Delusions of Gender

by Cordelia Fine

  A striking demonstration of this was provided by Mary Murphy and her colleagues at Stanford University. Advanced maths, science and engineering (MSE) majors were asked to give their opinion on an advertising video ‘for an MSE summer leadership conference that Stanford was considering hosting the next summer’.8 Under the cover story that the researchers were also interested in physiological reactions to the video, heart rate and skin conductance were recorded, to give a measure of arousal. After watching the ad, the students were asked questions to assess how much they felt they would belong at such a conference, and how interested they were in attending. There were two, near-identical videos, depicting about 150 people. However, in one video the ratio of men to women approximated the actual gender ratio of MSE degrees: there were three men to every woman. In the second video, men and women were featured in equal numbers. Women who saw the gender-equal video responded very much like men, both physiologically and in their sense of belonging and interest in the conference. But for women who saw the more realistically imbalanced version, it was a very different experience. They became more aroused – an indicator of physiological vigilance. They expressed less interest in attending the conference when it was gender unbalanced. (Interestingly, so did men – although this was probably, one can’t help but think, for different reasons.) And although women and men who saw the gender-balanced video very strongly agreed that they belonged there, the conviction of this agreement among women who saw a gender imbalance was significantly lower. Under the naturalistic condition of male dominance, they were no longer so sure that they belonged.

  Being outnumbered by men is simply a fact of life for women in MSE domains – as is being exposed to gender stereotypes in advertising. At first, it’s not obvious why an advertisement depicting, say, a woman bouncing on her bed in rapture over a new acne product might serve as a psychic obstacle to women looking to enter masculine fields. However, images of women fretting over their appearance or in ecstasy over a brownie mix, although they have nothing to do with mathematical ability directly, nonetheless make gender stereotypes in general more accessible. Paul Davies and his colleagues showed either these or neutral commercials to women and men who were invested in doing well in maths. They were then given a GRE-like exam that had both maths and verbal problems. Men in both conditions, and women who had seen neutral ads, attempted more maths problems than verbal ones. But women who had seen the sexist ads showed exactly the opposite pattern, avoiding the maths questions. Their career aspirations were also influenced, with a flipping of occupational preferences, from those that require strong mathematical skills (like engineer, mathematician, computer scientist, physicist and so on) to those that depend more heavily on verbal abilities (such as author, linguist and journalist).9 Ads that trade in ditzy stereotypes of women also, Davies and colleagues found, reduce women’s interest in taking on a leadership role. Male and female university students were equally interested in leading a group – except for women exposed to the gender-stereotyped commercials, who were more likely to choose a nonleadership role instead.10

  Entrepreneurship is another male-dominated arena, and one in which the traits usually assumed to be vital for success – strong-willed, resolute, aggressive, risk-taking – have a decidedly male feel. Here, then, is another occupational niche to which women could easily be made to feel that they don’t belong. Female business school students were given one of two fabricated newspaper articles to read. One described entrepreneurs as creative, well-informed, steady and generous – and claimed that these qualities are shared equally between men and women. The other article, however, depicted the prototypical entrepreneur as aggressive, risk-taking and autonomous, all traits that belong firmly in the male stereotype. The women were then asked how interested they were in being self-employed, and owning a small or high-growth business. For women who scored low on a proactive measure (the tendency to ‘show initiative, identify opportunities, act on them, and persevere until they meet their objectives’) it made no difference which article they read. But what about the highly proactive women? As you might expect of these go-getting women, their interest in an entrepreneurial career was high but significantly reduced after reading the entrepreneurship-equals-male news article.11

  What psychological processes lie behind this turning away from masculine interests? One possibility is that, as we learned in an earlier chapter, when stereotypes of women become salient, women tend to incorporate those stereotypical traits into their current self-perception. They may then find it harder to imagine themselves as, say, a mechanical engineer. The belief that one will be able to fit in, to belong, may be more important than we realise – and may help to explain why some traditionally male occupations have been more readily entered by women than others.12 After all, the stereotype of a vet is not the same as that of an orthopaedic surgeon or a computer scientist, and these are different again from the stereotype of a builder or a lawyer. These different stereotypes may be more or less easily reconciled with a female identity. What, for example, springs to mind when you think of a computer scientist? A man, of course, but not just any man. You’re probably thinking of the sort of man who would not be an asset at a tea party. The sort of man who leaves a trail of soft-drink cans, junk-food wrappers, and tech magazines behind him as he makes his way to the sofa to watch Star Trek for the hundredth time. The sort of man whose pale complexion hints alarmingly of vitamin D deficiency. The sort of man, in short, who is a geek.

