Following an audit test failure where I agreed that a question should be closed as too subjective, and was told that there were no major problems with it and I should have left it open, I have two closely related questions:

  1. What is the basis for the statement that there are no major problems with the question? Does the audit system only select questions for the "Close" queue that have been nominated for closure and kept open? Or does it assume that any question which has been open for a few days and attracted some up-votes must be ok?

  2. What is the correct response to the situation above? (This depends partly on the answer to the first question: if the question has already been kept open then nominating it for closure isn't a suitable reaction, although opening a meta thread to discuss where the line is drawn might be).

  • 10
    $\begingroup$ Wow. This whole "test" business strikes me as quite bizarre, or, at least, quite poorly implemented. $\endgroup$ Sep 9 '13 at 10:43
  • 1
    $\begingroup$ The question that was used in your audit has 8 upvotes and has not been closed in the more than two weeks it existed. That should be a solid indication that the question is usable as an example for an audit. The actual problem here in my opinion is that there is no consensus on where to draw the line for such subjective recommendation questions. $\endgroup$
    – user9733
    Sep 9 '13 at 10:56
  • $\begingroup$ @MadScientist, there's a hint from the current comments that some other comments have been deleted, and that they may have been debating its subjectivity. $\endgroup$ Sep 9 '13 at 11:12
  • 1
    $\begingroup$ @PeterTaylor: I have inspected the question under argument, and I can say that there have been no comments deleted from the question itself (and a single "greetings" comment deleted from one answer). One comment was edited twice (as I believe you can see), but even I don't have access to its previous versions. $\endgroup$
    – user642796
    Sep 9 '13 at 11:37
  • 1
    $\begingroup$ What question is it? $\endgroup$ Sep 9 '13 at 12:45
  • $\begingroup$ @MadScientist "The question that was used in your audit"... How do you know the question? $\endgroup$
    – Did
    Sep 9 '13 at 13:02
  • 2
    $\begingroup$ @Did I checked the recent activity history of the asker and found this failed review: math.stackexchange.com/review/close/95025 $\endgroup$
    – user9733
    Sep 9 '13 at 13:03
  • $\begingroup$ @MadScientist Thanks for the information. $\endgroup$
    – Did
    Sep 9 '13 at 13:07
  • $\begingroup$ @CarlMummert, although there's enough information in this comment thread to identify the question, I'm not going to edit it into the question because I've tried to make my question stand independently of the specific test instance. My goal is to clarify a broad principle, and if the community position seems to require a follow-up question about the specific instance then I would make that a separate follow-up question. $\endgroup$ Sep 9 '13 at 13:19
  • $\begingroup$ @Peter Taylor: as a broad principle, you can vote to close any question you like. In this case I agree with you that the question was written in a way to encourage subjective answers, so I would also vote to close. $\endgroup$ Sep 9 '13 at 13:23
  • 2
    $\begingroup$ I must have hit about two dozen audits so far, and always passed. They seem to have been pretty straightforward so far, but I would also be upset if I got caught by one I felt was ambiguous. At any rate, everyone is probably going to be caught a couple times depending on how often they review. The thing that matters is your frequency of failed reviews, right? $\endgroup$
    – rschwieb
    Sep 9 '13 at 16:33
  • $\begingroup$ @rschwieb: I managed to "fail" one for math.stackexchange.com/questions/466281/…. I played around with the review queue after the comment below. The system lied and said that the question was closed, when it really was not, and asked me if I wanted to reverse the decision, only to complain when I clicked a button to indicate my opinion that the supposed status quo was acceptable. $\endgroup$ Sep 9 '13 at 17:05
  • $\begingroup$ @rschwieb, I'm not actually upset. All systems have flaws, and there are grey areas where, as Mad Scientist comments above, there isn't a clear consensus. What I'm trying to clarify is whether when I hit an audit which I disagree with it means that the particular instance is almost certainly in a grey area and in need of discussion on meta - a running close/reopen battle isn't productive - or whether I should just treat it as any other instance of something bringing to my attention a question which I would nominate to close. $\endgroup$ Sep 9 '13 at 17:13
  • $\begingroup$ @PeterTaylor Sorry, I didn't choose any of my words very well. The audit in question was certainly not as obvious as the ones I came across, so I agree with you. $\endgroup$
    – rschwieb
    Sep 9 '13 at 17:27

While I don't have an authoritative answer, on the basis of further experience it seems to me that questions are selected on the basis of survival and possibly having a couple of upvotes. The aspect of the audit test failure which caused me uncertainty and led to this question was that I agreed with the closure reason mentioned at the top; while I've seen a few audit tests since where I agreed that the question should be closed, it hasn't been for the reason mentioned. I therefore now suspect that the reason is selected randomly rather than on the basis of someone else having selected it.

Given this, it seems to me that where the audit draws my attention to a question which should be closed the correct thing is to vote to close it.


This is at least the second issue with the audit system in the last few days. Please upvote this answer if you would like to see it turned off on this site. This will help gauge the overall opinion about the audit system.

  • $\begingroup$ The other issue I am referring to is meta.math.stackexchange.com/questions/10836/… $\endgroup$ Sep 9 '13 at 12:44
  • 5
    $\begingroup$ (-1) I think it's too early to be judging the system's effect in the MSE case. I would like to give it the benefit of the doubt for now. $\endgroup$
    – Lord_Farin
    Sep 9 '13 at 12:58
  • $\begingroup$ (-1) I think this suggestion is too broad: from what I've seen, only the close audits seem to be a problem. $\endgroup$
    – user14972
    Sep 25 '13 at 2:56
  • $\begingroup$ Also, I think it is premature (although that may be a symptom of the extent I've paid attention to the issue rather than the actual truth) -- a tiny self-selected sample does not often paint an accurate picture. $\endgroup$
    – user14972
    Sep 25 '13 at 2:58

You must log in to answer this question.

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