The impact of exam changes on students explained - RSA

The impact of exam changes on students explained

Blog 1 Comments

  • Picture of Hannah Breeze
    Hannah Breeze
    Former RSA Researcher
  • Education
  • Higher education
  • Schools

It became apparent early on in this pandemic that exams would have to be cancelled. The Government boldly promised that no child would be at a disadvantage.

The announcement that a school’s historic exam performance will play a role in how grades are standardised has raises concerns that the potential of disadvantaged pupils and turnaround schools will be undermined.

How are exams being calculated this year

Never before have grades for whole cohorts in every school had to be systemically calculated.  

Schools and colleges were asked to use their available evidence - mock exams, classwork and homework assignments – to predict the grades a pupil would have most likely achieved. Teachers were then asked to rank pupils within each grade boundary from the most to the least secure.

Schools' predicted grades would then be standardised by Ofqual. The exam board regulator has based its statistical model on ‘the prior attainment of this year’s cohort, expected national grade distribution and school’s historic results’.

A recent Guardian article highlighted the effect of this in the case of Lexie who’s predicted A* grades could be adjusted down to reflect the previous statistical performance of her school - no student has achieved above a C-grade in the last three years in her subjects. It’s high attaining students at historically low performing schools that are most at risk of having grades moderated down.

Just 26 per cent of teachers think this year’s exam system will be fair to all - and as with many things during this pandemic, the pupils in the most deprived areas will be at the greatest risk of disadvantage.  

Legacy of results

School leaders have warned that standardising grades by analysing school performance over the previous three years without considering any trajectory of improvement leaves pupils in rapidly transforming schools at a substantial disadvantage. No child should lose out because of the legacy of their school.

Ofqual tried to quell these calls from school leaders by pointing to SchoolsDash data that suggests just 45 schools made year-on-year improvements of +0.6 of a grade or more.

While this rapid improvement may only be true of a handful of schools it still accounts for over 5,000 Year 11 pupils. Written evidence to Ofqual’s examination consultation suggested these ‘turnaround’ schools will be undermined by historic data and ‘are likely to disproportionately serve poorer communities’.

Ofqual has stated overall this year’s grades are set to be slightly higher than 2019. This will be of little comfort to pupils who may be effectively penalised for their ‘choice’ of school and up to three historic cohorts’ prior achievement.

I use ‘choice’ here because the idea that all pupils and families have real and equal choice in our school system is not accurate. Sutton Trust research shows children from disadvantaged backgrounds attend schools with a much lower proportion achieving the gold standard of at least 5 GCSE grades than their more affluent peers.

A Select Committee of MPs criticised Ofqual for lack of transparency, calling for them to publish further details – they should ‘not be afraid of scrutiny or open debate over whether its model offers the fairest outcome for every student and provider’.

From the Scottish Higher results, we can see the most deprived areas did lose out. After moderation, the pass rate for pupils from the most deprived backgrounds was reduced by 15.2 percentage points compared to only 6.9 percentage points for pupils from more affluent backgrounds. In a U-turn the Scottish government has agreed to reverse 125,000 downgraded marks back to teacher assessed grades.

Optimistic grades downgraded

Figures show ‘very optimistic’ teacher-assessed grades would have resulted in a nine percent improvement in GCSE results this year – an unprecedented increase. Consequently, a ‘substantial number of pupils will receive results where at least one grade has been adjusted to avoid grade inflation’.

The highest percent increases in GGSEs would have been at Grade 4 and 7 (previously C and A grades) and B grades at A-Level. It’s not surprising that these grades might have been ‘optimistically’ predicted. Achieving Grade 4 in Maths and English remains one of the biggest predictors of life chances and a dropped A-Level grade can be the difference between a university place or not.

Rigour is needed to ensure grades awarded this year can be granted with integrity. ‘Optimistic’ teachers are not using this crisis to bolster their schools’ position in league tables. Jon Coles, Chief Executive of United Learning sums it up, ‘the cost of undergrading is so much higher than the cost of overgrading’ to young peoples’ life chances.

To try to protect pupils and to counter Scotland's recent turn around, the Department for Education has promised that pupils will have the option to accept their predicted standardised grade or opt to change it for a mark gained in a mock exam.  

