2

Is there a way to pull the stats on the demographic of people who are answering questions?

I get the impression there are only a few of us and we close questions we don't know the answer to and write the same answers to reoccurring questions where we have strong opinions.

If this is the case, it might be sensible to change the rules or application of the rules to encourage more participation.

3
  • 2
    Are you saying that you, and maybe other people, are closing questions just because you don't know the answer? In other words, you know that's happening because you're doing it? Or, are you saying you suspect that's being done by unidentified other users?
    – Makyen
    Aug 10 at 15:25
  • Just to be clear, I'm trying to get at a distinction between you, or other users, just not knowing the answer and the question being poor enough such that it can't be answered in its current state and the answer to the question not existing or being unknowable. The way this is written, it sounds like you, and others, are voting to close questions just because you, personally, (or they, themselves) don't know the answer to an answerable question. If that's not what you intended, please edit the question to clarify.
    – Makyen
    Aug 10 at 16:24
  • my feeling is that with a small group of answers there is the risk of unconscious bias in question closing and downvotes.
    – Ewan
    Aug 10 at 22:11

1 Answer 1

Reset to default
4

There are no per-site dashboards or tools that expose this information.

This seems like something you could get from the Data Explorer. I haven't looked, but perhaps someone wrote a query that can be run against Software Engineering. Otherwise, one can be written. I haven't had the time to see what was there, but knowing what data is exposed there, that seems like the best bet to get the information. For someone with some time, this seems like an interesting exercise.

11
  • didn't know about this. quick query give 67 users with answers scoring > 0 since this time last month. so more than i thought.
    – Ewan
    Aug 10 at 14:10
  • @Ewan I'd have to think more about thresholds. 1 answer may be insufficient - it's not that hard to write one good answer. But is it 3 in a month? 5? Is it 1 up vote or 5? How old are the accounts - are there a lot of drivebys or lots of recurring?
    – Thomas Owens Mod
    Aug 10 at 14:29
  • its definitely skewed towards the usual suspects, it you say more than 10 posts its just 4. But I guess there are two questions here, 1. how can we get the data?, 2. are we getting to cliquey? 3. if we are how do we stop? THREE! three questions
    – Ewan
    Aug 10 at 14:32
  • I suppose if there are 100 people who regularly answer one question a month, you could say "don't close new questions for at least 1 month" because this then allows x% of the total answerers to get a chance
    – Ewan
    Aug 10 at 14:35
  • 2
    This is definitely an interesting line of questions. The SEDE is the best place to get the data, since it's refreshed regularly. But this is a hard set of questions to query. For example, I've gone weeks without answering questions because there either hasn't been a question that really piqued my interest or someone beat me to it so I just tossed an up vote and moved on, so depending on your threshold, I could be falling out of the data. So setting these activity thresholds could be interesting. I can ask about if the CM or data team has some bandwidth to toss some thoughts on this matter.
    – Thomas Owens Mod
    Aug 10 at 14:45
  • @Ewan This query was also brought to my attention. Something like this can be run in SEDE. Or give ideas.
    – Thomas Owens Mod
    Aug 10 at 15:07
  • 2
    data.stackexchange.com/programmers/query/edit/1774466 you can run this query, and sort by the second column (# posts). This is over 3 years, change the time frame to whatever you want. 1702 people posted a single answer, but there are 11 people who each posted over 200 answers each over that time frame, so those 11 people have more answer output than all the one-answer posters combined. Could easily modify the initial select to only include answers with score>1, etc.
    – Bryan Krause
    Aug 10 at 15:08
  • hmm im not quite sure about that query, but yes I don't think anyone is going to be surprised that a small number of users are superactive, I dont think thats a problem, unless they are closing questions before the main body of answerers has a chance to see them.
    – Ewan
    Aug 10 at 15:45
  • for an extreme hypothetical example, If thomas knows about display drivers and answers one question a month but the superactive users only know about RESTful route naming and close questions we don't understand. Questions about display drivers don't get answers
    – Ewan
    Aug 10 at 15:50
  • 2
    Thomas' related feature-request on MSE, inspired by this discussion: Community Health Dashboards
    – V2Blast
    Aug 10 at 19:10
  • 1
    @Ewan Which part are you not sure about? If you mean that it's a bit orthogonal to what you're asking about, yes, indeed, I only meant for it to be useful in sort of setting the scope of the problem by answering "how many answers are posted by a few people relative to the total number of answers" (the query was designed for something entirely different, I just noticed it could be helpful here). Finding out if those answers are focused on very specific topics, etc, how it relates to downvoting, etc - those are of course separate and not helped by this query.
    – Bryan Krause
    Aug 10 at 20:30

You must log in to answer this question.

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