Fairness in awarding

This summer will be my twelfth at Ofqual. Each one has been different, but this one has been truly exceptional. However, over that time there has been one constant – fairness for students. This has always been at the heart of what we do at Ofqual.

In a normal summer, our work is focused on maintaining standards, so that it’s no easier to achieve, say, a GCSE grade 4 with one exam board than another.

We use a range of tools and sources of evidence to help us, including the National Reference Test (NRT) – an annual test taken by a sample of year 11 students and designed to measure small changes in students’ performance over time in GCSE English and maths.

We will use the results of the NRT again this year to determine whether students’ proficiency in these subjects has changed, and publish on GCSE results day our decision about whether to adjust grade standards.

But of course, this is not a normal summer, and so our work to make sure students are treated fairly this year is a little different.

Since the Secretary of State announced that exams in England would be cancelled this summer, we have worked with exam boards to develop an approach so that students will get the results they need to move on in their lives and have confidence in the grades they receive. Our aim has been to make sure those results are as fair as they can be for all groups of students, and that they reflect the grades students would have been most likely to achieve if teaching and learning had continued and they had taken their exams as normal.

The sector as a whole has risen to this challenge. In a little over 4 months, teachers and senior leaders have considered the evidence in their centres and submitted over 5 million centre assessment grades and rank orders, despite the practical challenges of lockdown.

Since it wasn’t possible to put in place a system to standardise teachers’ judgements before they made them, it is inevitable that there would be some variation in the judgments made in thousands of centres up and down the country.

We did, though, provide guidance to schools and colleges about how to make their judgements as objectively as possible, so as to be fair to all groups of students – and our initial analysis suggests that there has been no widening of the gap in attainment between groups of students with particular protected characteristics or from different socio-economic backgrounds.

If we did nothing to standardise between schools and colleges, that would not be fair to this year’s students, or to those who took exams last year or who will take them in 2021.

Having seen the data, we know that some centres have been more optimistic than others in generating centre assessment grades (CAGs), and the combined effect of that would be unprecedented increases in overall results. This is likely to undermine confidence in the grades issued, and means a student in a more optimistic centre will be unfairly advantaged over students in other centres.

Exam board technical teams have worked with us to develop and test various statistical models to standardise teachers’ judgements between centres, to make sure that the one we used is the fairest possible approach, and we’ve also had input from external statisticians and assessment experts.

Of course, a system of calculated grades and a statistical model can never know how an individual student might have performed on the day. Some students might have done better, or worse, if they’d had an opportunity to take their exam; we will never know and they will never know. Colleagues across higher and further education understand this, and many have committed to showing flexibility in their admissions decisions.

To try to take account of this limitation, we’ve made decisions which work in students’ favour where possible, and that has built some leniency into the way the statistical model worked. We also know that any statistical model will be unreliable with very small numbers. Where centres have entered a small number of students in a subject (either this year, or in the historical data), which is much more the case at A level than GCSE, we will rely much more on the CAGs than the statistics. There is more on how the model works here.

With over 5 million results being issued this summer, and even with the checking that exam boards are currently doing, it is possible that there will be a small number of errors. Without any exams, the usual reviews of marking – which assume marking has taken place – are not relevant this year but we have put in place an appeals system that we believe is as fair as it can be.

An individual student cannot make an appeal against the judgements made by their school or college – that is, they cannot ask for the school or college to think again. This is because any appeal would have to be done by someone better placed than teachers to judge a student’s likely grade; in the circumstances this year, we do not believe there is such a person. However they can ask their school to check what they submitted, to check for errors.

A school or college can appeal if they believe the historical data used for standardisation was not a reliable basis for predicting their 2020 results. This might include situations where there’s been a substantial change in the demographic make-up of the centre, perhaps if a single-sex school has changed to co-educational. Or where there is concern about the way the statistical model could affect individual high-ability students who might be expected to receive results that are out of line with the school or college’s historical results. In these cases, the best way to judge this will be by the school or college reviewing the results for the whole centre, and, if appropriate, submitting an appeal.

Alongside the results in August, exam boards will provide summary information so that centres can see what data was used in the standardisation model. This will help centres which think they might want to appeal.

We have published further material to help students and their families understand how appeals will operate this summer, including information to help students understand whether they might have reason to complain about bias or discrimination. While we believe such cases will be rare, we recognise this is important for the confidence of those students affected, and of students in general, in the arrangements this year.

Exams, like thousands of other important life events, have been cancelled this year, but the vast majority of students will still get results this August and will be able to move on to the next stage in their lives. In the absence of exams, we believe these results will be as fair as they possibly can be.

We understand readers may have specific questions or concerns about their particular situation. While it isn’t possible to respond to individual questions via our blog, you can contact our Public Enquiries Team on 0300 303 3344 or via email: [email protected]