Last Updated on November 7, 2026 by Karl Thompson
Boys and girls still choose very different subjects once they have a choice. Girls are far less likely than boys to take physics, computing and engineering, and far more likely to take subjects such as psychology, sociology and English literature. Sociologists explain the difference in five main ways:
- gender domains: children learn early which tasks belong to their gender
- early socialisation in the family, through toys and activities
- peer pressure, which polices gender-appropriate choices
- teachers, careers advice and subject counselling, which steer pupils towards stereotyped choices
- gendered subject images: subjects come to be seen as masculine or feminine.
This post sets out what the patterns look like in the most recent figures, then each explanation in turn, then the evaluation. Gender and subject choice is one of the three main questions in the gender and education topic, alongside achievement and gender identity, in the sociology of education module (AQA Paper 1, 7192/1).
Gender and subject choice: the patterns in 2026
The most important point about the statistics is when the differences appear. Where a subject is compulsory, or nearly everyone takes it, entries are close to 50:50. The gap opens as soon as pupils get to choose.
GCSE
At GCSE, maths and the sciences are taken by nearly everyone, so entries are almost evenly split. Option subjects are a different story.
| GCSE subject (2026) | Entries from girls |
|---|---|
| Mathematics | 50.2% |
| Biology | 49.2% |
| Chemistry | 48.7% |
| Physics | 48.1% |
| Design and technology | 28.7% |
| Computing | 22.8% |
| Engineering | 14.8% |
Source: EngineeringUK analysis of JCQ GCSE results, summer 2026.
A-level
At A-level every subject is chosen, and the gaps widen. Physics is the clearest example: girls make up nearly half of GCSE physics entries, but only about a quarter of A-level physics entries.
| A-level subject (2026) | Entries from girls |
|---|---|
| Mathematics | 37% |
| Design and technology | 33% |
| Further mathematics | 26% |
| Physics | 24% |
| Computing | 19% |
Sources: EngineeringUK analysis of A-level results, August 2026; BCS analysis of JCQ figures for computing (18.9% across England, Wales and Northern Ireland).
The pattern is slowly changing. In 2017, girls were outnumbered almost ten to one in A-level computing. By 2026 they made up nearly one in five computing entries: still a small minority, but roughly double the share of ten years earlier.
University
The pattern carries on into higher education, even though women now make up the majority of students. In 2024/25 women were 55% of full-time undergraduates in the UK, but:
| Degree subject area (UK, 2024/25) | Students who were women | 2014/15 |
|---|---|---|
| Education and teaching | 88% | 84% |
| Physical sciences | 46% | 37% |
| Mathematical sciences | 33% | 38% |
| Engineering and technology | 21% | 16% |
| Computing | 21% | 15% |
Source: HESA data for 2024/25, as analysed for the Higher Education Policy Institute (May 2026).
Women’s share of physical sciences, engineering and computing has risen over ten years, but their share of maths has fallen, and education has become even more female. Subject choice is becoming less gendered in some subjects and more gendered in others, which is a good evaluation point against any claim that the old divides are simply disappearing.
Why it matters
Subject choice links education to inequality at work. The subjects boys are more likely to take, such as computing, engineering and economics, tend to lead to better-paid jobs than the caring and arts subjects girls are more likely to take. That is one of the main reasons feminists argue that girls’ higher exam results have not produced equality after school (see my post on the feminist perspective on education). It also explains part of the A-level top-grade pattern: boys are concentrated in maths and further maths, which award a high proportion of A and A* grades (see explaining the gender gap in education).
Gender domains
A gender domain is a task or activity that children come to see as belonging to their own gender or to the other one. The concept comes from Naima Browne and Carol Ross (1991), who studied how young children talk about and play with toys and activities.
Their argument is that children arrive at school already knowing which activities are “girls’ stuff” and which are “boys’ stuff”, having learned it mainly from the family and early experiences. Children are more confident with a task they see as inside their own domain, and approach one outside it with anxiety. This is about perceived territory, not actual ability: a girl who sees physics as boys’ territory is less likely to choose it, less likely to persist when she finds it hard, and more likely to expect to do badly.
Gender domains are the best concept to use when you’re explaining subject choice, because they link early socialisation (what children learn at home) to later choices (what they pick at 14 and 16). How pupils construct and perform gender identity in school more generally is covered on my post on gender identity and education.
Early socialisation
Gender socialisation begins in the family. Ann Oakley’s concept of canalisation describes how parents channel children towards toys and activities seen as appropriate for their sex: dolls and toy kitchens for girls, construction sets and vehicles for boys. Oakley’s full framework is covered on my post on gender socialisation in schools.
The link to subject choice is direct. A girl steered towards dolls and caring play is being prepared, without anyone intending it, for health and social care and childcare courses; a boy steered towards tools and building sets is being prepared for engineering and design and technology.
