159.Assessment in England during the 11–16 phase focuses on GCSEs, which are studied across key stage 4 and assessed primarily in year 11. Pupils are also able to take qualifications with a more technical or vocational focus known as Technical Awards, but GCSEs are overwhelmingly the dominant set of qualifications taken at this age.
160.We heard from numerous witnesses that the current key stage 4 assessment system places intense pressure on both pupils and teachers. This is partly the result of several recent changes to the nature of GCSE exams. First, as discussed earlier, there has been a significant increase in the GCSE curriculum content that pupils are expected to learn.303 Pupils who attended our roundtable session felt that the extensive content required them to ‘cram’ vast amounts of knowledge in the run-up to their exams. They argued that this reduces the opportunity to develop necessary skills and retain subject knowledge for future study or work, as teachers are forced to rush through content.304 Dr Mary Bousted, then Joint General Secretary, National Education Union, noted that this approach results in narrow learning techniques: “Teachers are under so much pressure … if you have to deliver such a big body of content knowledge, of course you resort to rote learning.”305
161.Secondly, there has been a move away from a modular approach towards terminal assessment at the end of year 11. This has resulted in pupils sitting many hours of exams over a short period. Evidence suggested that most pupils will “undergo more than 30 hours of assessment” during GCSE exam season.306 Dr Bousted cited a school leader who described this as “a test of endurance and resilience rather than learning”.307 Some also argued this leads to an undue emphasis on exam preparation throughout key stage 4: “the class time lost to teaching exam rubrics, exam technique, taking practice questions, sitting mock exams, taking exam leave and sitting the exams, can remove over a sixth of the total potential learning time in years 10 and 11.”308
162.The pressure of exam season has been intensified by the Government’s removal of non-exam assessment in many GCSE subjects. This means that pupils’ grades in core subjects are often based solely on their performance in two or three exams taken within a few weeks of each other. This was raised as a primary concern by several pupils who attended our roundtable sessions. They noted the increased impact of a pupil having an ‘off day’ during exam season, without being able to rely on marks gained from non-exam assessment earlier in the year, and highlighted the additional stress this causes.309
163.This intensive period of exams could be having a negative impact on pupils’ wellbeing. In the Youth Voice Census 2022, co-sponsored by the Edge Foundation, “49.1% of respondents stated that assessments have a negative impact on their mental health”.310 This was supported by evidence from Parentkind, a charity representing parents and carers in education, and the membership body for parent-teacher associations. They cited polling they had conducted which found that “almost half (48%) of parents were concerned about their child’s exam stress … which was the third biggest indicator of mental health concerns”.311 Alistair McConville, Deputy Head of King Alfred School and Co-founder of Rethinking Assessment, described the GCSE exam period as “an incredibly stressful experience” which pupils must undergo “at a time of their maximally sensitive development as adolescents.”312
164.The pressure of assessment at key stage 4 is intensified by aspects of the wider education system. We heard that pupils often need to receive specific grades at GCSE to allow them to progress to the 16–19 institution and/or course of their choice.313 A pupil’s performance in GCSE English language and maths is of particular significance, as receiving a grade 4 (a ‘standard pass’) in these subjects is generally a prerequisite for progression to level 3 qualifications and apprenticeships. Moreover, as noted earlier, pupils who receive a grade 3 or below in these subjects are generally required to resit their GCSE or take an equivalent qualification.314 GCSEs can also be used in university admissions processes. Following the Government’s recent reforms, A-levels (taken at the end of key stage 5) no longer include externally validated assessment at the end of year 12. GCSE results are therefore often the only definitive qualification data available for pupils who apply to university at the beginning of year 13.315
165.Teachers are similarly under considerable pressure to ensure that their students perform well in key stage 4 qualifications, as GCSE grades underpin the main accountability measures used to judge overall school performance in this phase.316 Matthew Glanville, Director of Assessment, International Baccalaureate, told us that “there is certainly a real tension … in the role of teachers. Is it to maximise the student’s performance, to maximise the school’s performance, or to be the guardian of a high-quality education?”317 Professor Graham Donaldson, Honorary Professor at the University of Glasgow, told us that this approach risks the primary purpose of assessment becoming school accountability. He suggested that in some cases “young people are there to serve the school’s reputation rather than the assessment serving the needs of the young person.”318
