https://ofqual.blog.gov.uk/2017/11/03/gcse-maths-grade-boundaries/

GCSE maths grade boundaries

Over the next two months we will be reporting on various aspects of summer 2017, including official statistics. Our very clear aim, in planning for the first new GCSEs in summer 2017, was that the transition should be as smooth as possible, and that the students taking them would not be unfairly disadvantaged by being the first to sit these new qualifications.

The transition to the new GCSEs was generally smooth. In maths, schools appear to have made appropriate tier entry choices for their students: there were fewer students this summer who were ungraded on the higher tier and no students scoring full marks on the foundation tier.

However, some commentators expressed concerns about the grade boundaries in GCSE maths. Much has been written about the ‘low’ grade boundaries and how they are lower than previous years. This has been interpreted as Ofqual lowering the boundaries in the first year of a new qualification.

Unfortunately, it’s really not that simple. The new maths GCSEs were designed to be different from the old A* to G GCSEs, so you really can’t compare new and old. Here’s why.

The overlap

First, the available grades on each tier are different and the overlap grades are different. Previously the highest grade on the foundation tier was a C. In the new GCSEs, the highest grade on the foundation tier is a 5, a grade which spans the top of a C and the bottom of a B. The overlap grades –those available on both tiers – are 5, 4 and 3. That’s higher than the overlap on the previous qualifications of C, D and E, which probably explains the shift towards the foundation tier entry this year.

The design of the papers

As well as those differences, for the first time we also set specific rules about how the papers should be designed. Why do that? Higher tier papers must strike a balance between testing the content aimed at grade 4 and 5 students, and providing challenging questions on the grade 9 content and it’s important that exam boards are consistent in this. Our rules say that:

  • In a higher tier paper, half of the marks should be targeted at grades 9, 8 and 7 and the other half of the marks should be targeted at grades 6, 5 and 4.
  • In a foundation tier paper, half of the marks should be targeted at grades 5, 4 and the top of grade 3 and the other half of the marks should be targeted at the bottom of grade 3 and grades 2 and 1.

This is shown in the infographic below. Targeting questions is always tricky, but this means that higher tier papers now contain more demanding questions and only about a sixth of the marks on those papers are designed for students working at grade 4. In that context, it’s not surprising that the grade boundary for a grade 4 on the higher tier papers was around 20% of the maximum mark. But that doesn’t mean there is anything wrong with the papers. Rather, it’s a consequence of having to discriminate more at the top end but also provide sufficient challenge across the ability range.

On the higher tier, half the marks are targeted at grades 9, 8 and 7; half the marks are targeted at grades 6, 5 and 4. On the foundation tier, half the grades are targeted at grades 5, 4 and 3; half are targeted at grades 3, 2 and 1.

If there were more marks targeted at grade 4, grade boundaries might be higher, but exam boards would be criticised for making their papers too easy, and it would mean fewer marks available to differentiate the very good students at the top end.

As schools and colleges become familiar with the new assessments, we might expect performance to improve slightly. And of course, grade boundaries will be set each year to reflect the difficulty of the papers. But we should not expect to get to a position where students have to score 50% of the marks to achieve a grade 4 on a higher tier paper, unless we redesign the papers to include many more questions targeted at grade 4.

Cath Jadhav
Associate Director for Standards and Comparability

5 comments

  1. Comment by Steve posted on

    There is nothing in here to disagree with in terms of 'fact' but I think we need to check that the boundaries pass the general public 'smell test'.

    20% for a 'standard pass' just doesn't sit well with many of the general public - it seems too low. I don't think it's realistic for the general populous to ever be aware of the workings of tiers of entry, overlap questions and the like so it's important that the qualifications and grades stand on their own two feet.

    Given that around 75% was enough to give students the new 'superstar' grade 9 and around 55% enough to get a grade 7 (equivalent to the old A grade) it seems possible to make the papers slightly more accessible whilst still leaving plenty of space for the most able to showcase their abilities on the exam paper.

    Interestingly this year a student could, just about, get a grade 7 without getting any of the grade 7/8/9 questions correct as long as they got 100% on the grade 4/5/6 questions.

    I wonder if a 60/40 split between the easier/harder questions on the paper might mean that the grade better reflects, for those at the lower end of the grade spectrum on a paper, mastery of topics rather than potentially just picking up lots of 'bits and pieces' method marks.

    This might mean that the boundaries for grade 7 and 9 move up to around 65% and 85% which seem quite reasonable and pass my 'smell test'.

    In terms of the marks required to get a grades 4 that would rise to around 30% for this 'standard pass' and around 40% for a 'strong pass' (grade 5). That would be an improvement but still may be considered slightly 'smelly'.

    Mt preferred option - as a maths teacher - would be to return to the days of 3 tiers so that students awarded grades 4/5/6 had to genuinely show mastery of some more complex topics to get those sorts of grades.

    Reply
  2. Comment by James Mook posted on

    It's the "targeting" I have problem with. You state yourself it is tricky, and indeed it is. What may seem like something targeted as 4, 5 & 6 may not be dependent on understanding of individuals. Or have you defined a general list of concepts, methods or even topics that are deemed to be grade 4, 5 & 6?

    Reply
  3. Comment by Louisa Wilson posted on

    I love Cathy's blogs. Any chance of one on how you set the grade boundaries for the November resit? It seems a difficult task to make fair so I'd be interested to hear how it's done. Now that most of the grade 3 students will have done foundation rather than higher at their schools there could be a shift this year.

    Reply
  4. Comment by Rozh Barzinji posted on

    Can this higher and lower tier grading system be extrapolated to Science?

    Reply
  5. Comment by J. Eleini posted on

    Foundation Tier is now best for many pupils, specially borderline ones where previously C was the goal but it was hard to decide which tier gave them the advantage. The title Foundation puts off many pupils and their parents who think they are doing very simple maths. It should be changed to a more appealing name such as Essential or Standard Maths, specially now the content has increased.
    Higher Level should be just that,
    namely to suit higher achievers . The difficulty at present is trying to be very broad in scope in the exam and span 4 to 9, resulting in 4 achieved at 20 per cent, or less, which frankly sounds odd.
    The best way forward is to make a standard exam of two papers for EVERYONE , spanning 1 to 5 or even 6.It could be called Essential or Standard Maths. Then those capable and aspiring to a higher grade can sit one extra paper targeted at 6 to 9.Their result from the 3 papers combine for a final grade. A high grade such as 8 or 9 would then be far more meaningful than under the present arrangement.
    This is very much like a driving test where everyone has to pass the basic one . Anyone wishing to go beyond takes an advanced course. With some tweaking this might be a way forward to remove perceived anomalies.

    Reply

Leave a comment

We only ask for your email address so we know you're a real person