Last Updated on October 17, 2026 by Karl Thompson
How useful are official statistics for understanding differences in educational achievement by social class, gender and ethnicity? This post sets out what the government’s GCSE figures show, then evaluates them as a source of data. It is written for the Methods in Context question on AQA Paper 1 (7192/1), and for 10-mark questions on official statistics.
For how official statistics work as a method in general, start with Official Statistics in Sociology. This post applies that material to education.
How do GCSE results vary by gender, ethnicity and social class?
The figures below come from two government sources:
- the Department for Education’s Key Stage 4 performance statistics, which report GCSE results in England. The latest release covers 2025/26 and was published on 15 October 2026. Its figures are provisional: DfE revises them in the following year after schools have checked their data
- the government’s Ethnicity facts and figures website, which breaks results down by detailed ethnic group. Its latest figures are for 2022/23
Notice that both sources cover England, not the whole UK. Scotland, Wales and Northern Ireland run their own exam systems and publish their own statistics.
The main measure used below is the percentage of pupils getting a grade 5 or above in both English and maths (a “strong pass”). The other headline measure is Attainment 8, a pupil’s average points score across eight GCSE subjects.
Gender: girls still outperform boys, but the gap has halved
Girls outperform boys on every headline measure. HOWEVER, the gap is much smaller than it was before the pandemic.
| Measure (gap between girls and boys) | 2018/19 | 2024/25 | 2025/26 (provisional) |
|---|---|---|---|
| Grade 5 or above in English and maths | 6.6 percentage points | 3.1 percentage points | 3.1 percentage points |
| Attainment 8 | – | 4.0 points | 3.9 points |
| Entered for the EBacc | – | 7.8 percentage points | – |
Source: DfE, Key Stage 4 performance 2025/26 (provisional) and 2024/25 (revised).
In 2024/25 the gap on English and maths narrowed because girls’ results fell by 1.3 points while boys’ rose by 0.3. That matters for interpretation: a narrowing gap is not the same thing as boys catching up.
Ethnicity: Chinese and Indian pupils achieve the highest results
| Ethnic group | Grade 5 or above in English and maths, 2022/23 |
|---|---|
| Chinese | 78.3% |
| Indian | 70.4% |
| Bangladeshi | 57.1% |
| White Irish | 54.6% |
| Black African | 50.2% |
| Pakistani | 46.1% |
| Mixed (all groups) | 45.5% |
| All pupils | 45.3% |
| White British | 42.8% |
| Black Caribbean | 31.2% |
| Mixed White and Black Caribbean | 30.5% |
| Irish Traveller | 17.5% |
| Gypsy/Roma | 7.7% |
Source: Ethnicity facts and figures, GCSE English and maths results (2022/23 data, updated November 2024).
Chinese pupils are the highest-achieving group, with more than three-quarters getting a strong pass in both subjects. Indian pupils come second.
Black Caribbean and Mixed White and Black Caribbean pupils have the lowest results of the larger ethnic groups, at around 31%.
Gypsy/Roma and Irish Traveller pupils have the lowest results of all. Fewer than one in ten Gypsy/Roma pupils got a strong pass in both subjects. For why, see Why do Gypsy, Roma and Traveller children have such low educational achievement?
The DfE’s 2025/26 release only reports broad ethnic groups. On English and maths, Black pupils improved most on the year before (up 2.3 percentage points), followed by Asian pupils (up 2.1), with White pupils up 0.4.
Social class: disadvantaged pupils are far behind
The DfE does not measure social class. Instead it uses a category called disadvantaged pupils, which is the closest thing it has to a measure of class, but isn’t the same thing.
Pupils count as disadvantaged if they have been “eligible for free school meals at any point in the past six years”. Pupils who have been in local authority care for at least one day, or adopted from care, are also included.
| Measure (gap between disadvantaged and all other pupils) | 2018/19 | 2024/25 | 2025/26 (provisional) |
|---|---|---|---|
| Grade 5 or above in English and maths | 25.2 percentage points | 27.3 percentage points | 26.8 percentage points |
| Attainment 8 | – | 15.5 points | 15.1 points |
| Disadvantage gap index | – | 3.92 | 3.82 |
Source: DfE, Key Stage 4 performance 2025/26 (provisional) and 2024/25 (revised).
A gap of around 27 percentage points is the difference between roughly one in four pupils and roughly one in two. Put simply, disadvantaged pupils are about half as likely to get a strong pass in English and maths as other pupils.
The disadvantage gap index is the DfE’s own summary of the gap, using pupils’ rankings in English and maths. It peaked at 3.94 in 2023 and has fallen since. HOWEVER, the gap on English and maths is still wider than it was before the pandemic.
