Many years ago I was struggling at Maths O-grade. My parents (bless them) paid a maths graduate from the university for a couple of hours tuition once a week, for two or three months before I took the exam, and his careful explanations helped a lot. I’d concluded I wasn’t good at maths. He asked me, towards the end of our last session, to add the numbers 1 to 100 together and give him the answer. “In my head?” I asked him. “Any way you like,” he told me. I had pen and paper and a calculator to hand. My first thought was to start adding 1+2+3+4… and then I thought, no, if I add 1+100 that’s 101, if I add 2+99 that’s 101, there will be 50 such additions, so the answer’s 5050. And I told him. It took me about a minute. He smiled, and he told me this story about Carl Friedrich Gauss.
A teacher had given the class some busywork to do – just that problem, add together the numbers one to a hundred. The teacher expected this to occupy the class for quite a while as they added 1+2+3+4+5… but the boy Gauss thought about the problem for a few minutes and got the answer. It is a well-known story among mathmos, but not one I’d ever heard, and I’d never been presented with that problem before.
In 1989 the Barbie Liberation Organization was formed. In 1992 they carried out the best customer action against sexist toys ever:
Taking advantage of similarities in the voice hardware of Teen Talk Barbie and the Talking Duke G.I. Joe doll, er, “action figure,” they [bought] several hundred of each and performed a stereotype-change operation on the lot.
The surgery was no simple matter — circuit boards had to be trimmed, a capacitor moved, and a switch re-engineered. [The BLO published a step-by-step explanation of their method, but it's no longer available online. I have a copy, which I'll publish if anyone's interested.]
The BLO returned the altered dolls to the toy store shelves, who then resold them to children who had to invent scenarios for Barbies who yelled “Vengeance is mine!” and G.I. Joes who daydreamed “Let’s plan our dream wedding!”
Each Teen Talk Barbie (or GI Joe) could say four phrases from a set of 270, making it mathematically unlikely that any two Barbies would say exactly the same set of phrases. (The odds are rather worse than one in five billion, assuming that selection was truly random and that there was some kind of protection to ensure Barbies always said four different phrases.) But the odds that any one Teen Talk Barbie would say one specific phrase are much smaller: out of every five dozen Barbies sold, at least one of them would say “Math class is tough!” (often misquoted as Math is hard!). In October 1992, Mattel announced it was removing that phrase from the set – Barbies would now say any four of 269 phrases – and any child who still had a Barbie that said “Math class is tough!” could send her back to Mattel and get a new Barbie that would say “Physics is fun!” instead. (Kidding.)
The perception that ‘boys do maths and science, and girls do humanities and arts’ is one that most people would scoff at today, but research into the options chosen by boys and girls at 13 suggests that gender-stereotyping still exists in schools and is narrowing the career options of thousands of teenagers.
Worryingly, recent tests have shown that that teenage girls in Britain are lagging further behind boys in science than anywhere else in the Western World. A study of 57 nations by the Organisation for Economic Co-operation and Development (OECD) analysed the results of tests taken by around 20 million 15-year-olds. It found that boys in Britain were 10 points ahead of girls in science, a bigger margin than most of the participating countries, apart from Indonesia and Chile. (TomBrown.com, 10th June 2009)
Many studies of why there is a gender gap in science, mathematics, and engineering, have appeared to show that the stereotype threat – girls aren’t good at maths because girls and their teachers and their parents and the bous all know that girls aren’t good at maths – is responsible for this gap.
David Geary (Curators Professor of Psychological Sciences in the University of Missouri College of Arts and Science) and Gijsbert Stoet (Institute of Psychological Sciences, University of Leeds) conducted a review that cast doubt on the accuracy of this theory, showing that numerous studies had major methodological flaws and utilized improper statistical techniques, and that many studies had no scientific evidence of this stereotype:
“We were surprised the researchers did not subject males to the same experimental manipulations as female participants,” Geary said. “It is reasonable to think that men also would not do well if told ‘men normally do worse on this test’ right before they take the test. When we adjusted the findings based on this and other statistical factors, we found little to no significant stereotype theory effect.”
The researchers believe that basing interventions on the stereotype threat is actually doing more harm than good, as vital resources are being dedicated to a problem that does not exist.
“These findings really irritate me, as a psychologist, because this is a science where we are really trying to discover what the issues are,” Geary said. “The fact is there are still a disproportionate number of men in top levels of science, technology, engineering and mathematics. We need more women to succeed in these fields for our economy and for our future.”
