Till Westermayer (till(a)tillwe.de) [050529 21:37]:
catching up after holidays, I'm not sure if
anybody mentioned it, but at
least theoretically it would be possible to make statistical
comparisions between anonymous ratings and logged-in ratings (if anon/
logged in is stored), so after a while we could show if there a
statictically significant differences between both ratings.
Yep. Having gathering the data and processing the data as two completely
separate things will be a good way to go forward, because we can build
tools on top to analyse the data any way we happen to feel like.
And on a related topic: I think ratings (who rates
what how) should be
visible, meaning raters are conscious about their decisions and the
consequences.
Yes. Though it'll need to be noted on the ratings page - anon users who
rate things on other sites wouldn't normally expect the source of a rating
to be world-visible. Something like:
"Note: <b>All ratings are readable, just like article edits, with the
person and time.</b> Like edits, they are part of the process of improving
Wikipedia and so responsibility is attached to them. See the <a
href="[whatever]">privacy policy</a>."
This will discourage some anon raters, but if we don't then many people
will get a nasty surprise and be quite upset.
That second sentence, giving the rationale, needs work.
- d.