Quick fixes on stereotypes won’t mean more female scientists

时间:2019-02-28 10:07:02166网络整理admin

Eric Raptosh Photography/Getty By Simon Oxenham Women make up just 12.8 per cent of the UK’s workforce in science, technology, engineering and maths, according to one recent analysis. But what’s behind this gender gap in the STEM subjects? A popular explanation is the idea that a person conforms to a perceived stereotype about themselves – something called stereotype threat. For example, girls performing less well at maths because they have heard that boys are better at the subject. The effect of such stereotypes may then go on to affect subject choices and career paths. This idea has become one of the most studied theories in social psychology, and has been tested in hundreds of experiments. But the latest results suggest the consequences of stereotyping by race or gender are less clear than we previously thought. The term stereotype-threat was coined by social psychologists Claude Steele and Joshua Aronson in 1995 in an experiment that looked at race. They gave the same test to black and to white students, but told some of these that the test was of intellectual ability, while telling the others that it was a problem-solving task that gave no indication of intelligence. After adjusting the results to account for variability between each person’s ability, they found that the black students’ scored lower when they were told that the task was a test of their intellectual ability. There was no difference in white students. In 1990, a whopping 60 per cent of people in the US still thought white people were more intelligent than black people. According to Steele and Aronson, such attitudes cause black people to doubt their own abilities, and this leads to poorer academic performance. Their groundbreaking finding appeared to vapourise the effects of centuries of prejudice: tell a black student that a test is not judging their intelligence, and they will get a better result. What’s more, in a follow-up experiment, the researchers found that simply having to specify their race before taking the test was enough to drastically worsen a black student’s performance, while when white students did this, their performance actually increased. These experiments shocked researchers in the fields of social psychology and education. Four years later, Steele produced similar findings while working with Steven Spencer and Diane Quinn. This time, they looked at undergraduate psychology students taking a maths test. They found that when participants were told beforehand that there tended to be a gender difference in performances on this particular test, the women scored less well, and the men performed better. This effect evaporated when women were told that there were no gender differences before taking the test. Subsequent research has found that this kind of stereotype threat can be neutralised by telling people that a maths test is just a test of problem solving. Women also do better at maths tests if they are told that the test is designed to identify students with strong ability, rather than pinpoint those who are not very good. Since then, the theory has expanded to include everything from women’s athletic ability to men’s social sensitivity. Stereotype threat doesn’t just seem to apply to race and gender – it seems to extend to socioeconomic background too. The research has led to widespread efforts to educate children about stereotype threat and research has shown that teaching undergraduate students about this effect appears to improve their performance in a maths test. But all this was called into question in 2012 when researchers examined attempts to replicate Steele’s gender results. Of the studies they examined, only 55 per cent got the same results as Steele and his colleagues. Across all these studies – positive and negative – half of them had statistical problems. But of the experiments that were statistically sound, only 30 per cent replicated the original finding. A year later three studies involving nearly a thousand school-aged girls failed to find any evidence that their mathematical performances were impacted by stereotype threat. Then in 2014, a meta-analysis of performance of girls in various stereotyped domains found that none of the factors previously identified as things that can make the effect of stereotype threat stronger, such as the presence of boys in the room, had any effect. The researchers also reported that “there were several signs for the presence of publication bias” in a way that might “seriously distort the literature”. In essence, studies that showed stereotype effects have reported results that statistically seem too good to be true. We would expect to see more studies announcing failed attempts to replicate these findings, but we don’t – suggesting that there is a publication bias, and positive results are more likely to be published by a scientific journal. All this has made the supposed effects of stereotyping appear clearer and stronger than they really are. Earlier this year, researchers published a large study in which they assessed 590 women in the US for stereotype-threat. This was a far greater number than participated in the original experiments by Steele and his colleagues, which involved between 20 and 30 participants per experiment. The study founds no statistically significant effects. This could be for a number of reasons – perhaps because not all the people taking the test were students – but the debate rumbles on. All this suggests that if stereotype threat is a real phenomenon, it is one that is hard to reliably demonstrate under carefully controlled conditions. It is probably not something that can be altered meaningfully by a throwaway remark here or there – such as an encouraging statement before taking a test – but rather is something that builds up over a lifetime of hearing that you are not as good as the dominant group. This means that changing testing conditions is unlikely to make enough of a difference to the number of girls who go on to work in science, engineering or maths. It seems that more obvious factors are behind the STEM gender gap. These include outright sexism, as well as working environments that discourage female participation To bring about lasting change, we need to do more than just trying to stop people from conforming to inaccurate stereotypes. We need to smash those stereotypes and thereby take away the ammunition they give to sexism. A good step might be to teach children about the large contributions that women have made to these fields – achievements that all too often don’t make it into school textbooks. Campaigns in the UK like WISE and Ada Lovelace Day try to make up for this, but they may catch girls when it is too late to make a difference. Read more: Exploding the myth that boys are better at maths More on these topics: