2009/5/21 Robert Rohde <rarohde(a)gmail.com>om>: Do we have a rough estimate of qualifying voters who didn't vote? 17000 is pretty good, but it occurs to me I have no idea how large the editing community really is! -- - Andrew Gray andrew.gray(a)dunelm.org.uk

On Thu, May 21, 2009 at 10:53 AM, Thomas Dalton <thomas.dalton(a)gmail.com> wrote: I believe there are around 600,000 qualified accounts (roughly half of which from enwiki). If you choose a historically active account at random, there is somewhat less than a 10% chance that any given account will edit in any given month. Which suggests that the pool of people who are qualified and likely to have seen the site notice is around 60,000. If that number is in the right ballpark, then actual participation from qualified people who saw the announcement was perhaps 20-30%. -Robert Rohde PS. Incidentally enwiki has 9.7 M registered accounts, but 70% of these have exactly 0 edits and 90% have less than 5 edits.

On Thu, May 21, 2009 at 1:07 PM, Thomas Dalton <thomas.dalton(a)gmail.com> wrote: As of last September, exactly 297467 enwiki accounts had made 20 or more edits. [1] The distribution is compellingly log-log, which implies about 255,000 accounts on enwiki with 25+ edits on that date. Extrapolating the rate of increase forward 6 months gets you to about 300,000. Other analysis generally indicates that scaling enwiki results to the global wiki community generally requires a factor of a little more than 2. (Taking Sue's number at face value, which strikes me as a little high, would imply a factor closer to 3.) Doubling 300,000 is 600,000. Maybe that needs to be higher, e.g. 750,000 or something, but I'm fairly confident it is in the right ballpark. Also, there is a good deal of overcounting with individuals having accounts on multiple wikis. Incidentally, Tim Starling actually generated a list of qualified accounts during the setup process. Assuming he still has it lying around, we could possibly get an exact count. -Robert Rohde [1] http://en.wikipedia.org/wiki/Wikipedia:Editing_frequency/All_registered

On Thu, May 21, 2009 at 2:14 PM, Sue Gardner <susanpgardner(a)gmail.com> wrote: These numbers make sense to me. 10% voter turnout is relatively low when compared to public elections in general. But then again there aren't too many online votes of this size held among volunteers of charitable organizations, so there isn't a lot of direct precedent to compare to. I doubt any of the larger projects ever get to even 10% turnout for their various community discussions, and important decisions are routinely made in the project with far fewer votes then that. --Andrew Whitworth

On Thu, May 21, 2009 at 10:00 AM, Robert Rohde <rarohde(a)gmail.com> wrote: Yes -- I think this is definitely the largest group of Wikimedians to ever collectively express an opinion on anything! It'd be worth figuring out why the vote was successful, if possible (long period of voting? ubiquitous sitenotices? Important topic? Lots of outside interest?) -- phoebe

On Thu, May 21, 2009 at 5:46 PM, Michael Snow <wikipedia(a)verizon.net> wrote: And yet the "threshold for eligibility" hypothesis has not been tested on the projects. You have no idea whether allowing only those with the most biased opinions to vote (as most project votes are conducted) skews the outcome towards or away from the rational or optimal choice, or whether it has any effect on the outcome at all. Indeed, we have no idea whether the wording or presentation or usability of the votes matters. It could matter a great deal, changing the outcome in a statistically significant matter, or it could matter not at all, rendering the threshold for eligibility hypothesis meaningless. The current methods amount to folk statistics because nobody has any clue what matters and what doesn't. That's why I continue to encourage the WMF to adopt scientific thinking.

2009/5/22 Michael Snow <wikipedia(a)verizon.net>et>: Do we have any statistics for what the turnout was among different demographics? In particular, do we know how many people voted that wouldn't have been eligible under the board election suffrage rules? If it isn't many then that can't be the explanation.

2009/5/22 Anthony <wikimail(a)inbox.org>rg>: You're probably on to something there.

2009/5/22 Anthony <wikimail(a)inbox.org> 1) at which project (and please dont use enwiki abbreviations) 2) does it matter? :) eia

2009/5/22 Anthony <wikimail(a)inbox.org> Ah, meta? :) - seriously, i don't see why only to consider enwiki methods. Sure, most voters came from there, but this is an interwiki decision, where interwiki measures would make more sense. But well, not that it matters anyway, these things should be decided /beforehand/ not afterwards. well, since those are about people, and this isn't, I dont see any relevance :P Usually the two categories are treated differently. -- eia

No opinion means no opinion and should not be interpreted in any way, the group represents an uncertainty in the result. John Erik Moeller skrev: