3

TL;DR: I looked at some Data Explorer data for Stack Exchange's language sites. It seems

  1. we create a lot of answers compared to our peers (is this good or bad?),
  2. we really should be voting more (we're behind our peers), and
  3. we should be making more noise here on meta.

Data

Here's a snapshot of some statistics of Stack Exchange's language sites:

                        (total)      (1-month period)
                    #Quest  #Meta  #quest #ans #votes (*)
                                                 up  down

English.SE          108940   3279     890 1101  922   321
ELL.SE               71092   1013    1019 1146  627   162
Japanese.SE          20344    699     185  191  518    45
German.SE            12878    482     102  251  458    70
French.SE             9671    251      96  200  261    19
Chinese.SE <---       7546    221     122  322  128     2
Spanish.SE            7010    416     112  186  217     4
Russian.SE            3908    158      31   68  133    13
Latin.SE              3332    130      66   99  275     7
Italian.SE            2966    135      34   49  218    14
Ukrainian.SE          1858     92      24   36   61     6
Portuguese.SE         1753    116      22   36   52     1
Esperanto.SE          1336     77      27   40   27     2
Korean.SE             1203     71      18   17   43     2
LanguageLearning.SE    978    161       7   11   18     4
ConLang.SE             279     49       6   10    8     1

(*) for a month starting 14-th November 2019 the number of questions an answers from this data explorer query, and for November 2019 the number of votes is from this data explorer query counts votes made within 31 days after the question was created.

With this data explorer query (and this modified version), we can find the average post scores per site:

                      average score
                      quest  answer

ConLang.SE            8.51    5.64
LanguageLearning.SE   8.11    4.02
Latin.SE              6.46    5.23
Esperanto.SE          6.12    4.69
Portuguese.SE         5.82    3.94
Ukrainian.SE          5.79    4.21
German.SE             5.46    4.46
Russian.SE            5.13    3.70
Japanese.SE           4.54    4.14
Spanish.SE            4.41    3.45
French.SE             4.38    3.47
Korean.SE             4.20    3.00
Italian.SE            3.94    3.95
English.SE            3.90    3.49
Chinese.SE <---       2.85    2.16
ELL.SE                1.88    2.54

(Further statistics are on Sites.)


Conclusions

Or at least, what I think we can conclude:

  • Recently, we've been creating a lot of answers. I'm not sure if that's is a good thing, or if this simply implies we create many "meh" answers.

  • We're probably not voting enough (see: Vote Early, Vote Often!). Other sites of comparable size cast more votes on fewer posts.

    1. Scant nice questions/answers. In fact, searching for is:question score:10 and is:answer score:10 shows that for posts made in 2019, there has only been one "nice question" or "nice answer" badge awarded (here).

    2. Hot Network Questions. There's a formula for question "hotness"

      (MIN(AnswerCount, 10) * QScore) / 5 + AnswerScore
      -------------------------------------------------
                (QAgeInHours + 1) ^ 1.4
      

      Getting questions on the "Hot Network Questions" list is great exposure for the site (often resulting in 1000+ visits), but we can't hit it unless users upvote both the question and its answers shortly after posting.

  • Meta Chinese.SE seems a bit quiet. It's not a huge problem, but I feel users should be making more noise on meta: discussing if/how we're doing things right; coming up with creative ways to encourage participation, improve quality etc.

1
  • 1
    @everyone remember: If you take the time to answer a question, you should interpret it as a good question that hasn’t been on this site before, and upvote it! Conversely, why would you answer a not good question that’s already been answered on the site or shows no research?
    – dROOOze
    Dec 15 '19 at 17:29

You must log in to answer this question.

Browse other questions tagged .