The question of why academically able girls are less likely than boys—even less accomplished boys—to enroll in science, technology, engineering, and mathematics (STEM) programs, especially in physical sciences, technology, and some engineering disciplines, has long bedeviled researchers and educators . The discrepancy persists “despite efforts to overcome preparation deficits , provide role models and mentoring , and build communities  for women in sciences," writes Marie-Claire Shanahan of the University of Alberta in Edmonton, Canada, in an essay at Inside Higher Ed  (links in original). "Accordingly, we must acknowledge this is a more complex problem," she adds.
A recent study of students in the Netherlands  explores one possible aspect of that complexity: the influence of average differences in competitiveness on students' choices of curricula. At age 15, university-bound Dutch secondary students must choose among four possible programs. Ranked in order of prestige and presumed academic difficulty, they emphasize science and technology, health, social sciences, and humanities.
The science track, considered the most prestigious and difficult, attracted 40% of the boys in the study's sample but only 17% of the girls, even though the girls, on average, had math records as good as the boys and better overall grades. Humanities, the least prestigious, attracted 15% of the girls but only 8% of the boys.
The researchers devised an academic game to test the students' competitiveness and found that the girls, on average, are significantly less competitive than the boys. They also found "that competitiveness varies strongly and significantly across [the curricular choices], with students that are more competitive selecting more prestigious" programs, write Thomas Buser and Hessel Oosterbeek of the University of Amsterdam and Muriel Niederle of Stanford University in Palo Alto, California, the study's authors. "Competitiveness is as important a predictor of [program] choice as gender," they write. "More importantly, up to 23 percent of the gender difference in [program] choice can be attributed to gender differences in competitiveness."
Some time back, we reported on other research  showing that differences in the importance that female and male college students ascribe, on average, to caring, serving, and helping as opposed to attaining mastery, power, and prestige correlate with the likelihood that they will choose non-STEM or STEM careers. The more a student valued helping and caring, the smaller the likelihood was that he or she would choose to study STEM.
In her essay, Shanahan argues that student choices are strongly influenced by "tangled webs of expectations that influence all students' experiences in science degree programs." These expectations "paint a picture for students of whether their science program is really for someone like them. And it's here where many female students encounter difficulties in meeting the expectations."
Clearly, this is a complex problem that cannot be explained by any single factor or rubric.