Thanks for the response Andreas. I've updated with the 8.5% source.
I'm not selling Hill as a panacea either; there are actually lots of techniques to correct biased sampling and using another survey as a benchmark only works well if the demographic questions are the same or at least very closely matched. I haven't compared the Pew and UN surveys in detail but I'm sure it could be done better (that's pretty much how these things always go)!
One thing I didn't criticise you for yet (but will now!) is to dismiss the claim of family status's effect on contribution based on the data you provided.
I agree that the fact that even young women are very unrepresented means it likely doesn't account for a large portion of the gender gap. But your argument that the fact that the *bulk* of wikipedians are younger people means that family status isn't an issue is particularly erroneous, because it assumes that age demographics and family status are independent- which they are quite clearly not. It could be, for instance, that gender gap is smallest in the younger demographic *because* they don't have families yet, and the proportion of women drops with age because they drop out to have families. (Not an actual hypothesis I'm proposing, just an example of how assuming independance goes wrong fast.)
It's important to keep in mind that the actual reason for the gender gap is probably a large number of very small things and no particular one of these things likely accounts for a very large portion of the gap- if that's the case we'll need much better statistical power to detect them and more sophisticated analyses.
Anyway, thanks for writing this post- got us talking, and regardless of how the actual numbers kick up it's still pretty clear there aren't very many of us.
On Wed, Aug 27, 2014 at 8:42 PM, phoebe ayers phoebe.wiki@gmail.com wrote:
On Wed, Aug 27, 2014 at 12:22 PM, Andreas Kolbe jayen466@gmail.com wrote:
I will have to look into Hill & Shaw, but would note that the Wikimedia Foundation itself reported the figures from the UNU survey as they stood (see e.g. p. 8 of the February 2011 Strategic Plan: "According to the study, over 86% of contributors were male").
NB., that was before the Hill & Shaw paper was published, which was 2013 :) Hill & Shaw is *probably* the best estimate of the gendergap we have so far, but everyone -- including the WMF and the researchers involved -- knows that the data can be improved. And hopefully it will be, with future editor surveys and more research!
-- phoebe
Gendergap mailing list Gendergap@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/gendergap