166.Some witnesses argued that the English system of high-stakes assessment at 16 is an outlier internationally. Professor Gordon Stobart, Emeritus Professor of Education, University College London, suggested that other countries focus on exams at 18, and that “we are out of kilter with the United States, Canada, Australia, New Zealand and France.”319 In contrast, Tim Oates, Group Director of Assessment Research and Development at Cambridge University Press and Assessment, argued that many of the best-performing systems in the world have high-stakes assessment at 16. He conceded, however, that in some of these jurisdictions pupils take exams in fewer subjects or are tested through other forms of assessment.320 In its report, The exam question: changing the model of assessment reform, the Institute for Government made a similar point, arguing that:
“Elsewhere, many assessments at 15 or 16 contain coursework or continuous assessment as well as or instead of written tests. England is unusual in the extent to which it relies on linear written exams and also in the number of papers young people sit”.321
167.The Institute for Government explained that the emphasis on extensive, exam-based assessment, and other characteristics of the current system, derive from the fact that secondary exams in England are used for several purposes: to assess pupils’ knowledge of the content they have studied; to provide a consistent measure of pupils’ performance that enables post-16 institutions and universities to select students; and to feed into the performance measures through which schools are held to account.322 Gavin Busuttil-Reynaud, Director of Operations, AlphaPlus, commented that “one of the immense pressures that is placed upon GCSEs is that they carry many purposes”.323
168.Dr Michelle Meadows, Associate Professor of Educational Assessment at the University of Oxford, suggested that GCSEs fulfil “the specific purposes that the Government laid down” for them. She argued that these, particularly the focus on obtaining a reliable measurement of student attainment, have led to “a particular design” for the system. She also asked: “Could we imagine a different set of purposes and a different emphasis? Yes, and that would lead to a different kind of design.”324
169.We heard several proposals for how the current system of high-stakes assessment at 16 could be modified in the long term. First, it was proposed that there could be reform to the accountability system, with a decoupling of performance measures from assessment results, or at least a reduced emphasis on them.325 This appears to be the approach being adopted in Wales as part of its wider education reform programme. Gareth Evans, Director of Education Policy at the University of Wales Trinity Saint David, told us that the Welsh system is looking to move towards data sampling. This is linked to “a disaggregation of assessment of the learner for the learner’s benefit and assessment of the learner for the system’s benefit. There is no longer by standard a publication of attainment outcomes.”326
170.Several witnesses addressed recent calls for the introduction of ‘learner profiles’327, describing these as “a digital record of achievement”328 or “a kind of learning LinkedIn”.329 Mr McConville suggested that they would “capture a broader range of evidence and remove the accountability pressures that tie everyone to eight, nine or 10 GCSEs.”330 Describing an example learner profile produced by Rethinking Assessment, Olly Newton, Chief Executive, Edge Foundation, said that “employers love how it shows not just young people’s academic achievements but their skills, creativity, communication and broader development.”331 Parentkind cited polling which suggested that 74% of parents “would support the introduction of a learner profile which records their child’s skills, qualifications and achievements throughout their education.”332
171.Others have recommended moving away from GCSEs completely. We heard about several independent schools which have significantly reduced their use of GCSEs in recent years and introduced their own qualifications and assessment systems. For example, Bedales School in Hampshire now plans to offer GCSEs in only English language and maths, with pupils primarily studying their own Bedales Assessed Courses at key stage 4.333 These schools argued that GCSEs no longer need to be a definitive final measure of a young person’s achievements at 16 in the way they once did, given that all pupils must now remain in education or training until 18: “GCSEs were created at a time when many young people left school at 16 and needed to be able to demonstrate what they could do at this point of transition. With the majority of pupils now going on to study until 18, that idea is horribly out of date.”334
