Predicting the Impossible: Grade Boundary Review

Author: Barnaby Grimble

Category GCSE, Teacher Toolkit, Grade Boundaries

Life without boundaries

With the announcement of the change to 9-1 exams, the exam boards announced a change in grade boundaries that left many teachers across the country feeling perplexed. This was largely because the exam boards and Ofqual said it wasn't possible to predict. We understood teachers’ frustrations with the exam boards & Ofqual not releasing any grade boundaries. They said it was not possible to predict a reasonable range or provide any specimen & practice papers.

In the absences of any guidance, we took matters into our own hands and set out to provide a baseline for teachers to use to help predict grades within the new boundaries and take away some of the anxiety teachers were experiencing. We understood that despite the lack of clear direction, teachers and schools would still be under pressure to give predictions of how their students would perform at GCSE. The results are in and we’re now able to compare our predictions with last week’s GCSE results. So, how close were our estimations? And did teachers even need grade boundaries to achieve high results?

How did our boundaries match up?

Each year the exam boards will adjust the grade boundaries to ensure the spread of students achieving each grade remains approximately constant. This is normally kept below ±5%, but this year we have seen them vary from -3.3% to +8.3% in English & Maths compared to June last year.

To get grade boundary estimates for the new courses to within 6% would be a huge achievement, and within 10% would still be impressive. Out of all of our subjects that used grade boundaries before January 2018, 26 out of 45 were within 6% and 36 out of 45 were within 10%. The group that didn’t make the cut were primarily from MfL, which we’ll discuss later.

Why was our approach successful?

We used the limited information about the way boundaries would be set, existing data and our own classroom experience of curriculum changes to set our boundaries. We applied these boundaries to our tracking systems to support teachers through this challenging year. We also gave advice on using assessment data, pinpointing areas of weakness to maximise the marks students could achieve. We provided practice exam questions written by moderators with mark schemes and exemplar responses to fill the gap of lack of past papers (see our Resource Centre). And now we have been able to see how close our grade boundary estimates were to the real thing.

In calculating our grade boundaries, we looked at the reduction in percentage raw marks seen in English and Maths in 2017 compared to the previous 3 years average percentage raw marks. This was to account for the challenges that come with delivering a new course, lack of past papers for practice, the speed of the curriculum changes and the added challenge in the new GCSEs.

Using this % decrease, we then added 5% to where we thought they would be to ensure there would be added challenge for students. This was referred to as our New Course Adjustment Factor (see table below) – a figure based on statistical evidence and refined by our knowledge and experience as teachers to create our realistic, but challenging grade boundaries. We therefore expected most of our boundaries to be above the real June 2018 boundaries to ensure the challenge was still there. For more info on how we set our boundaries see our videos & posters here

Screenshot 2019-05-15 at 15.49.40

Some of our estimates were very close; Art, Drama, Geography, Music & Food Preparation all had the majority of exam boards coming in below 5% difference. This puts it within range of the expected variance of grade boundaries in a year where they said grade boundaries couldn’t be predicted.

Why did MfL vary so much from previous years?

Modern Foreign Languages saw the biggest variance from the previous 3 years raw mark averages and was seen across all exam boards (see our grade boundaries here). The percentage of raw marks to get a grade 1, 4, & 7 (G, C & A equivalent) dropped significantly, and much more than in other subject areas. We will be using the changes seen in French, German and Spanish to adapt our estimated boundaries for 2019 for the rest of the new MfL courses on the new specification.

Since results day there has been a lot of media attention on the changes in MfL – the reduction in number of students taking GCSE languages will have had an impact. The students choosing to take MfL and encouraged by schools also seems to have changed significantly, most likely due to the pressure of filling bucket 2 to maximise progress 8.

Looking at the Grade boundary comparisons the foundation tier grade boundaries and lower end of the higher tier decreased the most, suggesting that a much bigger spread of abilities of students were taking Modern foreign languages. The challenge level of the MfL courses also must have been significantly harder that the old specification, more so than other courses, reflected in the lower raw mark percentages required in the course to achieve a grade.

Spotlight on Physical Education

Our estimations for PE set our grade boundaries higher than we saw in August. Three of the exams boards came in at higher than we expected, but well within the yearly variation range. All four exam boards had lower boundaries than we had set which aligned with our approach of setting challenging grade boundaries. The change to Progress 8 has impacted the GCSE PE cohort across all exam boards significantly compared to other foundation subjects with a decrease of over 20% (from 126,013 in 2017 to 95,873 in 2018). Such a large change in the cohort could account for the lowering of average raw mark percentage seen for a grade.

AQA was slightly further out than the other exam boards. The boundaries for grade 5-1 showed the biggest difference suggesting that the ability range of students taking this course required a greater spread of grades at the lower end to differentiate their achievement. This would account for the overall average being slightly lower.

We received very positive feedback from our PE community about the 6 mark questions, exemplars and markschemes (see our Resource Centre). The style and level of the questions matched those seen in the exam and gave the students ideal practice, especially with the lack of past papers. One of the practice questions we published was also very similar to one of the exam questions which really boosted the confidence of students and teachers alike.

What does this mean for next year?

Now that we have grade boundaries for 2018, and two years worth for English and Maths we will be using this as a basis for our grade boundary setting for the coming year.

We still recommend adding a challenge to the grade boundaries you use to keep stretching your team and students and to account for any variation. Our trackers will be updated ready for use in September with the 2018 grade boundaries and you will be able to choose the percentage that you add for your challenge level.

Was having Grade Boundaries really that important?

As you would expect, one of our most frequently asked questions for the past two years has been, ”how can you predict grades with no-one publishing grade boundaries?”. Of course we had to come up with our own boundaries in the absence of any guidance, just like every teacher had to decide what they would use.

Our focus has always been on identifying where to focus your efforts and the best ways to maximise marks; if you maximise the marks, you will maximise achievement regardless of the grade boundaries. We designed our Assessment Platform based on our own experience to support teachers with increasing the accuracy of forecasts based on the marks and exam board specific course design.

Our trackers and reports provide teachers with an easy way to identify and communicate the areas of focus.

The grade boundaries really come into play when communicating to parents where their children are at, to compare to national and expected performance measures and at a whole school level. Naturally senior leaders and academy trusts need to use grades as it is the standardized measure that helps them to effect change at a broader level. But that also means they need to have trust in your grade forecasts. Being able to clearly show where the marks are coming from across the course and how that affects the grade predictions helps to improve that trust.

At a department level the grades helps give the overview, but it is the fine detail of where the marks are coming from that informs your teaching and planning. Regardless of what the grade boundaries, if you and your students know where they are at and where they can improve, then you are already on a road to success.


Summary Table: Overall Comparisons All GCSE Subjects: Unit Comparisons

See how easy it is to adjust Grade Boundaries in Pupil Progress

Simple and fast Grade Boundary adjustment is one of the core features in our Pupil Progress Assessment Platform. See it in action in our short video here.


6 Week WF - Grade Boundaries


All previous year’s and our estimated grade boundaries are pre-loaded in the system, meaning you can choose which exam series grade boundaries you want to use

Increase or decrease the boundaries by a percentage to change the level of challenge for your students. All grades on the tracker are instantly updated to reflect the changes in the boundaries.

If you'd like to give this and our many other features a try, don't miss our six-week no-obligations free trial for the Pupil Progress platform.

Try the Pupil Progress Tracker Free for Six Weeks (with no obligation)