Unconscious bias

Even before the process of standardisation, young people expressed concern about this year’s predicted exam results. 43 percent of working-class pupils are concerned the new assessment system will negatively affect their grades and consequent future pathways.

This concern is not unwarranted. While teachers have done a remarkable job under the current circumstances, evidence shows unconscious bias in teacher assessments can underpredict certain ethnic minority, FSM and SEND pupils. The Sutton Trust suggests around 1,000 high-achieving disadvantaged students have their grades underpredicted per year. The Equality Act Review even called for pupils from black and minority ethnic backgrounds to have their teacher-assessed grades inflated between 1 and 10 per cent to account for unconscious bias. For those pupils who may have already been underpredicted, the standardisation process may only knock down their grade further.

The ‘well-heeled and sharp-elbowed’

Schools can appeal if they believe something systemic has gone wrong in relation to their overall results. However, individual students need to make an appeal through their school that there has been serious malpractice by the school itself or they have been unfairly discriminated against.

The appeals criteria set-out by Ofqual has been accused of favouring the 'well-heeled and sharp-elbowed' middle-class parents who know how to navigate the system. Instead, most pupils will be pushed toward sitting exams in the autumn to improve their predicted mark creating a further issues of inequality.

A recent NFER survey suggests that a third of all pupils aren’t engaging in any distance learning but disadvantaged pupils have faced additional barriers attributed to a digital and technological divide. 93% of Headteachers from the most deprived schools report access issues to online-learning compared to 73% from the least deprived.

Some 230,000 laptops were promised to the most disadvantaged Year 10 pupils preparing to sit exams next year. No provision was made for disadvantaged pupils currently in Year 11 & 13 who wish to continue studying in preparation for autumn examinations.  

Disadvantaged pupils are hit hardest by lengthy periods out of school. Researchers estimate that these pupils could have already lost as much as between four and six months progress.

It’s hard to see how disadvantaged pupils who may have suffered the greatest impact to long-term progress with fewer resources will be on an equal footing to their peers to take autumn examinations.

What next?

On Thursday pupils will receive their A-Level results with GCSE results released a week later – but what next?

Qualifications are stepping-stones into the world beyond school whether that be Further Education, University, apprenticeships or entering the world of work. Come Thursday, every educator’s priority will be supporting pupils in making the next transition successful wherever it may be.

Ofqual has urged post-16 institutions and Universities to be flexible in their admissions criteria to ensure pupils who get “one or two lower grades” progress onto their courses regardless.

The Association of Colleges called for a ‘September Promise’ from the DfE to ensure every young person, in particular those from low-income families and those previously due to start an apprenticeship, have a guaranteed place come September.

Those looking to secure a University place may find that even the most competitive of institutions will be more lenient. Oxford University has confirmed they will take into consideration postcode and school region when deciding on accepting top-performing pupils from prior low performing schools who may miss out on their offer through this years predicted system.

It’s easy to be critical of this year’s process. It was never going to be a perfect model, fair to all. Many have argued that the pre-pandemic assessment system was still weighted in favour of more affluent pupils and teacher bias still crept into the system. It could be that this year’s approach was the only possible one given the unprecedented circumstances and rapid pace of the pandemic.

Statistically this year’s grades may not look that different to previous years in terms of distribution. Only results day will tell what the cost of unconscious bias is towards disadvantaged pupils or an accountability system that places more emphasis on statistical models than recognising improving schools. However, with the threat of a second wave we must avoid more pupils losing out in an unequal system next year by tackling inequalities as a whole within education.

Join the discussion


Please login to post a comment or reply

Don't have an account? Click here to register.

  • Well done for getting this out before it all came to pass on Thursday. Your account seems to fit pretty well everything I have read since except I think some of the most egregious examples of grade reductions have come from the requirement for schools not just to grade students but to rank them - so if they have 4 students they want to predict B's for but the distribution only allows them 2 B's and one C this leads to someone having to be given a D. This seems fundamentally flawed, applying standard distributions to small groups must produce perverse results. When my wife was teaching English by continuous assessment there was a rigorous system for allocating and moderating grades - something like that could have been used and grades based on the evidence of work done by students but moderated to limit grade drift. In France they have used teacher assessments and increased the number of university places to accommodate the extra students. That seems reasonable too.