The study, “Can stereotype threat explain the sex gap in mathematics performance and achievement?” will be published in the journal Review of General Psychology. (MU News Bureau, 18th January 2012)
Another study from 2007, in the journal Psychological Science, shows something very simple: women graduating in maths, science, and engineering have risen, but women aren’t getting senior posts at university faculties, for two sets of reasons: research has revealed a bias against women, where evaluators rate
resumes and journal articles lower on average for women than men
and women still do more family work than men do. Getting a high-level faculty position requires putting long hours at work, which in turn requires either having no family, or having a live-in housekeeper/childminder who will do the necessary work on your behalf. Faculty promotion is based directly on the presumption that an ambitious man with the will to rise in his profession will never be in want of a wife.
And Mary Murphy (Stanford University) demonstrated that women are less likely to participate in settings in which they are outnumbered by men. She ran an experiment in which participants with majors in maths, science or engineering watched two videos disguised as advertisements for a Math/Science/Engineering summer leadership conference.
While identical in content, the seven-minute videos showed around 150 people in either an unbalanced gender ratio (3 men to 1 woman) or a balanced ratio of 1 to 1.
While watching the videos, students were equipped with body sensors that measured their physiological responses, including heart rate, skin temperature and level of sweating.
Female students showed faster heart rates and more sweating while watching the gender-unbalanced video compared with the gender-balanced video. They also reported a lower sense of belonging and less desire to participate in the conference when gender was skewed toward men.
Helen Powell, a freelance chemical consultant, a Stemnet speaker (Stemnet promotes careers by getting people who work in jobs from biologists to builders to talk to schoolchildren about what they do) wrote in the Guardian in response to the Missouri/Leeds research, that she agreed with it:
But encouraging schoolgirls into university and careers is not all. As is typical in most sectors, I see a number of female engineers at the entry- and mid-levels of companies, but precious few at the top. I have seen too many well-qualified women who fail to make it to top positions due to poor company culture and the lack of flexible working to allow returning mothers to the workplace. This is a huge waste of talent. The report has done well to challenge the myths behind women’s underachievement in schools, but more work still needs to be done to address the problem of women’s lack of achievement in the workplace.
A movement has begun in British politics to oppose panels at political conventions which are all-male – more at the F-word. How many organisers of science, maths, and engineering conferences work to ensure they have no all-male panels? How many arrange for there to be a creche and childcare facilities available at the conference, paid for as part of the membership fee?
If those conference organisers are sure that they base their invitations to speakers purely on merit, I’d invite them to take a look at the evidence that orchestras started to hire more women after the practice of having the audition take place with the musicians playing behind a screen – so that selection had to take place exclusively on musical ability, with gender bias eliminated.
Using a screen to avoid gender bias increased by 50 percent the probability that women would advance out of preliminary rounds, and if used in the final round, increases “by severalfold, the likelihood that a female contestant will be the winner in the final round”:
Cecilia Rouse, assistant professor of economics and public affairs, used personnel records and rosters from several symphony orchestras to track the hiring of women musicians as orchestras adopted the practice of “blind” auditions during the 1970s and 1980s. While the use of screens in the final round is still uncommon, Rouse and Harvard Professor Claudia Goldin report that the use of screens in preliminary rounds, now a widespread practice, accounts for much of the success of women in winning orchestra slots in the past 25 years.
“The switch to blind auditions can explain between 30 percent and 55 percent of the increase in the proportion female among new hires and between 25 percent and 46 percent of the increase in the percentage female in the orchestras from 1970 to 1996,” the economists write. The study notes that the surge of women in symphony orchestras has occured despite the fact that the number of positions is highly fixed and turnover is slow. (Princeton University, 25th April 1997)
It makes no sense, if you are auditioning musicians for an orchestra, not to choose the best players. Yet the evidence shows that the auditions panels were doing just that – selecting musicians that weren’t as good, simply because they were men.
There are solid reasons why women may feel discomfort with male-dominated environments. See Schrödinger’s Rapist and the Open Source Women Back Each Other Up Project. One reason was illustrated extensively last year in the atheist/skeptic community, when a speaker from the US, invited to speak at an atheist conference in Dublin, got propositioned in a lift at 3am by a conference participant. One creepy guy in a lift may not seem much of a problem, but the issue was proved to be a much wider problem when the speaker said mildly and briefly in a vlog about the conference “Guys – don’t do that” and was hit with a storm of condemnation and abuse from a wide range of people, even including Richard Dawkins. (More discussion at Pandagon and Pharyngula. Link roundup here.)
