[Foundation-l] LA Times article / Advertising in Wikipedia

Tue Mar 11 21:17:50 UTC 2008

On Tue, Mar 11, 2008 at 4:45 PM, Mike Godwin <mgodwin at wikimedia.org> wrote:
>  Andrew Whitworth writes:
>  As someone who was a statistician in another life, I feel compelled to
>  point out that even this wouldn't necessarily elicit a representative
>  sample of the community's opinion -- it samples, at best, only that
>  subset of the community who (a) votes in elections, and (b) would be
>  willing to offer an opinion in the course of voting in an election.

True, but it's slightly more representative then the opinions that we
see posted here on foundation-l.

>  What makes this more complicated, though, is the question of whether
>  Wikimedia should gear its strategies in response to what the
>  "community" (however defined) wants, as distinct from what the world
>  needs.

Another good point, Mike. However, if getting a representative sample
of our community is not likely, getting data from a representative
sample of the whole world is far less so. Either way, we're living in
abject data poverty here. We could make the stretch that Wikimedians
themselves represent a good cross-sample of the world's population
(and this is a BIG stretch), and then we could say that a good sample
of our users is also an acceptable sample for the world at large.

A little bit of statistical gerrymandering never hurt anybody, right?

--Andrew Whitworth