Skip to main content

Blog The Ofqual blog


Why 9 to 1?

Posted by: , Posted on: - Categories: A levels and GCSEs

The new GCSE grade scale has been the subject of much debate in recent months. Why did we change it? That’s the question we’re being asked a lot this summer, as we await the first of the new grades.

To answer it, we have to go all the way back to 2013.

In February that year, the Secretary of State wrote to Ofqual to set out the purpose of reformed GCSEs – primarily to provide evidence of students’ achievement against more demanding and fulfilling content. The new qualifications were also to be linear, so that all exams would be taken at the end of the course. The Secretary of State also asked Ofqual to consider whether a new grade scale would be required, to make clear that these were different GCSEs.

Changing the grading scale is a big deal. Government is responsible for deciding what is included in the curriculum, but Ofqual is responsible for grading. We were asked to consider a new grade scale, and we took that very seriously.

We wanted to gather evidence. We looked at what other countries did (a variety of different things) and the strengths and weaknesses of the existing approach. We started from first principles, considering whether there were alternatives that might be more appropriate. And we considered the possible consequences of any changes we were proposing. We listened to experts in the exam boards and from around the world, and we spoke to teachers and headteachers.

In June 2013, we consulted on a number of design features for the new GCSEs, including tiering arrangements, whether to include non-exam assessment, and the grading scale. The consultation document is still available on our website if you want to take a look. We proposed a numbered grade scale from 8 to 1.

Responses were mixed, with some in favour of keeping the letter grades and others, including the exam boards, in favour of number grades. There was support for more grades above C, to better differentiate between the individual achievements of students, but not if that meant far less differentiation below C. That’s why we ended up with 9 grades, to increase the differentiation at the top end.

The June 2013 consultation also set out some of the other alternatives to grades that we considered.

We considered reporting marks rather than grades, but the disadvantage of marks is that they are not directly comparable from one year to the next. If a paper is more difficult one year, students will get lower marks compared to those in previous years. We also considered reporting standardised scaled scores, which deal with the issue of changes in difficulty between years. This would be similar to what is now used at Key Stage 2. But it is not easy to implement across multiple exam boards, so that the same score across different boards represents the same level of achievement.

We considered reporting a percentile score, but the disadvantage with that approach is that a student’s score is dependent on the rest of the cohort. This is particularly problematic in the science suite where the single science cohort is, on average, stronger than the science/additional science cohort.

Having considered all those options, we concluded that the fairest approach was to continue to report grades. A system based on grades strikes a reasonable balance between providing enough information so that users can differentiate between students, without suggesting unrealistic levels of precision. That was our view, but the consultation set out the alternatives and asked for views. Those that responded generally wanted us to continue with a grade scale.

With more grades above C in the new GCSEs, there was a strong case for a new grade scale – otherwise how would users compare an A in 2016 with an A that meant something else in 2017? In October 2013, the Ofqual Board agreed to implement the new 9 to 1 grade scale and in November 2013, we started a separate consultation about how we would set standards in 2017. At that point we proposed the ‘anchor points’ at 7/A, 4/C and 1/G, so that links could be made with the A* to G scale.

1 to 9, or 9 to 1?

Finally, how did we decide whether 9 should be the highest grade, or 1?

It wasn’t easy. There’s no obvious precedent. The most recent examples in England (CSEs) had 1 as the highest grade, but most international systems that use numbers use 1 as the lowest grade. We spent a lot of time thinking about the pros and cons, and in the end we decided to follow international practice.

So the grades being awarded in a few weeks’ time have been the subject of much thought over the past 4 years. During that time, we’ve been talking to teachers, parents, students, further and higher education and employers to explain the new scale. We have written more about that here, and we’ll keep doing that in the coming weeks, and as more new GCSEs are rolled out in 2018 and 2019. We consider that hugely important, because just as we want students receiving their new numerical grades on 24 August to be able to show clearly that they have sat the new, more challenging GCSEs, we want others to recognise it too.

Sharing and comments

Share this page