7
$\begingroup$

Anecdotally, I get the impression that more questions are closed at night then in the morning (my timezone is UTC -5). Are there times of day where questions of the same quality remain open longer? Is there a time of day where low-quality questions (which sometimes do reveal intriguing ways of thinking about chemistry) have a higher chance of getting an answer? Once the question is closed, the only way the OP will get feedback is through comments, if folks even spend the time to look at closed questions.

$\endgroup$
  • 6
    $\begingroup$ Not a terrible topic to bring up, but could you state some of the implications you think this timing might have/what you want for input from us? There isn't much to answer. $\endgroup$ – A.K. Feb 24 '19 at 19:03
  • 4
    $\begingroup$ Since closing is (mostly) done by the community, questions get closed whenever people are active. Anecdotally, most people who close are in Europe/US, so it makes sense that it happens at night. I don't see a problem, though. If a question should be closed, then there's nothing wrong with closing it, and if they need feedback on the question itself then the comment section is the correct place for it, as the answer section is for answers. If questions are being closed unfairly, then that's another issue, but it doesn't seem to be what you're concerned about. $\endgroup$ – orthocresol Feb 25 '19 at 11:47
  • $\begingroup$ I agree that there is no problem. It is possible that new users might get a different impression of the site depending on when they use it. Sometimes, they will see that most of the recent questions are closed, and sometimes they are not. $\endgroup$ – Karsten Theis Feb 25 '19 at 17:33
7
$\begingroup$

I ran this script to return the number of questions closed per hour. I'm in Greenwich Mean Time so I assume these results are too.

SELECT
  DATEPART(hh, ClosedDate),
  COUNT(*) AS ClosedQuestions
FROM PostsWithDeleted
WHERE PostTypeID=1 AND ClosedDate IS NOT NULL
GROUP BY DATEPART(hh, ClosedDate)
ORDER BY DATEPART(hh, ClosedDate)

Closed Questions GMT

It seems to correlate pretty well with the number of questions ask per hour.

SELECT
  DATEPART(hh, CreationDate),
  COUNT(*) AS QuestionAsked
FROM PostsWithDeleted
WHERE PostTypeID=1
GROUP BY DATEPART(hh, CreationDate)
ORDER BY DATEPART(hh, CreationDate)

Questions Asked GMT

$\endgroup$
  • 1
    $\begingroup$ That's interesting. So it seems that the community is always active (the more question asked the more closed) or we have a shift of one day.. would it be possible to map the query to see which countries ask more questions and which one close more questions? $\endgroup$ – Jojostack Mar 4 '19 at 9:43
  • 1
    $\begingroup$ I don’t think so. Users input their own location so any queries on location return things like “221B Baker St.” and “Gryffindor House”! $\endgroup$ – Jacob Mar 4 '19 at 10:24
  • $\begingroup$ Hmmm, the absence of a lag between asking/closing and the timing around European dinner/American lunch time suggests either "question fatigue" (overburdened reviewers) or "hunger/post-prandial-induced poor question intolerance". That spatial data for poster/reviewer location would be nice to have (if only the continent). $\endgroup$ – Buck Thorn Mar 7 '19 at 11:29
  • 2
    $\begingroup$ I think most of the bad questions come from Slytherin.... $\endgroup$ – Buck Thorn Mar 7 '19 at 11:30
  • 2
    $\begingroup$ I've created this graph for India, USA and UK: oi64.tinypic.com/2ce15xs.jpg. It's based on manually chosen search terms so it's probably not very representative. I used this very primitive code data.stackexchange.com/chemistry/query/1000827#resultSets. I'll update it when I get time. $\endgroup$ – Jacob Mar 7 '19 at 15:35

You must log in to answer this question.

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