172.Latymer Upper School has also announced plans to reduce dramatically the number of subjects in which they use external assessment.335 From 2027, they will offer GCSEs in only English language and maths, with performance in all other subjects being judged through a range of internal assessment methods including vivas, presentations and extended projects. Pupils will have the chance to study a combination of long and short courses, covering topics such as robotics, world poetry and printmaking, which the school suggests will enable greater breadth and depth of learning and allow cross-curricular and project-based work.336 Latymer argued that the new approach will better prepare pupils for “university, work and life”.337
173.Latymer said they were inspired by proposals put forward by the Times Education Commission, which argued that GCSEs should be replaced by a slimmed-down set of exams in five core subjects: “This would allow children to progress to the next level and provide accountability for schools, but lower the stakes and reduce the amount of time spent on preparing for and taking exams”.338 This proposal sits alongside suggested reforms to post-16 education. The Commission called for A-levels and other qualifications to be replaced by a “British Baccalaureate” in which most pupils would study a much broader range of subjects. The Commission’s proposals for key stage 4 and 5 draw on the structure of the International Baccalaureate (IB), which is widely respected by employers and universities and is used around the world, including in some UK schools.339
174.The Institute for Government, however, cautioned against recent calls for wholesale reform:
“While GCSEs and A-levels are certainly imperfect, the proposals for overhauling the system typically exaggerate the benefits while failing to acknowledge the costs. Education systems are interconnected and changing one major component dramatically can cause upheaval elsewhere.”340
They highlight, for example, that removing externally validated assessment at 16 would mean shifting to a post-qualification university admissions system, based on actual A-level results rather than predicted grades and GCSEs. Several witnesses also underlined the need to retain some form of national assessment at 16, since more than 50% of pupils move to a different institution at the end of key stage 4. 341
175.In October 2023, the Government announced plans to reform post-16 education over the next decade through the introduction of a new qualification, the Advanced British Standard.342 This would bring A-levels and T-levels together into a single qualification, with “every student also studying some form of maths and English to age 18.” The intention is for pupils to study at least five subjects as part of the “Baccalaureate-style” qualification;343 currently most study three A-levels or one T-level.
176.The Government said that it will review GCSEs in light of this proposal.344 It stated that maintaining externally assessed GCSEs is the “best and fairest way to ensure children learn and retain knowledge”. It also noted that “GCSEs can be onerous for students and teachers” and announced its intention to “look at where they can be streamlined, while still retaining their inherent rigour.” This will include exploring whether “the number and/or length of papers that children sit” could be reduced, and whether digital solutions could allow performance to be assessed “in more innovative and less onerous ways.”345
177.We have heard that the high-stakes nature of key stage 4 assessment in England necessitates an emphasis on terminal, exam-based testing. Witnesses have suggested that the current focus on exam-based assessment places considerable pressure on pupils and can have a detrimental impact on their learning experiences in the 11–16 phase. While there remains a need for some kind of formal assessment at 16, given the number of pupils who change institutions at this age, the current exam burden is disproportionate, since pupils must now remain in education or training up to 18.
178.We recognise that radical reform of GCSEs would constitute a major shift in the current secondary assessment system. This transition would need to be made as part of a long-term programme, alongside changes to the post-16 phase, and extensive consultation would be critical. This clearly goes beyond the Committee’s remit for this inquiry and so we have not made detailed recommendations in this area. We do, however, support recent proposals to move towards a slimmed-down form of assessment at 16, with externally validated assessment used across a smaller set of subjects.
179.We urge the Government to consider proposals to reduce more dramatically the amount of external assessment undertaken at age 16, as it reviews options for a less onerous GCSE assessment model. The Government should set out further details of its proposed review of GCSEs in its response to this report.
