Based on their abilities, more girls could choose careers in science. But they don't, even though failing to do so means missing out on lucrative and prestigious careers with a high level of job security. So why do these talented girls still not go for those study programmes and what can we do about it? Girls still choose scientific study programmes less often than boys, even when they took a scientific profile at secondary school and in spite of the fact that they are better at maths. One reason for these gender differences in study choices might be the influence of friends at secondary school, but not much is known about this. What we do know is that friends play an extremely important role in the social environment of adolescents and influence many other behaviours and choices. Research We researched if and how friends affect the study choices of secondary school pupils. In our research, we asked 744 pupils from 174 classes in senior general secondary education (HAVO) and university preparatory education (VWO) to fill in questionnaires about various aspects of their lives. At that time, the pupils were in their third year at secondary school (15 years old). They were asked who their five best friends in their class were and since the whole class took part in the research, this gave us information about their circles of friends. The same pupils took part in the research again after they had left secondary school (17 - 19 years old). This allowed us to evaluate which field of study they have entered. With these data, we were able to study whether friends in secondary education affect a student´s field of study choice after secondary education. Our research shows that gender differences in study choices are partly caused by the social expectations of friends. Young people, just as everyone else, have certain ideas about what 'appropriate behaviour' is for boys and girls. These normative ideas are also known as gender role expectations. Even in a modern country like the Netherlands, these expectations are still fully present and influential. Traditional gender role expectations dictate, for example, that scientific study programmes are for boys and not for girls. Young people behave accordingly, out of fear of being rejected. If their circle of friends has stronger ideas about how men and women should behave, there is a greater likelihood of adolescents behaving according to these gender role expectations. In other words, this means that girls who have friends with stronger gender role expectations will not choose a scientific study programme after secondary school. In our research, we found that girls are discouraged from deciding on a scientific study by the traditional gender role expectations of friends. This is consistent with research that shows that girls are more heavily influenced by their social environment than boys. So talented girls ignore a potential career path because those around them feel it's wrong for them. We also looked at whether it made a difference whether young people were friends with boys or girls. We know that boys are more often friends with boys, and girls with girls. Our question was: is there more pressure to conform to gender role expectations if someone is only friends with people of the same gender? It might be that with only girls around you, it's even more important to behave 'like a girl'. Or for boys, 'like a boy'. However, the effect of gender role expectations was not dependent on the composition of the circle of friends. One remarkable finding is that when boys have more friends of the same gender, they more often choose a scientific study. But this was not because they felt more pressure to conform to gender role expectations. An alternative explanation might be that boys share scientific activities and interests and therefore are more likely to choose a scientific study. For example, if boys often play computer games together, it's a plausible assumption that they encourage each other, even unconsciously, to study IT. One final important finding is that the gender role expectations of friends are more important than those of the whole class or of a person's own gender role expectations. This emphasises the importance of the role friends play in girls choosing, or not choosing, scientific study programmes. How can we reduce the effect of these gender role expectations? Girls don't choose scientific studies after secondary school because their friends feel it's not something for them. One way of making sure more girls opt for scientific careers is by breaking down normative expectations about what is 'correct' male or female behaviour. This won't be easy. Society is still permeated with stereotypical expectations about male and female behaviour. This even starts before children are born when we decide on the colour of the baby's room. But there are things we can do. One initiative that has been found to be effective is bringing girls in contact with women in technical careers through, for example, lectures, speed dates or 'student for a day' visits. It´s important to have role models. It also helps if female maths teachers or female teachers of technical subjects encourage girls to choose scientific studies. Our findings show that this contact needs to take place early on in secondary school, because gender role expectations already influence field of study choices at age 15. Another option is to better inform pupils at school about the content, diversity and possibilities of a scientific study and potential career. Scientific studies still suffer from a nerdy or stuffy image, even though there is so much diversity in scientific professionsNot to mention that science fields are rapidly expanding and every day new occupations are being . In short, the diversity of professions is growing - now we have to make sure that diversity among future professionals grows too. Original text (in Dutch): Maaike van der Vleuten, assistant professor of Sociology at Radboud University. Image by Priscilla Du Preez through Unsplash. This article was previously published on Van twaalf tot achttien. It is based on the scientific article 'Gender norms and STEM'.