Updated 26.06.25. to include guidance on Tiering.
It's the time of year when teachers ask questions on various fora about how to interpret assessment data and how to respond to management, pupil and parental expectations regarding predicted grades.
Pasted below is the advice shared by David Blow, (Executive Head of South-East Surrey Schools Education Trust) at the ALL London June Event 2025. A download of the document is available here. A version including how to deal with Tiering is available here.
Teachers and HoDs are likely to be being asked to give grades for the Year 10 students taking the new GCSE in French, German and Spanish. This aims to give some practical suggestions on how to do so.
which includes the following key points:
If an exam is easier than in previous years, the grade boundaries for that paper will be higher. If it is harder, the grade boundaries will be lower.
The difficulty of exam questions varies year to year, even though exam boards try to keep the level of demand consistent. That’s because it is impossible to determine how difficult students will find a paper until it is taken.
This is why new grade boundaries are set each year – to reflect the difficulty of that particular paper, and to ensure that it is no easier or harder to get a grade in any given year.
Grade boundaries are decided after students take exams and when marking is nearly complete
https://ofqual.blog.gov.uk/2017/03/10/comparable-outcomes-and-new-a-levels/
When qualifications change, we follow the principle of comparable outcomes – this means that if the national cohort for a subject is similar (in terms of past performance) to last year, then results should also be similar at a national level in that subject. So exam boards will control for the impact of the changes such that this year’s cohort is not unfairly disadvantaged. They will be relying heavily on the statistical evidence to do this, but also using senior examiners to check the grade boundaries that the statistics are pointing to.
which draws on the official Ofqual advice
Our additional advice
The above gives a clear explanation about why grade boundaries are not set in advance, and how the exam boards will determine what they are after the papers have been taken in June 2026. However, that does leave a real challenge for teachers in determining how to interpret “the standard itself has not changed, even though the assessments have.” because it is not just the assessment which has changed, there has been a significant change to the subject content, in particular with a restricted word list and a focus on greater mastery of a more specified content.
Our advice would be to assume that the ability and likely outcomes of the pupils will be similar to previous years, and to use a similar distribution of grades to that obtained by Year 11 in Jun 2024 and previous years. Pupils can be ranked on the assessments they have taken, and then allocated a grade using the likely distribution. It can also be very helpful to look at the distribution English and Maths grades obtained by ML students in the actual exams in 2023 and 2024 and what was “predicted” in Year 10), and cross-reference to the English and Maths grades being given to the current Year 10 ML students. It may well be worth asking your data manager to help with this process. Note that there was a slight increase in grading in 2023 and 2024 arising from Ofqual’s work on standards, but it is small and will be subsumed in the wider variations.
Worked example
Step 1: convert the distribution of grades in 2023 and 2024 to percentages, and average. So, ave. of 10% of students gained a grade 9.
Tiering
Tiering is going to be an additional complication and ALL, ISMLA, ASCL and others are liaising with Ofqual and the exam boards about providing advice to teachers, and we are looking to the Autumn Term for this.
In the meantime, Year 10 pupils in the same school may have sat different papers depending on their ability level e.g. “Foundation” and “Higher”
Follow the same procedure for Steps 1 to 2 for the combined group – this gives you an estimated final distribution once you have combined the grades for “Foundation” and “Higher” together.
There will be a degree of iteration to this process, and do ask your data manager or another colleague for assistance.
Now Steps 3 to 5 separately for “Foundation” and “Higher” so you will get two distributions of grades. Combine them together and compare with your outcome from Steps 1 and 2. Apply Step 6 to both. Make adjustments to each distribution, and compare at student level using your local knowledge and the grades in other subjects for the crossover students. When you are happy with both the separate distributions and combined distribution, use the allocated grades
This is a useful opportunity to remind people of the severe grading in GCSE ML which is a historic anomaly going back to O-level and first identified 50 years ago!
https://ffteducationdatalab.org.uk/2025/05/grading-severity-at-key-stage-4-in-2024/
