Reading the Runes

No doubt anyone with an interest in GCSE exam results will have followed the stories emerging from this year’s data set. This was always going to be a more complicated picture with the impact of revised specifications and the new grading system. Furthermore, this year, being a transition year with a mixture of old courses in most subjects but new ones in English and Maths, presents challenges in terms of identifying underlying trends.

However, what I want to try and do here is to draw out some key messages for GCSE science courses. What can we learn from the restructuring of English and Maths courses and what does this tell us about what we can expect next summer?

The first point is key. Ofqual have maintained that they will apply the principle of ‘comparative performance’ so that students demonstrating the same degree of performance will come out with the same grades. Even though the grading system has changed and there is no ‘one to one equivalence’ between old (lettered) grades and new (numbered) grades, there are three ‘pegged equivalences’. These are the thresholds to get a grade 1 (same as that to get a grade G), a 4 (same as a C) and a 7 (same as an A). We have to be careful – this doesn’t mean that a grade C is the same as a grade 4. Lots of students get a C and in the new world some of these now will get a 5, as the grade boundaries in this area will get narrower. What it does mean that the performance needed to just get a C should, after the changes have gone through, just get a candidate a 4. This was planned by Ofqual; Sally Collier, head of the regulator, said that they “had required exam boards to use the same system on comparable outcomes to ensure fairness”.³

This seems to have worked as intended. In English Language 16.2 % got an A or A* last year and 16.8% got a 7 or higher this year. 69.7% got a C or higher last year and 69.9% got a 4 or higher this year. Note that last year’s data here includes iGCSEs as many schools used those; this year that was not an option. Maths showed a similar trend, with 19.7% getting A or A* last year and 19.9% a 7 or higher this year. 69.9% got a C or higher last year and 70.7% got a 4 or higher this year.¹ With the same principle being applied to science next year, we can expect a similar effect. Obviously, this applies to headline figures and behind this, of course, there can be significant rises and falls in the fortunes of particular schools.

There will need to be a job done, if it hasn’t been done already, of managing expectations. A grade 5 is now being defined as a ‘strong pass’ and clearly if the C grade threshold is pegged to the grade 4 threshold the percentages of students getting a 5 or higher is very likely to be lower than those currently getting a C or higher.

Secondly, top grades are calculated in a different way, though the nett effect may not be dissimilar.  Whereas there is a very direct comparison between thresholds for students getting 1s, 4s and 7s with the previous system, there is a different mechanism for determining 9s. At one point it was going to be the top 20% of those getting a 7 or higher but it’s now calculated as 7% + 0.5 × (percentage of students awarded grade 7 and above). This designed to reflect the fact that intakes for particular courses are not necessarily representative of the whole cohort. This is likely to make a difference next year in the case of separate sciences; students who are entered for this are not usually drawn from across the full ability range.

Let’s explore what this means using a hypothetical example. This year 42.1% of the entrants to GCSE Biology in England scored a grade A or higher.²  If this figure is reflected in next year’s results, remembering the principle of comparative performance and we assume 42.1% of candidates get a grade 7 or higher, then 7% + 21.05% of those students will get a grade 9. This 28.05% is obviously higher than the 20% previously proposed. I make this 11.8% of the total intake, a little under the 13.1% who got A* this year.

Science (as in core science), of course, is affected the other way. This year 4.2% of the entrants got A or higher; the grade 9 formula would then give us 9.1% of those entrants getting a grade 9, a lot less than the 20% talked about before.²  This comes out as just over 0.38% of the entry, slightly above the 0.3% who got an A* in science this year.

It’s worth while emphasising though that none of the grades are awarded using grade specific criteria – there is no ‘grade 7 content’ or ‘grade 9 content’; it is the overall performance that is determining outcomes.

That ought to steady the nerves a bit regarding separate sciences. The approach being taken can be discerned, it’s in line with what Ofqual have been saying and even though the exams have become more challenging, students exhibiting the same level of performance are coming out with pretty similar grades.

However, it’s trickier to extrapolate forward to next year’s Combined Science. Unlike Biology, Chemistry and Physics where there is a one to one correlation between the old course and the new one, the nearest to Combined Science is ‘Core + Additional’. However, it’s not quite as simple as that – not everyone who does Science also does Additional Science and the Science entry consists of some Y10 students and some Y11 students, except that this year it doesn’t because the Y10s couldn’t be entered as they’re on the pathway leading to the new exams. Hence the Science entry dropped by over a quarter. Furthermore, it is the Y10 entrants who tend to do slightly better, so the A*-C percentage looks a bit sick. However, a more useful picture is conveyed by the Additional Science performance, which isn’t subject to such factors; the A*-C percentage dropped by 1.5% to 58.2%. The drop was less towards the extremities of the awarding; A*-A dropped by 0.3% and A*-E by 0.4%.² The DfE’s figures showed that stripping out the Y10 factor reveals a positive underlying trend – looking at Y11 performance across ‘core and additional’ shows an increase from 47.6 to 48% getting C or higher.¹ The longer-term trend, however, is less encouraging. As The Guardian pointed out, the four year trend in outcomes in science is pretty solid in separate sciences but less encouraging in science and additional science.⁴ This predates specification or grading changes.

So how should we approach next summer then? It looks as if the system is holding with the stated equivalences and if students enter the exam hall able to perform the same way they would have done this year, they’ll get the grades they deserve. Drawing on experiences of ‘what will get a C (or an A)’ will hold (even though some of the marks may be trickier to get). Students will need to be reassured about this. The cherished grade 9s (“a new grade, designed to recognise the very highest performing students”¹) will be awarded using the formula. The challenge is more with the lower attaining pupils. As more of them end up doing two GCSEs the focus is on keeping those guys engaged.

Sources

1. gov.uk/government/news/guide-to-gcse-results-for-england-2017
2. jcq.org.uk/examination-results/gcses/2017/gcse-full-course-results-summer-2017/gcse-full-course-results-summer-2017
3. bbc.co.uk/news/education-41023354
4. theguardian.com/education/2017/aug/24/proportion-of-students-getting-good-gcse-grades-falls-after-reforms

Ed Walsh