Yesterday, this paper hit my Twitter feed. It essentially argues that the gender gap in academic disciplines is due primarily to the specific perception by members of that discipline that raw talent is essential for success. That is, the disproportionate number of of male Ph.D.s in physics or math is due to physicists and mathematicians believing that raw talent is the most important element of success in their field. For comparison, in my field (molecular biology), 54% of Ph.D.s go to women, and that is explained primarily by the fact that we value hard work and training more than raw talent.
I initially responded to this paper (in the latest issue of Science) by asking if all of science was being trolled by the journal. I mean, the abstract reads like an article from the Onion. It turns out, though, that the paper makes a lot of good points, and uses data (albeit soft poll data) to do it. It’s also interesting because it doesn’t JUST consider stereotypically “low female” fields in science, technology, engineering and math, but also social science fields. For example, male Ph.D.s in Philosophy and Economics far outnumber female Ph.D.s, while there is parity in Political Science and History, and an excess of female Ph.D.s in Education and Psychology. What’s really interesting is that the data in this paper support the hypothesis across fields.
In the paper, the authors test their hypothesis (higher “field-specific ability beliefs” = fewer female Ph.D.s) against three competing hypotheses. The first of these is that fields in which the participants work more hours attract less women (I call this the “Mommy Hypothesis”). The Mommy Hypothesis is entirely unsupported by the data – there is no correlation between hours worked and percent of Ph.D.s going to women across disciplines.
The next hypothesis tests the selectivity of the discipline against number of Ph.D.s who are female, and, again, finds no significant correlation. (The measure of selectivity used was the percent of Ph.D. applicants who are accepted each year – lower acceptance rate = higher selectivity.) While I may quibble with this as a real measure of selectivity (if your institution is famous but not that rigorous, you may still have a LOT of applicants per space, and, of course, money is always a factor in how many applicants an institution may take, so it’s possible that selectivity might have nothing to do with overall applicant quality), the authors nonetheless found that the OPPOSITE (but statistically insignificant) trend is present: more selective disciplines have, on average, more female Ph.D.s. The authors attempted to backstop their measure of selectivity with scores on the Graduate Record Examination (GRE) within all fields for which data are available – higher scores on the GRE should reflect higher-quality applicant pools. Again, there was no correlation between GRE scores and percent of female Ph.D.s
Finally, the authors tested the hypothesis that the TYPE of work done in a field might influence the numbers of women vs. men awarded Ph.D.s. Specifically, they looked at whether the field tended to be more “systemizing” or “empathizing” – that is, is it a cold, hard, logical boy place or a soft, warm, emotional girl place? Here there was a significant correlation, but only at the most basic level. Not to get to deep into the nitty-gritty details of the analysis, but the correlation was there when all of the data were considered together, but the “systemizing vs. empathizing” score could not predict whether a field’s Ph.D.s are male-dominated.
The next question is whether this perception of “innate” talent being important predicts the attitudes of its members towards women. That is, do the members of fields valuing raw talent also think that women, on average, LACK that talent. The answer, unsurprisingly, is yes. On average, in fields where the field-specific ability belief score is high, members of that field were “more likely to endorse the idea that women are less well-suited to high-level scholarly work”, and that higher endorsement of this idea correlates with lower female representation. Finally, participants were asked to rate how welcoming their fields are to women, and…well, the results are exactly what you’d predict.
So, what does this mean? Well, maybe nothing. Correlation does not equal causation. But there’s a lot of correlation here, and I think it must mean something. The authors offer several potential explanations for their results, which are not exclusive of one another. Women may be discriminated against in “brilliant” disciplines because there is a stereotype that women, generally, are far less likely to have innate genius than men. Women may be less likely to enter fields where they know they are perceived to be less likely to succeed, due to stereotype threat. Women may be less likely to enter or stay in a field because it is perceived (by them, and by those IN the field) as unfriendly to women (I’m looking at you, Physics and Engineering). Women may be more likely to enter a field that values work over innate talent because women value hard work more than innate talent. (Incidentally, they tested this last hypothesis, and found that, while it is true that women, on average, are more likely to perceive hard work as the most important aspect of success, and the reverse is true for men, this doesn’t explain the results of the study. That is, while women do value hard work more than talent, they don’t seem to be choosing their fields of study based on a similar value system existing within that field.
So, what do I think? Well, I think some of this makes sense. We know stereotype threat is real, and we know that the stereotype that women are somehow less likely to be “brilliant” than men is real. Those stereotypes are well-documented. It’s nice to see that the Mommy Hypothesis, which makes me want to throw textbooks with great force, is unsupported by the data (again). I suspect that the explanation is more complex than just one factor, but that a combination of stereotype (this is what women are good at; men are more likely to be geniuses), field-specific ability beliefs, and the expectations imposed on kids when they are still very young (girls are bad at math, elementary school teachers are all women) is ultimately responsible for the gender gap.
How do we fix it? Stop with the stereotyping on ALL levels, for a start. What do you think? Suggestions/comments welcom