[Update: I'd just posted this when the building controversy about UniLad popped up on my timeline. (The link goes to freewebsitereport, not to the UniLad website itself, as I agree with the consensus not to give them the hits.) UniLad are a website and a magazine circulating on UK university campuses, purporting to "help lads get laid": they recently published an article celebrating how easy it is to get away with rape, they sell t-shirts with "rape jokes", and they react to criticism with homophobic/misogynistic abuse. Stavvers dissects their behaviour and reaction in an open letter:
Your reaction to the controversy suggests that deep down you have some semblance of a clue that what you are doing is wrong, and here’s why: what you are doing is reinforcing a culture which facilitates rape.
How prevalent is the attitude among male students that rape is some kind of joke - or worse, a crime easy to get away with?]
With all of that, it’s possible that at the extreme edge of maths-nerdiness, there are more men there than women.
Geary and Stoat make no contention about the gender gap itself. Their study makes a strong case for ruling out a self esteem-based explanation of the gender gap, but an increasing number of scientists believe the gender gap is illusory in the first place.
Recent years have brought mounting evidence against the idea that, other things being equal, women are worse at math than men. A 2011 study published in Psychological Bulletin found evidence of gender gaps in various countries, but noted that in some countries, such as Jordan and Bahrain, girls had the edge. (Huffington Post, 19th January 2012)
But Stephen J. Ceci and Wendy M. Williams (Cornell) demonstrated that the gender imbalance in universities far outweighs any ability imbalance, even at the far end of the scale.
In other words, there’s a higher percentage of math-nerdy women than the percentage who are employed in math-nerdy positions. “When you look at the cognitive and biological data, they’re not enough to explain the difference,” says Ceci.
Women graduate with baccalaureate degrees in math at the same rate as men, and since the mid ’90s more than a quarter of the math PhDs have gone to women. If they’re good enough to get that far, they’re pretty damn amazing at math. But at the next level — tenure-track associate professor — the proportion of females shrinks to single digits. (Time.com, 28th October 2010)
The reason for this, they too say, is perfectly clear: the system for advancement in universities presumes that you will not spend any time at all having and raising children. Men have the option of having a family and having someone else raise your children: women do not. Rebecca Goldin a tenured math professor at George Mason University, and a mother of four, comments:
“When having children and trying to be serious about mathematics, you can feel like your entire intellect is being judged,” she says, “and that if having children disrupts your publication or teaching efforts, you are a failure.” Among her friends, women have left mathematics because they felt marginalized, not because they didn’t like math.
There may be some truth to the stereotype theory, but the theory itself may be damaging girls’ abilities in class:
In one experiment, when a class of girls and boys were told they were being tested at math to see if the stereotype held true, boys generally outperformed girls. Yet, in groups told before the test that studies showed the math gender stereotype was almost never significant, girls and boys scored about the same. (The Glass Hammer, 1st June 2011)
And there’s a consistent problem with British media consistently presenting academic success at school for girls as a “worrying trend”:
Just under a fifth – 19.8% – of boys have achieved the two top grades, creating a record gap between the genders of 6.7 percentage points. Last year, the gap was six percentage points, with 25.5% of girls achieving the top grades, compared to 19.5% of boys.
The gender gap has also widened at C grade and above, with girls 7.5 percentage points ahead, compared to 7.2 percentage points last year.
The same story notes a small increase in more boys than girls getting slightly higher grades in maths as a success story:
Boys have beaten girls at GCSE maths for the third year in a row, following a decision to drop coursework in the subject. The proportion of boys getting A* to C in maths has risen to 58.9% from 58.6% last year. The proportion of girls achieving at least a C has increased from 58.3% to 58.6%.
However, girls are significantly ahead in English. Some 72.5% of entries from girls has achieved a C grade or better, compared to just 58.7% of entries from boys. Guardian, 25th August 2011
An article from 1999 points out that this is not a new problem:
Although GCSE results suggest that boys have fallen further behind in Britain over the past decade—a change which has coincided with the increased emphasis on continuous assessment—the evidence suggests that boys have always done worse than girls to some degree. It is a difference that was once tolerated, even catered for.
