The cross sectional data is interesting and certainly puts a dent in simplistic explanations that only the patriarchy is preventing gender parity in STEM. But then again, variation in male/female average preferences by time and place works overall in favour of arguments that cultural factors like stereotyping influence career choices (and against biological predispositions being the one true explanation for female underrepresentation). It's not entirely impossible that women are simply more inclined to pick professions other than tech as barriers to other careers are removed (I'm sure the stereotypes that women don't belong in finance or law are stronger in the UAE and were stronger in the 1980s US). But once one acknowledges that alienation [and anticipation of discrimination, and role models] probably plays some role in career selection, the question becomes why it wouldn't be a driving factor in [self]selection for a field where US gender gaps were much smaller when it was a non-traditional niche attracting comparatively little attention than when it was a mainstream white collar career choice but one whose male nerd stereotypes are firmly ingrained in public consciousness.