Michael Arrington's real argument was: Few women choosing to study CompSci\Engineering -> Few women in tech _startups_. As long as you agree that a subset can

Michael Arrington's real argument was: Few women choosing to study CompSci\Engineering -> Few women in tech _startups_. As long as you agree that a subset cannot be larger than its parent set, I don't see how you can disagree with that. It's as close to observational fact as you can get.