180.When GCSEs were introduced in 1986, subjects were assessed by both coursework and exams, with the proportion of each varying from subject to subject. In 2006, the Qualifications and Curriculum Authority (QCA), the regulator at that time, announced that controlled assessments would begin to replace coursework in many GCSE subjects. These assessments took place in the classroom and reduced the opportunity for outside help compared with coursework assessments. This change followed several years of criticism relating to the reliability of coursework and suggestions that it was often being amended, or even written, by teachers or parents.346
181.In 2011, the Office for Qualifications and Examinations Regulation (Ofqual), which had replaced the QCA as regulator, commissioned a survey of teachers’ experiences of controlled assessment, followed by a more comprehensive review in 2013. It found that, in general, controlled assessment was “well received” and that respondents felt it guarded against malpractice, provided a fair assessment of performance and assessed a broad range of skills. However, there were concerns about the impact on teaching and learning time, and the inconsistency with which rules were being followed, with many feeling there was too much room for schools to “interpret the guidance differently”, or ‘game’ the system.347
182.Many of these concerns were reiterated to us by Dr Jo Saxton, Chief Regulator at Ofqual. She commented that “history has shown us that high degrees of non-examined assessment end up being very scaffolded by staff and not necessarily the best reflection of what students understand and know and can do.” She added that this is particularly the case “when the qualifications are high stakes”.348 Sir Ian Bauckham, Chair of Ofqual, also noted that some students were previously doing up to 100 hours of controlled assessment. As this was done in class, it consumed large amounts of time in school that might otherwise have been used for teaching.349
183.In light of these concerns, from 2015 almost all GCSE subjects became assessed purely by examinations taken at the end of the course. Other types of assessment are now used “only where they are needed to test essential skills”.350 GCSEs in English literature, maths, biology, chemistry, physics, history and geography are all now assessed purely through examinations. English language GCSE does contain an element of non-exam assessment in the form of an oral presentation, but this component no longer counts towards a pupil’s final grade. The only English Baccalaureate (EBacc) subjects that still contain non-exam assessment that contributes towards the final grade are the modern foreign language qualifications.351 Some non-EBacc GCSE subjects contain a significant proportion of non-exam assessment. For example, design and technology has 50%, music has 60% and art has 100% non-exam assessment.
184.Much of the evidence we have received has called for a broader range of assessment methods to be used across a greater number of GCSEs. Dr Bousted told us that exams do some things well but are not “a valid way of assessing attainment” in key aspects of many subjects. For example, she highlighted the importance of project-based programming in computer science, extended research in history and fieldwork in geography, suggesting that all of these assess “essential fields of knowledge” in each subject which an exam cannot.352 The Historical Association told us that the terminal exam in GCSE history is “more a test of how much the student can remember and does not allow them the room to show what they know”.353 They also argued that this assessment method is detached from real-life tasks: “In no other circumstances in the working world would someone be expected to produce a coherent, detailed and analytical extended response or report in 20–25 minutes of writing time.”354
185.The Association of Colleges argued that the use of a narrow set of assessment methods is not adequately preparing pupils for post-16 study, particularly for technical and vocational options, or future employment. They told us that a more diverse range of assessment methods “would help to familiarise these forms of assessment, including practical, oral and work-based assignments and projects.”355 Young Enterprise told us that the shift to purely exam-based assessment in many subjects is “an example of where more practical applications of learning have been removed from education over time”.356 It stressed that coursework had helped to “build the confidence and motivation of those young people that found examinations challenging”, as these pupils could demonstrate their understanding of the subject “in a more practical way”.357
186.Rethinking Assessment highlighted a range of other assessment options that could be used instead of, or alongside, written exams. These include performance-based assessment, such as a viva or a practical assessment in music or sport, and extended investigations or projects.358 Mark Marande, Principal of The Petersfield School, described the GCSE assessment system as “very one-dimensional” and proposed “project qualifications, interdisciplinary learning or micro-credentials” as alternative forms of testing that merit consideration.359
187.Pearson told us that in recent polling of pupils, parents, teachers, school and college leaders, and employers, “respondents articulated a strong preference for regular assessments throughout the year—with more coursework, fieldwork, observations and internal assessment”.360 Parentkind noted that in their 2022 Parent Voice survey 76% of parents supported including some coursework in every GCSE subject, rather than examinations only.361
188.At our roundtable events, pupils told us that coursework helped them develop skills such as project and time management, and encouraged engagement throughout the year, compared to terminal exams. They argued that increasing the proportion of coursework in GCSE courses would reduce the pressure of exams and more fairly capture the abilities of all pupils.362 Several teachers called for an end to “100% terminal exams” at GCSE.363 A history teacher commented that coursework enabled pupils to “really explore what history is like”. Arguing that coursework is used successfully in A-level history, they suggested that any problems in implementation at GCSE are “fixable”.364
189.However, data from the survey app Teacher Tapp suggests that, in most EBacc subjects, the majority of teachers are against a return to coursework assessment at GCSE. Only among English teachers was there a majority in favour of a coursework component.365 Pearson concluded that while “there is definitely scope to evolve the methods of assessment incorporated into existing GCSE qualifications”, implementation would need to be carefully considered, “being mindful of teachers’ time and the volume of teaching hours required to implement different assessment methods.”366
Figure 4: Teacher survey: “Would you like there to be an assessed coursework component at GCSE?”
