 at 0:11
  • 2
    $\begingroup$ @Frank: Just to be clear, any question which has an answer with a positive score does not appear under "Unanswered". If it does, then it's a bug, please report it. $\endgroup$ – ShreevatsaR Apr 5 '14 at 3:58
  • 1
    $\begingroup$ @Frank: Well, in that hypothetical situation, if answers were to appear as unanswered despite being accepted, then it would remove one reason for accepting answers. :-) Anyway, again, you should not see any questions with a positive-score answer in the Unanswered section. If that happened it's a bug; please report it. $\endgroup$ – ShreevatsaR Apr 6 '14 at 1:48
10
$\begingroup$

People accept answers as a kind of "thank you." Its a way of giving 15 points to show your appreciation.

The problem is, until you have an accepted answer or you've read the FAQ, you don't know there's any points involved (this was me for about 8 months). So this whole nonsense with "accepted answers" just ends up seeming kind of pointless. You end up thinking: "Why should I bother fretting over which answer to accept? Its irrelevant."

So I guess one possible solution would be: when you roll your mouse over the tick/checkmark, it could say "click to accept this answer, conferring 15 points" or some such.

$\endgroup$
  • 16
    $\begingroup$ I think many new users don't even notice the check mark. $\endgroup$ – MJD Apr 3 '14 at 15:04
  • 5
    $\begingroup$ At least to me, the more important information would be “you can change your mind and accept a difference answer later on”. Otherwise, I'd be afraid that accepting an answer would prevent me from eventually accepting a better answer. $\endgroup$ – MvG Apr 14 '14 at 12:45
7
$\begingroup$

Suppose someone has asked twenty questions, but has accepted no answer on any of them. Is there not a strong case for preventing this person from asking any more questions?... What are your thoughts on this?

To first order, I think questions should be considered on their merit, to "play the ball not the man". If it's a good question, upvote it, and if you can, answer it.

Having said that, one of the main reasons I am engaged on SE sites is because they are interactive: I can get feedback on what I've done well or badly, and I can likewise give feedback by voting, adding comments, etc. So I find it quite unsatisfying if I answer a question and receive no feedback whatsoever. And I'm not ashamed to say that I value accepts or "thanks" -- in my professional life I get a great deal of satisfaction from satisfying customers and more generally helping people learn, and it's not much different online.

Moreover, this is a community and if someone deliberately violates community norms, whether it's being rude to other users, posting spam, or knowingly refusing to engage in voting, accepting answers or the common politeness of saying "thanks for the effort", then I'm likely to put in less effort to help them.

I'll quite happily prompt newer users to accept or vote if they haven't been, with text like the following

hi @user123456, as a relatively new user you might not realize that you can comment, upvote, or even accept answers to your questions. You've asked 4 questions and got 5 answers and the site works better when there's feedback from the person who asks the question. Please see meta.stackoverflow.com/a/5235/165400 for more information. (When you have more reputation you will be able to downvote as well.)

I've done this a few times on mine and other people's answers. And I've been pleasantly surprised -- the first time the user concerned replied to say they didn't realize, went back over all their questions and upvoted and accepted answers to most of them.

But your question is really about the longer-term user who refuses to engage in accepting. I have no problem about leaving a comment for such users if I come across them, along the lines of

hi @user7890ab, did you realize that you can comment, upvote (when you have 15 reputation), or even accept answers to your questions? You've asked 10 questions and got 11 answers and the site works better when there's feedback from the person who asks the question. Please see meta.stackoverflow.com/a/5235/165400 for more information. (When you have more reputation you will be able to downvote as well.)

However so far it hasn't had any affect when I've tried it! I think it's reasonable to give such feedback, however, as eventually it might encourage such users to engage. For this class of user, I would certainly tone down the amount of effort I put into answering their questions, all other things being equal. If you want to see users' accept rates, the User Accept Stack App will provide the statistic for you.

So, to summarize, do I think any additional mechanisms are necessary? No, I don't, but I think it is reasonable to give feedback within the mechanisms that already exist on the site. And, perhaps most importantly, everyone on the site needs to figure out how much effort they want to spend answering questions, in view of their own personal cost / benefit system, and then apply that as best they can.

$\endgroup$
  • 3
    $\begingroup$ Maybe it should be pointed here, only registered users can vote (or flag, for that matter). So telling unregistered users that they should vote is not going to work without them registering first. And not everyone wants to register. $\endgroup$ – Asaf Karagila Apr 7 '14 at 14:55

You must log in to answer this question.

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