[Foundation-l] Attribution survey, first results

Anthony wikimail at inbox.org
Wed Mar 4 18:56:48 UTC 2009


On Wed, Mar 4, 2009 at 1:48 PM, Robert Rohde <rarohde at gmail.com> wrote:

> On Wed, Mar 4, 2009 at 10:22 AM, Thomas Dalton <thomas.dalton at gmail.com>
> wrote:
> > I'm not a statistician, someone else can work out how large a majority
> > is needed from a sample size of 570 to be confident (at the 95% level,
> > say?) that a majority of the population as a whole agrees.
>
> If the 570 people are a RANDOM sampling of the underlying population:
> 307 people (53.5%)


> If 307 out of 570 people (53.5%) agree with statement X, you can be
> confident at the 95% level that at least 50% of the underlying
> population would agree with X.
>


Thanks for the specific number.  I was under the impression it was something
like that, but it's far outside my area of expertise.


> Of course the current sample is not random


Far from it, though it pretty much confirms my suspicions (actually, I
thought the number of people who wanted their name listed would be lower,
around 5-10%, not 20%).


> and I don't think rights should be apportioned by simple majority either.


Thomas's latest statement suggests that he doesn't either, but then, that
brings back up my question as to what *does* constitute a sufficient
majority.



More information about the wikimedia-l mailing list