The dominant model of vocational preferences today is John Holland’s (1959, 1997) RIASEC model. Fifty years of research on Holland’s model has supported the utility of conceptualizing human personalities and work environments in terms of their resemblance to six prototypical categories: Realistic, Investigative, Artistic, Social, Enterprising, and Conventional. Researchers who employ the Holland model typically refer to each vocational-personality type by the first letter of the type label; hence the acronym RIASEC. Of particular interest to the question of women in economics is the RIASEC classification for an economist and research on gender differences in RIASEC preferences. Economists, like research psychologists and other scientists, are considered to be primarily Investigative. Investigative individuals like working with ideas more than dealing with people. They do not mind laboring long hours in relative isolation. They also exemplify a cognitive style that Welsh (1975) termed high Intellectence. High-Intellectent Investigative persons disengage and distance themselves from the sensate world, preferring to relate to the environment indirectly through abstract symbols...
Psychological research has consistently demonstrated that males, as a group, score higher than females on measures of Investigative, Realistic, and Enterprising vocational preferences, while the reverse is true for Social and Artistic preferences.
And if you like understated ridicule:
If there is hostility in the sciences toward accepting women, Browne (2006) notes that it takes an odd, selective form by subfield. Women earn relatively few doctorates in mining/mineral engineering, but considerably more in bioengineering. Biology is apparently welcoming to women, as women earn 45% of all doctorates in biology. So is medicine, as over 40% of new doctors are women. But biophysics must be hostile, because relatively few women earn a doctorate in this area. Women earn 67% of all doctorates in psychology, but mostly in developmental and child psychology. Few women earn doctorates in psychometrics and quantitative psychology.
Personally, though, I think that gender balance in academia is already over-studied. I'd rather see some work on gender balance in the blogosphere. Does anyone want to channel Larry Summers?