Socialisation may also affect confidence and attitudes to competition. Laboratory experiments by economists have found gender differences in how people respond to competition: men’s performance tends to rise more under competition (Gneezy, Niederle and Rustichini, 2003), and men are more likely than women to choose to compete (Niederle and Vesterlund, 2007). If subjects like maths and physics are seen as competitive, with “one right answer” and clear rankings, differences in confidence could influence who chooses them.
Peer pressure
Once pupils reach secondary school, their peer group becomes a powerful influence. Choosing a subject seen as belonging to the other gender carries a social risk. A boy who chooses dance, or a girl who chooses engineering, may be teased, and verbal abuse (such as calling a boy “gay” for choosing a “girls’ subject”) is one way peer groups enforce gender norms.
Peer group pressure is one of the factors AQA’s own mark scheme credits for June 2017’s question on subject choice: it “leads to girls and boys choosing subjects which reflect stereotypical gender identities”. The way peer groups police gender identity is covered in more detail on my post on gender identity and education.
Teachers, careers advice and subject counselling
Interactionists argue that teachers and careers advisers can steer pupils towards stereotyped choices, often without meaning to. AQA’s mark scheme for June 2017 credits both careers advice (for example, girls being more likely to choose childcare courses because childcare jobs are seen as female) and subject counselling (teachers channelling boys and girls into different subjects in line with their own gender stereotypes).
Economists have tested this using large datasets. Victor Lavy and Edith Sand (2018), studying pupils in Israel, found that primary school teachers’ gender biases raised boys’ later achievement and lowered girls’, and made boys more likely and girls less likely to enrol in advanced maths courses in high school. Studies by Lavy and Megalokonomou (2019) and Terrier (2020) also found that teachers’ biases affect pupils’ outcomes.
Do teachers stereotype pupils according to gender?
A survey carried out for Accenture in 2016 asked young people, parents and teachers in the UK and Ireland about girls and STEM subjects. It found that:
- 57% of teachers and 52% of parents admitted making subconscious stereotypes about girls and boys in relation to STEM subjects
- 54% of teachers said they had seen girls drop STEM subjects because of pressure from parents
- 32% of young people thought more boys chose STEM subjects because they matched “male” careers or jobs.
The sample was large (8,644 people, including 400 teachers and 1,000 parents), but the survey was carried out online for a technology company with an interest in the result. It wasn’t a random sample, so we can’t be sure how representative it is of all teachers. And it measures what teachers say about their stereotypes, not how they actually behave in the classroom.
Gendered subject images
A gendered subject image is the idea that a subject is masculine or feminine, built up over time from who teaches it, how it is taught and who takes it. Once a subject has a gendered image, it becomes harder for pupils of the other gender to choose it, because doing so means stepping outside their gender domain.
The classic study is Alison Kelly’s “The construction of masculine science” (1985), which examined how science comes to be seen as a male domain. Kelly set out four different explanations of how this happens, and argued that they overlap rather than compete. Her conclusion matters for policy: because the masculine image of science is constructed, schools could help change gender identities rather than simply reproducing them.
Subject images are built up in several ways, all of which AQA’s mark scheme recognises:
- Learning resources: textbooks and examples that represent science as mainly a male activity (using cars and engines, for example, to illustrate physics).
- Who teaches the subject: if science teachers are mainly men, boys may see science as theirs.
- Who takes the subject: once a class is mostly boys, a girl choosing it is in the minority, which makes the subject look even more male to the next year group.
The physics figures show subject images at work. Girls take GCSE physics in roughly equal numbers to boys, but only about a quarter of A-level physics entries are from girls. Ability doesn’t explain that drop; the subject’s image is a much better explanation.
Single-sex schools
It is sometimes argued that girls in single-sex schools are more likely to choose “boys’ subjects”, because the subject images and peer pressures of mixed classrooms are weaker. The evidence on whether single-sex schooling affects results is weak: Skelton, Francis and Valkanova (2007) concluded that once socio-economic status and prior attainment are taken into account, there are no significant differences in achievement between pupils in single-sex and mixed classes or schools. That finding is about results, not choices, and is a reminder that any apparent single-sex effect has to be separated from the effect of pupils’ backgrounds.
Evaluating explanations of gendered subject choice
No single factor explains the pattern. Skelton, Francis and Valkanova (2007) concluded that pupils often choose subjects according to their own ideas of what suits their gender, and that gender stereotyping, or differing ideas about gender among pupils and teachers, runs through all of the explanations above.
| Explanation | Strengths | Limitations |
|---|---|---|
| Gender domains and early socialisation | Explains why the gap appears as soon as choice opens up, and why it’s hard to change at 16 | If socialisation were the main cause, gaps should narrow everywhere as attitudes change, yet some (maths at degree level) are widening |
| Peer pressure | Credited by AQA; fits the social risk pupils describe of choosing “the wrong” subject | Hard to measure, and pupils may overstate or understate it in interviews |
| Teachers and careers advice | Lavy and Sand show teacher bias affecting later choices in large datasets | Much of the strongest evidence is from other countries, and surveys like Accenture’s rely on self-report |
| Gendered subject images | Explains the drop in girls’ physics entries between GCSE and A-level | Images change over time (computing and engineering degrees are slowly becoming less male), so they are not fixed |
Three further evaluation points:
Girls’ choices are not simply “wrong”. Treating girls’ choices as a problem to be fixed assumes the subjects boys choose are the ones that matter. Feminists would reply that the issue is that the subjects girls choose lead to lower-paid jobs, which says as much about how society values caring work as about girls’ choices.