  Sapna Cheryan, a psychologist at Washington University, was interested in whether the geek image of computer science plays a role in putting off women. When she and her colleagues surveyed undergraduates about their interest in being a computer science major, they found, perhaps unsurprisingly given that computer science is male-dominated, that women were significantly less interested. Less obvious, however, was why they were less interested. Women felt that they were less similar to the typical computer science major. This influenced their sense that they belonged in computer science – again lower in women – and it was this lack of fit that drove their lack of interest in a computer science major.13

  However, an interest in Star Trek and an antisocial lifestyle may not, in fact, be unassailable correlates of talent in computer programming. Indeed, in its early days, computer programming was a job done principally by women and was regarded as an activity to which feminine talents were particularly well suited. ‘Programming requires lots of patience, persistence and a capacity for detail and those are traits that many girls have’ wrote one author of a career guide to computer programming in 1967.14 Women made many significant contributions to computer science development and, as one expert puts it, ‘[t]oday’s achievements in software are built on the shoulders of the first pioneering women programmers.’ 15 Cheryan suggests that ‘[i]t was not until the 1980s that individual heroes in computer science, such as Bill Gates and Steve Jobs came to the scene, and the term “geek” became associated with being technically minded. Movies such as Revenge of the Nerds and Real Genius, released during these years, crystallized the image of the “computer geek” in the cultural consciousness.’16

  If it is the geeky stereotype that is so off-putting to women, then a little repackaging of the field might be an effective way of drawing more women in. Cheryan and her colleagues tested this very idea. They recruited undergraduates to participate in a study by the ‘Career Development Center regarding interest in technical jobs and internships.’ The students filled out a questionnaire about their interest in computer science in a small classroom within the William Gates Building (which, as you will have guessed, houses the computer science department). The room, however, was set up in one of two ways for the unsuspecting participant. In one condition, the décor was what we might call geek chic: a Star Trek poster, geeky comics, video game boxes, junk food, electronic equipment and technical books and magazines. The second arrangement was substantially less geeky: the poster was an art one, water bottles replaced the junk food, the magazine
s were general interest and the computer books were aimed at a more general level. In the geeky room, men considered themselves significantly more interested in computer science than did women. But when the geek factor was removed from the surroundings, women showed equal interest to men. It seemed that a greater sense of belonging brought about this positive change. Simply by altering the décor, Cheryan and colleagues were also able to increase women’s interest in, for example, joining a hypothetical Web-design company. The researchers note ‘the power of environments to signal to people whether or not they should enter a domain’, and suggest that changing the computer science environment ‘can therefore inspire those who previously had little or no interest … to express a new-found interest in it.’17

  You might think that this is a nice sentiment, but that a narrowly focused, unsociable personality simply goes hand-in-hand with talent in computer science. But as developmental psychologist Elizabeth Spelke and Ariel Grace point out, ‘personality traits that are typical of a given profession often are mistakenly thought to be necessary to the practice of the profession.’ They provide, as a historical example, the assumption by an early-twentieth-century psychologist that his talented Jewish students could not succeed in academia because they did not share the traits of the predominantly Christian faculty: he ‘mistakenly assumed that the typical mannerisms of his Harvard colleagues were necessary for success in science.’18

  Underscoring Spelke and Grace’s point is a fascinating natural experiment in the Carnegie-Mellon computer science department that suggests that geeky traits may indeed be extrinsic to success in computer science. In the mid- to late 1990s, an intensive study of male and (the very few) female computer science students at Carnegie-Mellon found that the men were very focused on programming – the sort of person who ‘dreams in code’ – while the few women in the programme were more interested in the applications of computer science. But in the late 1990s, the admission criteria were changed so as to no longer unnecessarily and unfairly exclude applicants without a lot of programming experience.19 This led to a fivefold increase in the number of women, from about 7 percent to 34 percent. Lenore Blum and Carol Frieze took the opportunity of this situation to interview the students who entered the computer science programme in 1998. In 2002, when they were interviewed, these students were, uniquely, the babies of the old, hacker-favouring admission criteria, yet were now in a department with a much more diverse student body. Remarkably, Blum and Frieze found that interest in programming versus applications was now a point of similarity, rather than difference, between men and women. ‘Almost all students saw programming as one part of their interests and the computers as a “tool” for their primary focus, which was applications.’ But also, there was evidence that the ‘students were constructing a new image’, and one in which the ‘narrowly focused computer science student’ was no longer the norm:

  Our cohort included students who played the violin, wrote fiction, sang in a rock band, participated in university team sports, enjoyed the arts, and were members of a wide range of campus organizations. We found that men and women alike appear to be moving towards a more well-rounded identity that embraced academic interests and a life outside of computing. Students described themselves as ‘individual and creative, just interesting all-round people’, ‘very intelligent, … very grounded, not the traditional geek …’, ‘much more well rounded than people five or six years ago.’

