Which specific biological mechanism are you hypothesizing is responsible? In combination with which specific cultural construction of computer science that is somehow inalterable?

Whereas we have pretty good evidence that by improving the signaling of ambient belonging for women we can significantly increase their interest in computer science: http://ilabs.uw.edu/sites/default/files/Cheryan_Meltzoff_Kim...

Few things with a biological mechanism are that easy to manipulate.

The short answer is yes, all of the statistics which show a deviation from a priori gender balance suggest that there is "something" going on.

That "something" is almost certainly cultural, since there is little evidence that any group based genetic hypothesis has explanatory power, whereas cultural forces have clearly been shown to impact gender balance. e.g. the right to vote.

Yes. There is SO MUCH wrong with ballet. So much. Like, way more than is wrong with tech or computer science. Just... so much is wrong with ballet.

Of course, while a majority of ballet dancers are women the people who run ballets are almost entirely male, which should give you some insight into what is wrong with ballet.

Indeed, if you want to feel better about tech I have never heard of a tech company where developers were openly expected to sleep with the person running the team to get a promotion, so we are beating the New York City ballet by a mile.

Possibly? Or at least in the social pressure at school that stops boys from dancing, and/or the way that dance is taught.

"CHOICE" is not a get-out-of-sexism-free card. Anyone who builds products or sells anything to anyone ever should know that we have enormous influence over who chooses to consume our products.