For example, the old 11-plus examinations, which sorted students into secondary schools, were skewed deliberately: boys, considered less mature than the opposite sex, were not required to perform to the higher standard set for girls. Only in this way could the selective grammar schools enrol a balance of boys and girls. A 1923 Board of Education report dealt with boys’ under-performance simply by noting their habit of “healthy idleness”. (The Economist, 27th May 1999)
The 11-plus, for anyone who doesn’t remember it, was intended to sort out the smarter kids whose parents could not afford to send them to a private school. Pupils who passed the exam at age 11 went to selective state-funded grammar schools. There were many problems with the 11-plus and the officially-meritocratic grammar school system, but a huge one was never discussed: if the exam results had been marked gender-blind, and pupils passed on to grammar school solely on the basis of how well they did, grammar schools would consistently have been about 70/30 girls/boys. This was considered unacceptable, so the education authorities simply set up a system whereby girls had to do better than boys to get into grammar school – a girl and a boy could have got identical marks, and the girl would have gone to an inferior school.
Via hundreds of City Learning Centres in England and Wales, since 2005, 135,000 girls have been supported in their interest in ICT by Computer Clubs for Girls (CC4G). The Education Secretary, Michael Gove, has recently announced cuts to the budget for City Learning Centres.
And as for supporting women in science, maths, and engineering by childcare:
Parents spend almost a third of their incomes on childcare – more than anywhere else in the world, according to a study by Save the Children and the Daycare Trust. For four out of 10 families the cost of childcare is on a par with mortgage or rent payments, the study showed. … parents, regardless of income, cannot afford not to work but struggle to pay for childcare, and despite many parents cutting back their spending almost a quarter are in debt because of childcare costs. (Guardian, 7 September 2011)
Gove is planning to scrap the current ICT curriculum for schools in England and Wales, subject to a consultation launched on 18th January. While the computer industry appears optimistic about the input big firms like HP and IBM can now have on the education curriculum:
Computer games entrepreneur Ian Livingstone, an adviser to Gove, envisages a new curriculum that could see 16-year-olds creating their own apps for smartphones and 18-year-olds able to write their own simple programming language. (Computing – 11th January 2012)
the educationalists are less so:
However, although Gove announced what he was intending to scrap, he did not announce any intention to replace them with anything, beyond confirming that ICT would remain a compulsory subject. Instead, he simply stated that “schools will be able to offer a more creative and challenging curriculum, drawing on support and advice from those best positioned to judge what an ambitious and forward-looking curriculum should contain.” (UKAuthorITy, 13th January 2012)
The Scottish government is also considering changes to the ICT curriculum (The Future of GLOW) following comments made by Eric Schmidt of Google at the MacTaggart Lecture at the MediaGuardian Edinburgh International Television Festival, last year, on Computing in UK schools:
“I am flabbergasted to learn that today computer science isn’t even taught as standard in UK schools. Your IT curriculum focuses on teaching how to use software, but gives no insight into how it’s made. Not only that, we don’t teach kids how computers work and we don’t teach them how to create software for themselves.”
There’s an e-petition at the UK government website on beginning to teach computing to younger children: “we are losing the female coders and we need this generation to help us code a better country”. Worth signing.
It’s easy to see the appeal of the self-esteem story. The graduate student who told me I’d figured out the Gauss solution in my head wanted me to know I wasn’t bad at maths. But I never joined the computer club at school, because it was always boys-only, and even when taking my degree in Computing, where by alphabetical accident I was in a tutorial group of 7 men and 4 women (the whole year in that course was 80-20 men:women) the men were far more likely to speak up in the tutorial group or answer questions in the lectures than the women. Still, one-on-one it never seemed to me that the women taking the degree had any less self-esteem than most of the men.
Promoting the idea that “girls need more self-esteem” is cheap and sets the blame squarely on girls and their parents and educators.
But the scientific research actually disproves that view. Instead the chief factors would seem to be disproportionate numbers of men at events making them look unwelcoming to women; unconscious sexist selection bias at all sorts of levels where the candidate’s gender is known; and, overwhelmingly, the plain fact that the structure in academia and in the computer industry particularly is geared to prefer for promotion people who will work long hours for their employer without regard for any family responsibilities.
For all Michael Gove’s speeches in Parliament on improving ICT education, apparently in direct response to the MacTaggart lecture, it seems that if governments and institutions want more girls in maths, engineering, and science, they need to invest more effort in ensuring that universities, conferences, journals, and even panels at conferences are gender-balanced, and they need to invest more money in ensuring that childcare and other personal responsibilities do not prevent women from developing their careers.
Anyone care to calculate the odds on either one?
“Vengeance is mine!” said GI Barbie.