Source: Teacher Tapp, ‘ Ofsted, your colleagues and a return of GCSE coursework’: https://teachertapp.co.uk/articles/ofsted-your-colleagues-and-a-return-of-gcse-coursework/ [accessed 7 November 2023]
190.During the COVID-19 pandemic, examinations at key stage 4 were cancelled and grades were based solely on non-exam assessment, graded by teachers. This led some to raise concerns about grade reliability.367 There was a significant increase in the number of top grades awarded at GCSE in both 2020 and 2021, followed by comparable falls in 2022 and 2023 when exams were reinstated.368 A survey by Ofqual in 2021 found that less than 40% of the public had confidence in the A-level and GCSE grades awarded during the pandemic.369 Dr Saxton noted that pupils she had engaged with supported a return to exam-based assessment, stating that: “I was overwhelmed by the extent to which students wanted their exams reinstated … They felt trust in the examined system, where there are expert markers.”370
Figure 5: Percentage of GCSE grades awarded in all subjects (England only)
Source: Joint Council for Qualifications CIC, ‘Examination results’: https://www.jcq.org.uk/examination-results/ [accessed 10 October 2023]
191.Mr McConville cautioned against placing too much emphasis on the experience of teacher assessment during the pandemic, due to the specific circumstances: “Teachers were not prepared for that experience … that was not a totally fair set of evidence to use to compare.”371 Likewise, Mr Glanville argued that teacher assessment can be implemented reliably through a robust moderation process, such as that used as part of the IB’s Middle Years Programme:
“We can ensure that teacher marks are consistent across schools through the moderation process … It is about selecting a random element from the teacher’s marking and making sure that our examiners agree with the marks that the teacher has given and the reasons why. That can be done.”372
192.However, the thinktank EDSK noted in a recent report that even with moderation in place, “students often appear to perform better in assessments such as coursework and controlled assessments that are graded by their teacher rather than external examiners”.373 Moreover, moderation is not able to mitigate against different pupils receiving different levels of teacher support during the non-exam assessment, or any other attempts by schools to ‘game’ the system. Mr Oates highlighted the example of Sweden where, with increased use of teacher assessment, there had been a significant increase in pupils’ grade outcomes but “an almost mirror decline” in attainment, based on international surveys: “They found that they had relied on teacher assessment without a whole series of other safeguards in place. Sweden is now looking at introducing more testing, not removing it.”374
193.EDSK also highlighted that generative artificial intelligence (AI) tools such as ChatGPT present further challenges to ensuring reliability in non-exam assessment. They note that the ability of such tools to produce essays and projects with minimal input has significantly increased the risk of plagiarism. This issue was also raised by several witnesses, including Sarah Fletcher, High Mistress of St Paul’s Girls’ School. However, she argued that St Paul’s is at least partially mitigating this risk by ensuring that extended essays are written in controlled conditions.375 Mr Glanville commented that the increased prevalence of tools like ChatGPT could mean that teacher judgement will have an even greater role when non-exam assessment is used: “The people who are best placed to understand whether it is the student’s own work are, of course, the teachers.” However, he also noted that this can work only “if we have a system whereby the pressure on the teacher is not to maximise the student’s grade but to represent the result accurately”.376
194.Witnesses have suggested that increasing take-up of the Higher Project Qualification (HPQ) could enable more pupils to benefit from the opportunities of non-exam assessment in key stage 4. The HPQ requires students to carry out research on a topic of their choice that is not covered by their other qualifications. They use this research to produce a written report and, in the case of practical projects, an artefact.377 AQA, an exam board that offers the qualification, highlights that “a student can take inspiration from something studied in class or something completely unrelated to their studies”.378 Mr Newton told us that the HPQ “encourages the sorts of skills that excellent schools encourage—team working and problem-solving” and would be “an impressive thing to show an employer”.379
195.The HPQ is overseen by project supervisors, who guide pupils through the process, and is expected to take 60 guided learning hours.380 Schools and teachers are not assessed on their pupils’ performance in the qualification since the HPQ is not recognised in the headline school accountability measures.381 In 2023, the HPQ was taken by 5,347 pupils,382 just 0.8% of all 16 year-olds taking GCSEs.383 Jenny Clements, writing for the British Educational Research Association, highlighted that the take-up of the HPQ declined dramatically in the early 2010s. Over 20,000 pupils entered for the qualification in 2011, but this decreased to just 4,829 in 2016. One exam board, OCR, dropped the qualification completely in 2019.384
196.There is some evidence to suggest that non-exam assessment produces less reliable grades than traditional exams. However, we heard that non-exam assessment supports the development of knowledge and skills that are more difficult to assess in an exam context. Witnesses suggested that, used alongside exams, it can lessen the pressure of assessment for pupils, by reducing the significance of terminal exams to their overall grade, and help to capture the full range of their achievements. We are persuaded that an increase in the use of non-exam assessment at GCSE would bring benefits for pupils.
