Presumably the article's author is so used to "sexism" meaning "women are under-represented" that they don't realize "men are under-represented" is equally sexist.

Not that under-representation is necessarily evidence of bias, which is what the word "sexism" implies. The study's authors, Ceci and Williams, are aware of this:

> This led to their conclusion that gender bias did not cause the small number of women in STEM fields. They said the small representation of women in the sector was caused by their own reluctance to enter these fields.