Whenever you stick your head above the parapet and express a strong opinion in a high-profile venue, it’s almost inevitable that you’ll receive a large number of emails in response. Most of the follow-up I receive tends to be from people in the same position who recognize and share the sentiments expressed, writing to share their gratitude. But – probably of more interest – a smaller percentage of respondents are those who disagree and want to tell me why. (An even tinier fraction are from a certain sort of person eager to explain their revolutionary ideas about quantum uncertainty, string theory or government conspiracies, but these we can classify under ‘entertainment’ and swiftly move on.)
Some of the naysayers make excellent points that make me think, and the best of these cause me to shift my position slightly to accommodate the new perspective. The most effective of these are those who have gathered thoughtful evidence, usually in the form of official surveys from reputable establishments such as Ipsos MORI, or peer-reviewed data.
But these people are in the minority. Usually what I get is anecdotal evidence.
It is, I suspect, a very strong tendency of human nature to generalize one’s own personal experience into a broader worldview. Such an approach is, after all, entirely natural – we filter our whole universe through personal sensation and, though we can be influenced by outside opinion and evidence, the most intimate and immediate impressions are those we gather ourselves. If those two perspectives happen to clash, it’s understandable that many would prefer to place more value on their own. Indeed, part of basic scientific training is to teach people to rally against these natural tendencies in favor of objective, measurable outcomes.
A few months back I wrote a piece for Nature about women scientists being under-represented as pundits, in which I speculated that this might be related to detailed empirical observations I’d made over 20 years of conference attendance: in a gender-balanced audience, far fewer women tended to ask questions at lectures. (As an example, I’ve recently returned from the British Societies for Cell Biology and Developmental Biology annual joint meeting in Canterbury, and of the first eight talks I attended [seven male speakers and one female], out of a total of 24 questions, only 3 were posed by females; the audience was roughly 50:50.)
Anecdotal evidence is especially insidious when it derives from highly isolated environments that buck the trend. For example, one response I received was from a man from a British institution who wrote that, because more than fifty per cent of professors in his institute were women, there was clearly no problem with women in science. Leaving aside the fact that my piece was not discussing the professional representation of female scientists, it’s a perplexing extrapolation for someone to make: perhaps he was unaware of the stats overall. But when I pointed out that only about 9% of UK professorships in science and mathematics were held by women, it didn’t change his opinion. In this case, hard data could not sway personal experience – not even that of a scientist who should know better.
Another response I received was from a woman – let’s call her Mary – who was frustrated by the fact that no one was willing to admit that the real story about gender skewing in science was down to the fact that women were not as mentally capable of doing science as men. We’ll leave aside for the moment the fact that (a) my piece had nothing to do with the gender basis of scientific aptitude, and (b) I was speaking about a phenomenon that occurred in a gender-balanced population – the life sciences, where numbers are close to equal as far as the pool available to give opinions. In this case, Mary’s opinions had been formed by talking to other like-minded women, but (it seemed to me) more significantly on the basis of the view of her father, a highly respected engineer who used to teach post-grad trainees in a leading industrial equipment company.
According to Mary, her father ultimately came to the conclusion that, despite their degree in engineering, the women he was charged with training just didn’t have the right sort of brain for designing equipment. So this view helped shape Mary’s conviction that women are underrepresented in engineering not because they are intimidated, but because they simply lack the basic mental skills. In her email, she implied that the phenomenon held true in the scientific research sphere as well as in engineering.
And here’s where both the scientist and novelist in me started to feel intensely curious about this man and his group of female apprentices. I began trying to fill in the gaps with the limited information to hand. Of course one possible explanation for his observations was that women in general just aren’t as apt at design of industrial equipment – or that this group of women in particular were not. But it wasn’t difficult to hypothesize other scenarios. I was guessing that these subjects were children in the 1950s, plus or minus a decade. What was the upbringing of these women like? Were they given Lego, toy trucks and Scalextrics to play with, to help hone any innate aptitude during their formative years, or were they rather showered with dolls and tea sets? What was the parental reaction if any of the girls tried to play with their brother’s toys? When they expressed an interest in becoming engineers, did their parents and teachers give them support, or did they react badly? During their studies, were they taken as seriously as their male counterparts and given the same opportunities; did they have the same level of confidence as a result?
That’s just the cohort of women to consider. What about the male trainers of women trying to be engineers in that era, such as Mary’s father? Were they a big believer that women belonged in the field? Did they do their best to make these women feel welcome? Did they treat their female students as seriously and respectfully as the male students? Or did they have a private opinion that the girls just weren’t up for the job at hand, and did this become a self-fulfilling prophecy? It is well documented (see here for a nice summary) that a phenomenon called ‘stereotype threat’ can cause women to do worse on math exams than men if they’re reminded beforehand that they’re not supposed to be as good at it – an effect which disappears when they are not prepped. Also, there are plenty of peer-reviewed studies (as always, elegantly outlined in Virginia Valian’s book Why So Slow) showing that women and men with otherwise equal qualifications and talents are not perceived to be equal; were engineer trainers of the day susceptible to that extremely common and unfortunate bias? And of course, this was one company in one physical location, many years ago: is it really possible to extrapolate to the entire, present-day world on the basis of this man’s subjective experience, no matter how respectable? And finally, most importantly, can we extrapolate from engineering to every other scientific disciple to conclude, as Mary had, that women are just not cut out to be scientists and should instead focus their efforts on endeavors more appropriate to their innate skills, such as nursing?
For me, as a scientist – female or otherwise – , the answer is obviously ‘of course not’. We don’t have enough information to draw a reliable conclusion one way or the other about the women under her father’s care. The best we can do is look to the peer-reviewed literature, which despite many attempts over the years, to my knowledge still fails to show that the innate differences in scientific aptitude between the sexes is larger than the variation within them (see for example summaries by Spelke and Hyde – please do send me any contrary studies, as I’m genuinely interested).
I could, of course, counter her arguments with anecdotal evidence of my own. In the 21 years since starting my PhD, I have never noticed a gender-based difference in scientific aptitude among the many hundreds of scientific colleagues I’ve encountered in labs around the world. I’ve never noticed that the male students I’ve supervised have had a better grasp, or the female professors who’ve mentored me have had an inferior one. Of the thousands of scientific papers I’ve devoured over these two decades, there has been no tendency for the best ones to be authored by men, or the weak ones by women. Sitting through thousands of research talks and group meetings, I haven’t noticed that the men on average produce better research data or come up with more insightful experimental designs or hypotheses than their female counterparts. All I have noticed is variation from individual to individual: some people are brilliant scientists, some are weak, and the rest fall into a long continuum in between.
I could counter her arguments this way. But I won’t – because it’s not scientific.
But what, then, is the best way to help people think beyond the merely anecdotal ? Do we reply with statistics and citations and hope to sway minds that way? Can better education that emphasizes critical thinking and objective measurements help to counter our natural tendencies to believe what we think we see and to prize this view above all others? Or will personal experience always reign supreme?
Note added later: I should note that anecdotal, or personal experience, is a powerful persuasive tool and I don’t advocate avoiding it for rhetorical purposes. Ironically, my op-eds in Nature have contained such items, largely because the editor encourages their use (and references are not normal for that format).