What's interesting is when people do the opposite. When they A/B test the name. Same content. Half the people have male names, half female. Now, if gender doesn't matter, and people are judged purely on their achievements, and we are in a real meritocracy, then there should be no difference, right? If there a statistically significant difference in how people treat the male vs. female applicant, then there's some bias going on.
Turns out if you do that (e.g. http://blogs.scientificamerican.com/unofficial-prognosis/201... ), you get men getting more job offers and getting offered more money. And this is scientists judging applicants for a science job!
Something is wrong here....
The paper assumes independence between the gender of the applicant and the gender of the students of the professor, which isn't the case unless you hide the applicant in a room away from the current grad students.
Furthermore, the paper mentions that they used both Student's t-test and an ANOVA, both of which assume that the underlying data is normally distributed; there was no mention of a normality test anywhere in the paper. If the data is not normally distributed, which could be the case, then it's violating the normality assumption of both tests, which could potentially make them invalid.
Finally, are there other confounding factors in the study? For example, if the applications were sent in different batches, the time at which they arrived does have an influence. For example, a recent paper evaluated the decision of judges for parole and found out that judges were more lenient right after lunch and were much more strict right before[1].
As they say, further study is required.
I wish somebody woyuld at least look into possible reasons. Personally I suspect the elephant in the room is motherhood. If there is an x% chance for the recipient of your grant to quit the profession after a couple of years, it lowers the value of your investment quite a bit.
You think women can't be sexist against women? Of course they can be! A member of the marginalised group can easily continue that marginalisation. I watched a recent current affairs programme about same sex marriage, and they'd found a gay man to speak out against gay marriage.
If you name-blind the review process you would still expect to get 95% men unless there is a higher % of women submitting talks and being rejected because they are women.
The quoted article suggests doing the opposite of name-blinding and actively looking for female talks in the pile.
https://news.ycombinator.com/item?id=5036457
> Everybody can vote for DjangoCon 2013 presentations (Github required)
> The voting is deliberately anonymous at this first stage.We DO use names to discriminate but that often allows us to have "fresh" speakers and keep the lineup somewhat varied. It also allows us to filter for all the other spam-talks; i.e. marketers.
It is an issue of priorities but I would easily place quality of presentations over guaranteed blinding to prevent gender-bias selection.
I only wish it was possible for smaller organizations to take advantage of blind review processes for speakers.
The balance of evidence is that she is dismissive of the problem. She says she disagrees with Debra Mastaler, she uses scare-quotes around the word "bias", all but one of her "solutions" are aimed at female speakers individually, of the "do more to get noticed" variety. (The one standout is her accusation against Forbes Woman.)
Nowhere does she acknowledge that a) a number conferences have managed to get much, much closer to parity, without resorting to quotas, and b) a conference that still ends up with 95% males therefore very much has a problem.
This is known as "not doing your homework".
A tech conference that ends up with 95% male speakers is Doing Something Wrong. They have no right to be "happy with their process".
One of the "correct" solutions has already been discussed in many places - use a name-blinded review process, and make sure your CFP outreach activities include women-intensive groups.