Jump to content

Wikipedia talk:Labels/VE experiment edits

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Welcome!

[edit]

Hey folks. Welcome to the VE experiment edits campaign. Please feel free to ask questions and make proposals here. --Halfak (WMF) (talk) 20:00, 26 June 2015 (UTC)[reply]

Developing the form -- what questions should we ask?

[edit]

Before we start the campaign, we'll need to develop a form with questions that we'd like answered about each edit. I think that the list should be short so that we don't overload editors with questions while they are looking through the edits.

I have a proposed list to start us off:

  • Is this a productive edit?
    • If not, is it at least good-faith?
  • Is the edit summary useful?
  • (freeform) What types of changes are made in this edit.

--Halfak (WMF) (talk) 20:00, 26 June 2015 (UTC)[reply]

Maybe the type of edit could include these kind of things:
  • Improved formatting
  • Updated data (numbers, dates, names)
  • Rewrote text
Helder 15:21, 27 June 2015 (UTC)[reply]

I liked one of the suggestions from Meta: Can you tell (by looking only at the diff) which editing environment was used? This would require that all tags be removed from the diff. It would be most interesting to have this paired with a question about whether something is screwed up in the edit (e.g., stray markup, unpaired brackets, broken ref tags, etc.). Whatamidoing (WMF) (talk) 00:50, 29 June 2015 (UTC)[reply]

A few brief thoughts that may be helpful to you.
  • If you aren't aware of them, you might be interested in STiki and Huggle. For example, reverts using these tools include identification of "good faith" status, and the list of edits to check is created by some fairly effective automatic classification as well.
  • Based on the percentages at the STiki leaderboard, what different editors consider to be vandalism, AGF-revertable, or acceptable varies dramatically. From edits in the STiki queue, some editors revert as many as 95% of edits and others revert as little as 5%.
  • For your suggested types of edits, I would note that some fairly common types of vandalism fall into those categories, especially updating data (which also may not be obvious, e.g. changing or replacing names is a common theme).
--Sunrise (talk) 06:51, 29 June 2015 (UTC)[reply]