skip to Main Content

Even fewer grade 9s

It has become clear that far fewer grade 9s than A*s will be awarded in English (and Maths) in 2017 and subsequent years. This has some unexpected implications for teaching and learning.

Ofqual originally planned that “the proportion [of candidates] awarded a grade 9 overall should be less than, but not dramatically less than, the proportion awarded a grade A*”. To ensure comparable outcomes between the current grades and the new, the proportion of candidates attaining A*/A at present will receive grades 9 to 7. Of these, it was intended that grade 9 would be awarded to the top 20 per cent in each subject “to provide a greater level of discrimination than the present A* grade”, i.e. making 9 more difficult than A*, but not much more difficult. Grades 8 would then be awarded to the next 40 per cent and grade 7 to the final 40 per cent. Setting grade standards for the first award of the new GCSEs : Ofqual Board Paper 39/14 (Ofqual 2014) paragraphs 55 – 68, especially 66 and 67.

Ofqual has now accepted that some subjects achieve considerably more than 20 per cent of A*s – Latin and Classical Greek are given as examples – and has changed its policy to reflect this. The proportion of grade 9s will be ‘tailored’ to fit each subject’s history of A*s more closely. “The tailored approach uses a formula that would result in, across all subjects, close to 20% of those awarded a grade 7 or above being awarded a grade 9.”
Decisions on setting the grade standards of new GCSEs in England – part 2 (Ofqual 2016) section 4.

This looks like a minor technical change, but it isn’t. While some subjects will receive considerably more than 20 per cent of grade 9s, others including English will receive substantially fewer.

Ofqual provides the following table showing the impact of the different approaches to the award of grade 9 in English Language and Mathematics. The figures in blue in brackets beneath the percentages give the numbers of students expected to achieve A* and A in 2017, rounded to the nearest thousand, based on Ofqual’s analysis of the 2014 results.

  English Language[1] Mathematics
Entry 611,000 585,000
% of grades A and A* 18.8
% of grade A* 5.3
The 20% approach (% of grade 9) 3.8
The tailored approach (% of grade 9) 3.4
The tailored approach (% of grades 7, 8 and 9) 18.8

As you see, the number of candidates attaining the highest grade in English Language drops from 32,000 to 21,000 under the new ‘tailored’ approach, a drop of nearly one-third. In fact, the drop will probably be even greater because:

  • 21,000 grade 9s are 18.26 per cent of the 115,000 candidates attaining A* and A.
  • Ofqual’s formula for calculating the number of grade 9s per subject is: Percentage of those achieving at least grade 7 who should be awarded grade 9 = 7% + 0.5 × (percentage of candidates awarded grade 7 or above).
  • 50 per cent of 18.8% attaining A* and A = 9.4%
  • 9.4% + 7% = 16.4%
  • 16.4 % of 115,000 = 18,000 to the nearest thousand, a drop of 44% from the 32,000 attaining A*.

I have written to Ofqual asking for an explanation of the difference between the two calculations and will post it when I have received it.

Ofqual’s latest paper says that the ‘tailored’ approach will “result in, across all subjects, close to 20% of those awarded grade 7 or above being awarded a grade 9”. This isn’t so, as explained in the paper by Tom Benton which proposes the formula for calculating grade 9s which Ofqual has accepted. Benton writes:

“we suggest adopting the following simple formula:

Percentage of those achieving at least grade 7 who should be awarded grade 9 = 7% + 0.5× (percentage of candidates awarded grade 7 or above)

For example, if 20% of people get grade 7 or above then the percentage of these awarded grade 9 should be 17%. If (as for Biology grade A) 42% get grade 7 or above, then 28% of these should be awarded grade 9. If (as for grade A in Latin) around 74% of candidates get grade 7 or above, then 44% of these should be awarded grade 9.”

Benton then states Ofqual’s aims with greater frankness than Ofqual:

“This rule fits with the stated aims for grade 9 in that:

a. Overall, across all subjects, the percentage of those achieving at least grade 7 who will be awarded grade 9 is very close to the 20% suggested by the currently intended rule.

b. Across almost every subject, the percentage of candidates achieving grade 9 will be substantially lower than the percentage who achieved grade A*” [bold added].

See A possible formula to determine the percentage of candidates who should receive the new GCSE grade 9 in each subject [PDF document], page 6.

How much lower?

