5

I had a look at the question closure statistics (link for 2k+ users) and the following strikes me as potentially bad:

  • We close about 1/4th of our questions.
  • Of those, for roughly 40% it should be possible to edit them to make them fit (I'm looking at the close reasons "unclear what you're asking", "too broad", "missing context").
  • However, in these categories only few questions are edited:

    • Unclear what you're asking: 2/35 (5.71%)
    • Too broad: 3/22 (13.64%)
    • Missing context: 2/12 (16.67%)
  • Of those questions that get edited, almost none get reopened. In the last 90 days, this happened for two questions out of 16, one for the "this does not seem to be about philosophy" reason and one primarily opinion-based question.

  • In fact, the group of questions that gets edited the most are the ones that are closed because they push a personal philosophy (6/21, 28.57%). These questions are usually beyond help because the aim of the asker does not correspond with the aim of the site - yet somehow it seems we are not able to communicate that clearly and the user continues trying to argue through edits.

Perhaps we can try to identify the reasons for this and discuss what can be done to help (mainly new) users to understand (a) what is needed from them to get the question reopened, if possible, or (b) why their question will probably never work here (and perhaps guide them to other places where they might find help).

Or perhaps you think that there is no problem at all (because closing 25% is OK, or because most of the questions we close are beyond help, or ...) - we can discuss that too. Unfortunately, I don't have data from other sites (but I could get them if needed).

4

A few quick responses. (This is a great discussion to have and I hope others weigh in too!)

  1. Question-closure is important for indicating the on-topic line "negatively". It sends a signal to new users about what the community thinks is within its sphere -- but is also important even beyond the scope of just this stack, since it sends signals to users of other stacks who may be interested in directing people here/migrating questions (i.e., possibly it somewhat decreases the amount of not-helpful migrations if we are as clear as possible in terms of directives about what's appropriate here.)

  2. Many of the questions that are closed are structurally "beyond reasonable help". Now there is definitely a limit to how much is "okay" to change about a question (given you're not the author; and even if you are [the author], still: totally transforming a question is really not usually reasonable to ask/expect and can even be confusing, "stranding" any content logically attached to a previous formulation.) My take is generally that it's often better to push for new, better-framed questions that might be more inherently answerable.

  3. We could definitely do a better job describing how to ask! --Maybe some of this is down to improving/clarifying the language we use, refining the examples on the tour and maybe even improving tag wiki entries (e.g., really should not give license to any possible question about "reality" -- i.e., you still need to provide philosophical context and motivation.)

3

What can be done to help users improve their questions?

The site should insist on precision and brevity. The positive message could be that such qualities improve the site for everyone and advance the progress of philosophical inquiry.

There could be a set of Author's Guidelines with the section, "Why was my question closed?" followed by a list of common problems. Many questions are so general that the question becomes too ambiguous to answer. The common result is either no answers, or a collection of unrelated answers, all responding to a reasonable interpretation of the question. The text accompanying such questions, and even those more precise, tends to wander around an area of thought and run on for too long.

I would encourage the preparation of drafts offline, with a cut-and-paste into the stack exchange site.

As for the 25% closure rate, that sounds appropriate, given the frequent drafting problems. Closures tell posters that the site's editorial standards are taken seriously. The problem is that legitimate inquiry is often lost when the question is never reposted. Perhaps the closure could be accompanied by an upbeat statement that the site encourages questions and wants to see this particular question revised and reposted.

1

I agree with Joseph here, I suspect (and we are limited to guesswork in this) that the reason why the questions on hold are not edited is that the questioner's have no interest in asking a question that is within the scope of the site and they had simply mistaken what the site is about, such questions simply need cleaning out.

The problem is with Joseph's point 3.. As I have harped on about many times before, things like the strap-line, the FAQ, the highest voted questions/answers, the highest viewed questions/answers - all give the impression that the site is about (at least in part) exactly the sort of discussion these closed questions are initiating.

I will not re-write all my comments on previous meta posts completely, but in summary;

  • The strap-line says "Q&A for those interested in the study of the fundamental nature of knowledge, reality, and existence" - It does not mention canonical or academic study and yet lack of academic focus is a common close reason.
  • The top voted answer to the FAQ question "What's this site about?" still states "either be part of a particular, widely-studied school of philosophical thought (e.g. Epicureanism) or are such as to be focused on abstract reasoning and widely accepted methods of argumentation and be precise in that they rigorously adhere to the demands of the original question." In practice questions which invite answers based on abstract reasoning are getting closed (or downvoted) ostensibly because they do not refer to, or limit their scope to, a particular school of academic philosophy. (I think it's important to note here that most of the questions of this sort I've encountered I would still prefer to see closed, not because of their lack of reference, but because their "abstract reasoning" is so muddled as to be unintelligible).
  • A substantial proportion (I'm tempted to say the majority, but don't have the statistics to back this up) of the highest voted questions (and particularly answers) are entirely opinion-based pieces with little, or often no, reference to any academic philosophy at all. I have in my previous comments given dozens of such examples. Not only does this give the impression that this is exactly what we're looking for vote-wise, but it is re-inforced by the fact that moderators and high rep users entirely refrain from commenting to this effect, as they would on posts from newer users.

Basically, I think that the 25% have fairly widely misjudged the scope of the site and their questions are probably not remediable, but what we could do to help is provide some information out there as to what the scope of the site actually is - Change the strap-line, write an up-to-date answer to the 'scope of the site' type questions in FAQ, and stop up-voting, or at least start commenting on opinion-based answers, even when they come from high rep users.

  • I'm not sure if those help center texts have a big influence, it is my impression most people don't read them before asking. Nevertheless it's an important reference point and should reflect the current practices. Do you have any concrete suggestions for revisions? I think it's worthwhile to look into this. – user2953 Sep 12 '17 at 10:05
  • 1
    I agree that the FAQ is probably the least influential item on my list and I'm much more concerned about how to deal with opinion in answers, but with regards to revisions, I think the questions are already there ("What's this site about?", the many questions about subjectivity), what's lacking from most of the answers are examples. It's of little use having an ideal if it's not possible to achieve, we need to see a canon of 'ideal' questions and 'ideal' answers, that both reflects the site's objectives and represent a reasonable majority of the questions/answers we actually get. – Isaacson Sep 12 '17 at 10:49
  • I would suggest that the higher rep users might spend a little time looking over the FAQ type meta questions and adding some recent examples to the standard answers. My concern is that they might struggle to find any, but that's a rather different matter. – Isaacson Sep 12 '17 at 10:51

You must log in to answer this question.