A lot of people on HN seem to have this bizarre rejection of any evidence that wasn't collected by objective experimentation--something which isn't possible in this sort of context. So you do the next best thing: you survey people, and their responses (i.e., anecdotes) form the data. As with experimental data, in the real world, different things will happen to different (test) subjects, which is why you need to collect sufficient data to have a representative sample.
In this case--why does it matter that Scandinavian women are less likely to choose a career in STEM? How is that relevant to Silicon Valley, or the US computer programming industry? Why does it matter that women have a 2:1 preference for STEM jobs--that's just a reaction to the initial problem, namely--why aren't women pursuing STEM jobs?