How grades are set for GCSEs, AS and A levels
Once they have marked the work of most students, exam boards set grade boundaries - overall mark thresholds for achieving a particular grade - according to our rules.
The ‘comparable outcomes’ approach
The basic principle that exam boards follow when setting grade boundaries is that if the group of students taking a qualification in one year (the ‘cohort’) has a mix of abilities that is similar to the previous year’s group, then the overall results (outcomes) should be comparable.
This also means that if a particular year’s cohort is - overall - more (or less) able than the previous cohort, then results should be higher (or lower).
For this approach to work, exam boards need to understand how the mix of abilities in the current year’s cohort compares to previous cohorts. They must use a range of evidence to do this, including:
- predictions based on prior attainment/achievement (for AS and A level, exam boards use predictions based on the prior attainment at GCSE; for GCSE, they use predictions based on the prior achievement at key stage 2)
- students’ work (marked exam papers and controlled or non-exam assessment) from the current and previous years
- reports from senior examiners and moderators about how the exam questions worked
How exam boards and Ofqual use predictions
Predictions based on prior attainment/achievement are not used to predict the grades for individual students.
Instead, they are used to predict the overall (national) proportions of grades that should be expected for the current year’s cohort.
Before any results are issued, exam boards share the outcomes of their awards with Ofqual. Exam boards must provide evidence to justify any circumstances where the expected results are markedly different from the predictions. We consider that evidence according to a published process. We will either accept the explanation provided by exam boards or challenge the results if the argument is not backed by sufficient evidence.
This also means we can judge whether the grade standards are in line across exam boards.
If necessary, we can require the exam board to change its grade boundaries if proposed results are not supported by the evidence.
The impact of new qualifications
When new qualifications are introduced, it is usual to see a dip in performance - even if the overall mix of abilities has not changed - as schools, colleges and students get used to new content, or new styles of assessment.
To deal with this, we require exam boards to place particular emphasis on predictions in the first year of awarding a new GCSE, AS or A level. This helps make sure students are not disadvantaged by being the first to sit a new qualification.
Variability in a school’s or college’s results
While national results may remain steady from one year to the next, we know that schools and colleges can see variability in their year-on-year results, even when qualifications don’t change.
When qualifications do change, we know there is the potential for more variability in individual school results. This is because individual teachers and schools will respond differently to the changes, which increases year-on-year variation at school level. We have published information on this in recent years (for example, see ‘Variability in GCSE results: 2012 to 2015’ and our June 2017 letter to schools)
Separate reporting of non-exam assessment results
For some reformed qualifications, students’ performance in their non-exam assessment is reported in a grade or outcome that is separate to that for the exam assessments for the qualification. This is the case for:
- the spoken language assessment in GCSE English language
- the practical assessment in A level science qualifications