5

Not a question, but an FYI for those who are interested in some site stats.

Intrigued by user6726's impression that "no syntax trees" was one of the main close reasons on this site, I asked the Stack Exchange API about it (code). Here are the close reasons sorted by number of questions (in brackets) closed for that reason in the past ~3 years on Linguistics SE (Σ = 245):

  1. (69): off-topic: language-specific
  2. (39): needs details or clarity
  3. (37): duplicate
  4. (34): needs more focus
  5. (20): opinion-based
  6. (18): off-topic: other
  7. (16): off-topic: language identification and translation
  8. (12): off-topic: syntax trees
  9. (0): off-topic: other site in the network

As it turns out, syntax trees are in fact the second-least common close reason on our site according to that statistics. However, note these numbers count only those closed questions that are still on the site. Closed questions meeeting certain criteria for very low quality are automatically deleted by the system -- these low-quality questions, as well as questions migrated to other network sites, are excluded here. It may well be that "Please draw me a syntax tree" and "What language is this/Translate this for me" questions are in fact way more common, just that they tend to be more badly received with negative score and no answers than e.g. duplicate questions are, and are therefore more likely to be automatically removed from the site and hence underrepresented in this list. I am currently not aware of a way to automatically access the "live" close data.

7
  • (1) I'm dying to get the stats on "closed and obliterated" and (2) apparently there's such a thing as "awareness bias", i.e. surprise!
    – user6726
    Mar 30, 2020 at 23:51
  • Re (1): Now that you say it, I just realized that accessing the API was a bit of cracking the nut with a sledgehammer, as you can actually partially see apparently precisely these stats in the so-called mod tools, which afaik you as a >=5k user should be able to access as well: Review/"Tools"/tab "Stats"/section "Links"/ "question close stats" -- though only for up to the past 90 days, and not separating alive vs. obliterated. In these statistics, off-topic reasons are generally higher ranked (which somewhat makes sense) with syntax trees on place 5 overall; top #1 and #2 remain the same. Mar 31, 2020 at 0:15
  • People with 2000 rep can also see the close stats here: linguistics.stackexchange.com/tools/…
    – curiousdannii Mod
    Mar 31, 2020 at 1:46
  • Your stats can't be right. We've closed 29 questions as language-specific just in the last 3 months! And there are 681 closed questions that haven't been deleted. Your stats add up to 35% of the total. I'm not sure why though.
    – curiousdannii Mod
    Mar 31, 2020 at 1:47
  • @curiousdannii Because many of these will already have been deleted, and because, as written, I restricted my search to questions active the past 3 years. With that, the numbers roughly match up (there are still a few questions more in the search for that time range; it's possible that duplicate and migrated questions are treated differently than those with the [closed] annotation). Mar 31, 2020 at 10:03
  • @lemontree Oh, I missed the 3 year time limit. Yes with that the numbers make sense. Sorry!
    – curiousdannii Mod
    Mar 31, 2020 at 10:13
  • I was writing a SEDE query for this and was over an hour in when I found out that the sub-reasons of Off-Topic aren’t accessible in SEDE D=
    – Jan
    Apr 13, 2020 at 15:05

0

You must log in to answer this question.

Browse other questions tagged .