  Recall that these students had been chosen according to the old criteria. They were the geeky programmers. And yet, as the researchers suggest, the years spent in an increasingly gender-equal environment ‘had shaped their image of themselves. We might also speculate that such a transitional culture gave the men “permission” to explore their nongeeky characteristics’.20

  Both women and computer science are the losers when a geeky stereotype serves as an unnecessary gatekeeper to the profession. And recent work by psychologist Catherine Good and her colleagues shows that a ‘sense of belonging’ is also an important factor in women’s intention to continue in maths. This feeling of belonging, however, can be eroded by an environment that communicates that maths ability is a fixed trait and not something that hard work can increase, especially in combination with the message that women are naturally less talented than men, Good and colleagues found.21 Philosopher Sally Haslanger has suggested that a difficulty even today for women (and minority) philosophers is that ‘it is very hard to find a place in philosophy that isn’t actively hostile towards women and minorities, or at least assumes that a successful philosopher should look and act like a (traditional, white) man.’22

  But choosing a career is not just about finding a place socially in which one can feel at home. It also entails finding a fit with one’s talents. People of course tend to be drawn towards jobs in which they are likely to succeed. If gender stereotypes can affect people’s perceptions of their abilities (as we now know that they can), then it would not be surprising to discover that this then has effects on career decisions. Sociologist Shelley Correll has shown that beliefs about gender differences in ability have an important role to play in people’s perceptions of their own masculine abilities and, as you might expect, this affects their interest in careers that rely on such skills. Correll used the data from the 1988 National Educational Longitudinal Study, involving tens of thousands of high school students, to carefully compare students’ actual grades with their own assessments of their mathematical and verbal competence. She found that boys rated their maths skills higher than their equal female counterparts. This was likely due to the culturally shared belief that males are better at maths, because boys were selective in their self-embellishment: they didn’t inflate their verbal competence. These self-assessments proved to be an important factor in the students’ decision making about their careers. With actual ability (assessed by test scores) held equal, the higher a boy or girl rates his or her mathematical competence, the more likely it is that he or she will head down a path towards a career in science, maths or engineering. Correll concludes that ‘boys do not pursue mathematical activities at a higher rate than girls do because they are better at mathematics. They do so, at least partially, because they think they are better.’23 For example, gender differences in self-assessment of maths ability fully explained the gender gap in calculus enrolments.

  Correll then went on to show just how easy it is to create a gender stereotype that diminishes women’s confidence and interest in a supposedly male domain. She used a contrast sensitivity test, in which the participant has to guess which colour, black or white, covers a greater area in a series of rectangles. Her participants, freshmen at Cornell University, were told that ‘a national testing organisation developed the contrast sensitivity exam and that both graduate schools and Fortune 500 companies have expressed interest in using this exam as a screening device.’24 (In truth, the test is a fake one: black and white appear in essentially equal proportions, so there is actually no correct solution.) Participants were then told either that males, on average, perform better on tests of contrast sensitivity or that there is no gender difference.

  The participants were all given the same feedback on their test performance, but how this score was perceived depended on the context – male-advantage or gender-equal – in which the test was presented. When the students thought that contrast sensitivity was a nongendered ability, women and men’s self-assessments were very similar. But it was a different story when the underlying assumption was that one sex had the upper hand. In this male-favourable context, men rated their contrast sensitivity ability more highly and claimed to have done better on the tasks. They also set themselves a more lenient standard against which to judge their performance. Correll then investigated whether, as in her real-world data set, higher self-assessments would lead to higher aspirations. She found that they did. When men thought that they were, as a group, better at contrast sensitivity, they were more likely than women to say that they would enrol for courses or sem
inars based on the ability, and to apply for graduate programmes or high-paying jobs that relied heavily on the skill. And it was their higher self-assessments of ability that appeared to bring about this greater interest in contrast-sensitivity-based aspirations. We like what (we think) we are good at.

  But of course many women do persist in male-dominated careers like mathematics, despite the stereotype threat and lack of sense of belonging. Luckily for them, there is an alternative to turning away from maths – and this is to turn away from being female. Emily Pronin and her colleagues found that female undergraduates at Stanford University who had taken more than ten quantitative courses were less likely than other women to rate as important and applicable to them supposedly maths-incompatible behaviours such as wearing makeup, being emotional, and wanting children.25 The researchers then went on to provide evidence that it is not simply that women who like to wear lipstick and fondly imagine having children one day are intrinsically less interested in maths. Rather, women who want to succeed in these domains strategically shed these desires in response to reminders that maths is not for women. The researchers recruited a group of Stanford undergraduate women, for all of whom maths ability was important. Half of the women read a (fabricated) scientific article about ageing and verbal ability. But the remainder of the women read a shortened version of an actual scientific article about gender and maths, published in Science.26 This was a study of the Scholastic Aptitude Test results in maths for nearly 10,000 high-achieving seventh and eighth graders. Boys were more likely to score highly than girls, and the article concluded that there is ‘a substantial sex difference in mathematical reasoning ability in favour of boys’,27 together with the assertion that this advantage reflects boys’ innate superiority in spatial ability.