OVER the past few decades, while all fields of science have increased the proportion of women at faculty level, physics has remained the most unequal, with women accounting for only around 10 per cent of faculty members. Is this because women are innately incapable of succeeding in the field (and/or innately prefer other careers), or because they are being discriminated against at more junior levels?
I was curious to know, so in 2004, while I was working as a physicist based at , I surveyed researchers there. Conference presentations are key to the career advancement of young physicists, as they provide exposure to potential employers, so one of my questions asked how many such presentations the respondent had been allocated over the preceding five years. A particle physicist cannot give a presentation unless selected to do so by the administrators of their experiment 鈥� a decision generally taken in a closed-door meeting of senior collaborators.
Among the 300 physicists who responded, women reported being allocated significantly fewer conference presentations than their male peers. This disparity, which was particularly marked among young graduate students and postdocs, was so great that I almost didn鈥檛 believe it, so I cross-checked the result using public databases kept by . These include records of all conference presentations, as well as all internal publications produced by collaboration members. Out of 48 men and nine women, I found that female postdocs were indeed allocated significantly fewer conference presentations than their male peers, even though they published significantly more papers. Based on productivity, the women were only allocated one-third as many conference presentations as the men.
Advertisement
鈥淭he disparity was so great that I almost didn鈥檛 believe it, so I cross-checked the result鈥�
I do not think the bias in conference allocations was intentional. It鈥檚 more likely that lack of oversight in the system allowed unconscious bias to creep in. This could be easily fixed by making the allocation process more transparent and democratic.
In 2006 I lodged a complaint with Fermilab, in the hope that such changes would be made. Fermilab said it had no legal obligation to investigate, so in 2007 I complained to the Department of Energy, which funds the lab. In April this year it had still not decided whether to investigate.
Frustrated by the lack of response, I on the arxiv.org physics preprint server, and this was picked up in a in Nature. Since then I have received many messages of support, several of them from female particle physicists who have noticed similar gender biases at their own experiments. One had even complained to the administration of her experiment at Brookhaven National Laboratory in New York, the week before my paper was released.
Discouragingly, I have also had many responses that were not concerned with my finding, but focused on attacking either my analysis or me. One disturbingly common criticism was that publications are not a good measure of productivity because women might choose to publish more often than men. Not only does this imply that the publication record of a woman is somehow worth less than a comparable record of a man, it also ignores my finding that women receive fewer conference allocations even when productivity is not taken into account.
Another common criticism is that my sample size was too small to prove gender discrimination. As with any statistical finding, it is never possible to prove gender discrimination, but we can make statements about the probability that gender equity is evident. The disparities I found were so extreme that the probability that men and women were being allocated conference presentations equally was less than 1 per cent.
Other attacks have been more personal, even vicious. I have been accused of falsifying my data, called a lunatic, and one critic (apparently a male chemistry professor at a US university) promised on an online chat board to present 鈥渁ll the dirt [he could] drag up鈥� on me. No scientific or statistical analysis I have ever performed has engendered such an extreme response.
Perhaps that鈥檚 why studies like this are so rare. One of the few comparable investigations was in 1997 by Christine Wenneras and Agnes Wold of Gothenburg University, Sweden (vol 387, p 341). They looked at how scientists鈥� work was rated when they applied for grants from Sweden鈥檚 Medical Research Council, and found that women had to be 2.5 times as productive as men to get the same peer review ratings. More than 10 years on, critics are now .
Since the release of my study, Fermilab has arranged meetings to ask female staff if they are concerned about bias. This is a step in the right direction, but it cannot be a substitute for transparent conference allocation procedures. Unless unconscious gender bias is acknowledged and tackled across physics, efforts to attract young women into the field will be in vain.
Careers 鈥� Find out how to make the most of your career in our comprehensive special report.