Talk:Strategy/Wikimedia Foundation/2016/Community consultation
|This phase of the consultation is closed. For information about the outcome, please watch the consultation main page, where a pointer to the next step in this process will be posted on or around February 26. While you are welcome to continue to use the discussion pages of this phase, please know that future submissions to this phase may not be reviewed by staff. We look forward to talking to you more about the themes that have emerged in the near future!|
On Wikimedia-L, the following two questions were asked by User:Legoktm:
- 1. How was a user's "home wiki" determined?
- 2. Page 29 says that 17 users have 0 edits. AIUI you had to edit the wiki to participate, so how do they have 0 edits?
While our strategy consultant is flying today and the person who helped collect data is off, I will try to get responses to these questions as soon as possible. :) --Maggie Dennis (WMF) (talk) 16:28, 29 February 2016 (UTC)
- Sorry for the delay answering this while I was out.
- The home wiki determined is quite simply the CentralAuth homewiki (which is delivered back through the API when requested). This is certainly not the most robust choice and has holes (users who started on English Wikipedia but then migrated to their native language or Wikidata for example) but so do other means of determining home wiki such as most edits/recent edits etc. I do think it is a good rough approximation especially for what it was used for (to get a general sense of how diverse the pool was) and, if anything, steers the result to 'less diverse' (it is more likely to say that someone's home wiki is English, for example, when it's really something else).
- Looking at the data behind the report it looks like the 0 edits were mostly IP editors (which got counted in the vote count but had 0 editcount because we were just looking at account edits). Unlike the 2015 Strategy consultation, where we ran banners to anonymous users, there were very few IP editors in this bunch so Suzie lumped them in with everyone else instead of trying to treat them differently or by country (which would have made it very hard or impossible to anonymize). This is especially true since it's likely those users were mostly (or completely) editors who edited logged out by accident. There were also users (non IPs) that resulted in 0 edit counts and appear to be because they signed/adjusted their comments to appear to be from non existent usernames when my script analyzed the comments. Which then, unsurprisingly, didn't find any edits. Jalexander--WMF 05:48, 2 March 2016 (UTC)