Unfortunately it isn’t possible to calculate the number of grade 9s for English Language in 2017 because the grades for English and English Language are published as combined figures. The nearest estimate is 18,000 as explained above.
But an accurate figure for English Literature can be calculated from the number of awards in 2016 published by the Joint Council for Qualifications –

English Literature
Entry in 2016 414,286
% of grades A and A* 21.3
% of grade A* 5.1
The 20% approach (% of grade 9) 4.1
The tailored approach (% of grade 9) 3.6[2]
The tailored approach (% of grades 7, 8 and 9) 21.3

The number of candidates receiving the top grades in English Literature will fall by almost exactly 10,000 between 2016 and 2017, from 24,856 to 14,969, a drop of 40 per cent.

Implications for teaching and learning

Competition for grade 9s will be fierce, especially from selective, independent and other high-achieving schools whose reputation depends on the numbers of top grades achieved. Independent schools aren’t subject to Progress 8, of course, but their attraction to parents depends largely on their examination results.

In these circumstances most schools are likely to achieve very few grade 9s, if any, in 2017 and subsequent years. The competition for 9s will also impact on 8s, making them more difficult to attain. And remember that Ofqual’s statutory commitment to consistency of awarding between Boards means that their standards of awarding will be closely monitored so that no Board can award higher grades more easily than another.
In these circumstances schools will do their best for their most able candidates, but they should also devote resources to the moderately and less able. This is because (a) enabling moderately and less able students to attain a grade above expectation counts as much for Progress 8 as with more able students (see Progress 8 above expectation 2016 [PDF Document]) and (b) this will be easier to achieve because there will be less competition for the lower grades.

Schools should therefore investigate teaching programmes which engage moderately and less able students in discussion of literary texts and techniques, building the confidence that will enable them to respond confidently to unseen texts and closed-book questions.

[1] Ofqual’s projections for English Language, based on aggregated 2014 data for English and English Language.

[2] 50 per cent of 21.3 = 10.7 + 7% = 17.7% of number attaining A* + A (84,514)

Leave a Reply

Your email address will not be published. Required fields are marked *

As the mocks approach

As the mocks for the new GCSE English exams are approaching, I’m concerned that PiXL is still recommending these nonsensical grade boundaries for the new GCSE English exams (the number of marks for each grade has been added):

Grade Percentage Marks for each grade
9 94% 7 marks
8 83% 11 marks
7 72% 11 marks
6 67% 5 marks
5 63% 4 marks
4 59% 4 marks
3 38% 20 marks
2 27% 11 marks
1 15% 12 marks

PiXL has subsequently provided grade boundaries for Language and Literature for each of the GCSE Boards.  They have similar mark ranges with only 4 or 5 marks for each of grades 4 – 6 and much wider mark ranges for higher and lower grades.  No GCSE has ever had this kind of mark distribution and schools that follow it will predict grades very inaccurately.

A more accurate set of grade boundaries can be derived from Ofqual guidance, viz

Grade Mark range (/100) Marks for each grade
9 99 – 100 3 marks
8 94 – 98 5 marks
7 89 – 93 5 marks
6 78 – 88 11 marks
5 67 – 77 11 marks
4 56 – 66 11 marks
3 41 – 55 15 marks
2 26 – 40 15 marks
1 11 – 25 15 marks
U 0 – 10 10 marks

For a full explanation of how these grade boundaries have been derived, see New GCSE grade boundaries PiXL and Ofqual [PDF document].

The DfE expects grades to fall as the new GCSEs are taken.  This is reasonable as all new exams result in lower results in their early years owing to teachers’ unfamiliarity with them.  The new GCSEs are also more demanding in several ways.  By 2018 the Government expects GCSE results to be at “ground zero” from which they can only rise.

Faced with this, schools can do several things to reduce the shock and prepare for the next few years.

  • Use sensible grade boundaries for predictions.
  • Accept there will be fierce competition for the limited number of grade 9s and grade 8s including from selective and independent schools whose reputation depends on them.
  • It will therefore be easier, with appropriate teaching,  for moderately and less able students to build higher Progress 8 scores by achieving grades above those expected on the basis of their KS2 fine-level score.  This will also be helped by the grade mark ranges for grades 1 – 3 being much wider than for 7 – 9.   A matrix is attached showing how less able students attaining grades above expectation can achieve P8 scores as great as more able students – Progress 8 above expectation [PDF document].
  • A cognitively rich teaching programme, developing the literary skills needed for the new GCSE specifications, will help students attain higher grades by increasing their understanding and confidence.
  • Progress 8 is designed to raise the attainment of students who currently achieve grade D and below.  This is cross-party policy to improve our productivity as a nation, so P8 is here to stay.  The unspoken subtext is that mixed ability teaching will lead to higher P8 scores.  It’s not too soon to start seriously considering this.