For more on class and achievement, see Social class and educational achievement statistics.
Strengths of official statistics on educational achievement
1. Good validity, as far as it goes
These figures aren’t schools’ own claims about how well their pupils did. They come from external examinations, set and marked by the exam boards, so they should give a reasonably true picture of how pupils actually performed in their GCSEs.
HOWEVER, we need to be cautious about this. In 2020 and 2021 exams were cancelled during the pandemic and grades were assessed by teachers, and DfE itself warns that changes in results across 2020 to 2022 mainly reflect the way grades were awarded rather than changes in standards.
2. Excellent representativeness
We get information on practically every pupil in the country, even the ones who fail! No sociologist could ever collect a sample this large themselves.
One thing to check: many headline figures are reported for state-funded schools only, so pupils in independent schools may not be in the table you’re looking at.
3. Easy comparisons between social groups
The data allow us to compare achievement by social class, gender and ethnicity, and to spot trends over time. They also let us look at two factors at once. Ethnicity facts and figures reports that in every ethnic group, pupils eligible for free school meals did worse than those who weren’t (2022/23, grade 4 or above). Yet the gaps between ethnic groups are large too, so poverty alone can’t explain the pattern.
4. Freely available
They are free to anyone with an internet connection. The DfE’s Explore Education Statistics service even lets you build your own tables, so a researcher can save a lot of time and money.
5. They help the government target policy
They allow the government to track achievement and to design social policies aimed at the groups most likely to underachieve. The disadvantage gap index, for example, exists so that progress on closing the gap can be tracked year on year.
Limitations of official statistics on educational achievement
1. They are confusing to non-specialists
If you look at the DfE’s Key Stage 4 statistics, you’ll probably find them bewildering. There are so many different measurements that they obscure the headline question of who got those two crucial GCSEs.
Attainment 8 and the disadvantage gap index are especially unclear to anyone other than a teacher: all you get is a number, which means little to most people.
Some measures disappear altogether. Progress 8, which compares pupils’ GCSE results with their results at the end of primary school, cannot be calculated for 2024/25 or 2025/26. These pupils never sat Key Stage 2 tests, which were cancelled in 2020 and 2021.
2. Changes to the measures make comparisons over time difficult
Until 2015, the headline measure was five good GCSEs (A* to C) including English and maths. Attainment 8 replaced it in 2016. From 2017 GCSEs moved from letter grades to numbers (9 to 1), and the government now focuses on English and maths at grade 5, Attainment 8 and EBacc entry.
On top of that come the pandemic years, and the fact that the latest figures are provisional and will be revised. So a sociologist comparing 2015 with 2026 isn’t comparing like with like.
3. They don’t measure social class
These statistics don’t actually tell us about the relationship between social class and educational achievement. Instead of a sociological definition of class, the government uses the narrow measure of free school meal eligibility, which is an indicator of material deprivation rather than social class in its fuller sense.
Marxist sociologists would argue that this is ideological: the government simply isn’t interested in measuring the effects of social class on achievement, and if you don’t measure it, the problem kind of disappears.
4. They don’t explain why the differences exist
This is almost certainly the biggest limitation. These statistics don’t tell us anything about why these variations exist.
They allow us to formulate hypotheses. HOWEVER, if we’re being objective, they don’t show us why disadvantaged pupils are half as likely to get a strong pass in English and maths. To find out, we need further research, usually qualitative methods such as interviews or observation in schools. That is the interpretivist case against relying on official statistics.
How this comes up in the exam
On Paper 1, official statistics can come up in Section B as the method in a 20-mark Methods in Context question, applied to an education topic in the item. The approach is covered in How to Answer Methods in Context Questions.
They also come up as 10-mark “outline and explain” questions on Papers 1 and 3. For a worked example, see Outline and explain two reasons why some sociologists choose to use official statistics in their research.
The figures in this post are also useful evidence in education essays on class, gender and ethnicity. Quote the year with any figure you use.
Signposting
This post has been written primarily for students studying research methods in context, as part of the AQA’s A-level Sociology course.
Official Statistics in Sociology covers the general strengths and limitations of official statistics, which this post applies to education.
Assessing the Usefulness of Secondary Qualitative Data to Research Education looks at the other kind of secondary data on schools: documents such as Ofsted reports and school prospectuses.
How does Educational Achievement Vary by Ethnicity? and Explaining the Gender Gap in Education explain the patterns shown in the tables above.
Please click here to return to the Research Methods page, or here to return to the homepage – ReviseSociology.com.
Yess yess lads lads lads and everybody! Yo sociology dudez are epic as! I would like someone to translate this! Yo to you lottt
Pingback: Assessing the Usefulness of Secondary Qualitative Data to Research Education | ReviseSociology