It's a tricky question to address. It doesn't help that the people addressing it are obviously not entirely clear on the concept of an unbiased sample. To quote

It's a tricky question to address. It doesn't help that the people addressing it are obviously not entirely clear on the concept of an unbiased sample. To quote the fine article: But let’s say a larger percentage of the pitches came from women. Then people are also assuming that of the pitches that came from women, 100% of them were awesome pitches that organizers passed up solely because the presenter would be wearing a bra. No, people are assuming that pitches from women and men are on average of equal quality. It's an assumption borne out by experience. So we do need to ask the question why there are less women. If indeed 95% of the proposals were from male speakers, it's not a problem with the conference. If there were more than 5% of the proposals coming from women, but the final selection looked like it did, the organizers should take a look at their process - something is likely to be off if that's the case.