Choice is shaped by class and ethnicity too. Cavaglia, Machin, McNally and Ruiz-Valenzuela (2021) found that boys in England are far more likely than girls to take STEM subjects in post-16 vocational education and at university, despite girls’ higher achievement overall. Gender interacts with the route a student takes, and vocational routes are strongly gendered.
Things are changing, slowly and unevenly. Computing and engineering degrees are less male than they were ten years ago, but maths degrees are more male. Any answer that says subject choice is “still exactly as gendered as ever” or “no longer gendered” is wrong in both directions.
How this comes up in the exam
Subject choice has been examined directly as a 6-mark question:
June 2017, Q02: Outline three ways in which factors within schools may shape gender differences in subject choice. [6 marks]
You can see a real student answer to this question, as marked by AQA’s examiners, with AQA’s mark scheme, on my guide to the 4 and 6 mark outline questions. Note that this question asks for factors within schools: careers advice, subject counselling, learning resources and teacher gender would all score, but socialisation in the family would not.
Subject choice also turns up inside other questions. Item A for June 2024’s question on the problems girls may experience at school opens with “Pupils may be able to choose what subjects to study at GCSE and beyond”, inviting an answer on subject steering ([LINK NEEDED: June 2024 Q03 post]). It also appeared in the AQA specimen paper’s 30-mark question, which asked students to evaluate the view that gender differences in both subject choice and achievement are mainly the result of the influence of wider society.
For practice in the 10-mark format, see my practice question on analysing two reasons for gender differences in subject choice.
Sociology of Education Revision Bundle
Revise Education properly
Notes at two depths, so you can start early and finish fast — mid-depth notes to learn and consolidate from, and one-page summary sheets for the last fortnight.
Written by Karl Thompson, who taught A-level sociology for eighteen years and spent twelve as an AQA examiner, marking all three papers on the current specification.
Signposting and Related Posts
This post has been written primarily for students studying the education topic, as part of the AQA’s A-Level Sociology course.
Subject choice is one of three questions in the gender and education topic. The other two are achievement, covered on explaining the gender gap in education and in-school factors, and identity, covered on gender identity and education.
For the history of policies designed to get more girls into science, see education policy and gender.
More on this topic on my main sociology of education page.
Please click here to return to the homepage – ReviseSociology.com
Sources
- EngineeringUK (August 2026) A level and Scottish Highers results 2026 and Level 2 results: GCSE, VTQ and National 5 (2026), from JCQ data
- BCS, The Chartered Institute for IT (2026), analysis of A-level computer science entries from JCQ data
- Joint Council for Qualifications, A-level results, summer 2026
- Higher Education Policy Institute (16 May 2026) Subject choice shapes gender balance across UK universities, from HESA data
- EB News (29 August 2017) ‘IT exam results show widening gender gap’
- AQA, A-level Sociology 7192/1 Paper 1, June 2017, mark scheme. AQA material is reproduced for educational purposes
- Browne, N. and Ross, C. (1991) ‘”Girls’ stuff, boys’ stuff”: young children talking and playing’, in N. Browne (ed.) Science and Technology in the Early Years: An Equal Opportunities Approach, Open University Press
- Kelly, A. (1985) ‘The construction of masculine science’, British Journal of Sociology of Education, 6(2), 133–154
- Lavy, V. and Sand, E. (2018) ‘On the origins of gender gaps in human capital: short- and long-term consequences of teachers’ biases’, Journal of Public Economics
- Gneezy, U., Niederle, M. and Rustichini, A. (2003) ‘Performance in competitive environments: gender differences’, Quarterly Journal of Economics; Niederle, M. and Vesterlund, L. (2007) ‘Do women shy away from competition? Do men compete too much?’, Quarterly Journal of Economics
- Skelton, C., Francis, B. and Valkanova, Y. (2007) Breaking Down the Stereotypes: Gender and Achievement in Schools, Equal Opportunities Commission, Working Paper 59
- Cavaglia, C., Machin, S., McNally, S. and Ruiz-Valenzuela, J. (2021) ‘Gender, achievement, and subject choice in English education’, Oxford Review of Economic Policy, 36(4), 816–835
- Accenture (2017) survey of 8,644 young people, parents and teachers in the UK and Ireland, carried out online in December 2016

