We are technologist. Many of us are academics. If someone writes or talks about how a compiler is slow, the community will demand more than anecdotes. We expect numbers, tested theories, and verified results. When people come to us with anecdotes, we often use "less than nice" ways to disregard what they say. hmm, I see parallels...

There looks to be a number of excellent evidence to prove that there is a gender equality issue in IT, but there is no scientific research on what the causes are. Is it the general culture in the world and IT has the same level as other industries? Is it because of the question/answer culture of engineers, which is an aspect currently being researched in regard to gender equality. Is it a lack of respect for women in the IT field?

I can think of plentiful number of ways to test those concept, work that anthropology might want to do if we asked them. We could also pay them to do it. Maybe that could be someones kickstarter project?

I have a strong conviction that if someone would bring scientific data on the issue on what the causes are and how to fix them, in a verifiable way, few would object. Can a strong worded policy document have an statistic proven effect in discourage sexism in a conference? If you think so, verify it with data. Get 10 conference to do it, and collect data from it and 10 other without any such document. If its verifyable, every conference should adapt one after that. Do you think negative IRC comments is the problem? Switch a number of projects to a moderated communication channel and test the concept. Compare a unmoderated IRC with a project page like stackoverflow. Bring data, do some science, show it to the community. Its the exact same thing we expect from anyone else when they want to do changes to the things the community care about.