Wednesday, October 1, 2014

Worrying News from the Royal Society

The Royal Society is the preeminent scientific club in the United Kingdom, and it is arguably the oldest such society in existence. Each year, the Royal Society offers its University Research Fellowships. These are highly coveted awards! Applicants may be from a broad range of scientific fields (including the noblest of the physical sciences, astronomy), and it provides both salary and research support for 5 years (with the possibility to extend to 8 years).

The fellowship site states the goal clearly:

"This scheme is for outstanding scientists in the UK who are in the early stages of their research career and have the potential to become leaders in the field."

Importantly, the fellowships have a key relationship as a feeder program to permanent research posts in the United Kingdom, including faculty spots.

The instructions for the current round indicate that they expect a success rate of 10%. What the instructions don't say is that this was true last year only if you are a man.

Indeed, when the 2014 University Research Fellowships were announced to some fanfare just over a week ago, a pattern was quickly spotted: Of the 43 awards, only 2 went to women. Yes, that's 4.6%. Any keep in mind that this is all fields of science, including many (notably the life sciences) for which the overall rates of participation by women generally far outstrip the lower rates seen in astronomy and physics.

Some of comments I have seen online run along the following lines: The Royal Society is mostly white men, and so its natural that they pick young fellows who are like them. Perhaps, but that doesn't explain why the rate of awards going to women has suddenly plummeted. I doubt the Royal Society has become suddenly more white and male than it was in 2013, when it awarded 17.1% of the fellowships to women. Or 2010, when fully 33.3% were won by women.

When Jayne Birkby, a NASA Sagan fellowship here at Harvard brought this to my attention, my first thought was: Where is the Royal Society leadership on this issue? Have they responded quickly, and are they investigating the matter?

While I remain tremendously disappointed in the Royal Society about the outcome, I was comforted somewhat to see that Sir Paul Nurse, the Society President, has responded quickly to present the data at the intermediate steps of the selection process, and to announce that he will investigate the matter. Dr. Nurse has shown good leadership in not just expressing concern, but pointing out the very negative message that this outcome sends to younger scientists in the UK (and worldwide). (Specifically, he wrote that it sends "a bad message to young female scientists". I am extending this to say it sends a bad message to all young scientists, not just the women.)

Let me reproduce the data from the selection process here:


There are some worrying long term trends here. First, the fraction of awards going to women has declined over the past 5 years, from 33% in 2010 to 4.7% in 2014. Second, the success rate of men has steadily increased, from 4.7% in 2010 to 12.8% in 2014.

Looking at a survival analysis for the 2014 selection process, it's clear that women faced a headwind, which grew in strength at the later stages: 23% (17/75) of female applicants advanced to the shortlist, compared to an advancement rate of 33% (109/321) for men. But while only 35% (6/17) of shortlisted women were selected for interviews, 57% (62/109) of men survived that step. Finally, while the large majority (66%; 41/62) of men who were selected for interviews ultimately received awards, only 33% (2/6) of the women interviewed would ultimately be successful. Said differently, only 12% of women who were shortlisted would ultimately receive this prized fellowship, whereas 38% of the men who were shortlisted (the first step of the process) would ultimately win one.

The motto of the Royal Society (as you can read below the smiling dogs, above) is Nullius in Verba,  which translates roughly as Take Nobody's Word For It. As Dr. Nurse and his designated officials investigate this matter, I hope they will keep this motto in mind. Perhaps they will be assured by those who ran the selection process that it was conducted in a fair and impartial manner. The data strongly indicate it was not: Gender was clearly a factor in a candidate's chance of success.