197.Concerns about reliability, plagiarism and the impact on teacher workload mean that an increase in the use of non-exam assessment should be approached cautiously. Careful consultation with teachers and schools will be vital to ensure that any increase in the use of non-exam assessment at GCSE is manageable to deliver. Increasing take-up of the Higher Project Qualification would enable more pupils to experience the benefits of non-exam assessment at key stage 4, without requiring any significant changes to the current suite of GCSEs.
198.As part of a longer-term review of qualifications at 16, the Government should introduce a greater proportion of non-exam assessment at key stage 4. In the short term, the Government should set out how greater take-up of the Higher Project Qualification at key stage 4 could be encouraged, to enable more pupils to undertake an extended project qualification alongside their GCSEs.
199.The vast majority of GCSE exams in England are still conducted using pen and paper.385 Even in GCSE computer science, most pupils do not use a computer to complete their exams. Sharon Hague, Senior Vice President of Pearson School Qualifications, highlighted to us that “Pearson is the only awarding organisation to provide an on-screen assessment in computer science where children are coding in the examination and being assessed on the quality of their coding. We are not the most popular.”386
200.There have been strong calls in recent years for a move towards greater use of technology in school assessments. The Times Education Commission recommended online testing as part of its overall proposal for a slimmed-down assessment system at 16. It suggested that this could bring significant savings to the education budget and noted that AQA, an exam board “which has to print, deliver, collect and mark about 12 million scripts a year”,387 is already trialling on-screen assessment. The Institute for Government was also positive about the potential for online assessment, listing a move towards this as one of a series of “smaller improvements” to the current system that could have “important benefits”.388
201.A number of witnesses emphasised the potential benefits of on-screen assessment. Pearson told us that it can “facilitate a far more modern, flexible, and inclusive examination model for young people” and that “70% of teachers feel on-screen assessment will provide faster, better insights about students’ performance, helping to improve teaching and learning”.389 The question of inclusivity was also addressed by the National Association of Special Educational Needs (NASEN.) They suggested that the dominance of paper-based exams presents teachers and special educational needs coordinators with “a difficult decision for students who benefit from assistive technology in the classroom, but who … would not be granted access to the same adaptive methods in exams”. They argued that these pupils may be being “set up to fail”.390
202.The International Baccalaureate Organization, which has offered digital assessments at the end of year 11 since 2016, highlighted that on-screen exams do not have to be simply “digitised conventional examination papers”, but can “take advantage of the digital environment to utilise a range of innovative features”.391 The IB Middle Years Programme assessments, for example, may require students to engage with video content or plot graphs, as well as offering features that can support learning “such as hover-text to explain or define words that may be difficult for students”.392 Analysis conducted by AQA in 2022 identified “a general expectation” among the teachers and school leaders surveyed that with digitisation “the examination experience would evolve to become a richer one with, for example, real archive documents in history, digital mapping in geography and the use of relevant applications.”393 AQA’s analysis also raised the important point that completing tasks on-screen will feel more familiar and relevant for many young people than completing them on paper:
“Digital technology is what they have grown up with and how they prefer to work. Editing of answers is a clear advantage. Many prefer to type rather than write and few were concerned about typing in an exam situation given enough time to prepare. Typing is seen as an essential skill.”394
203.AQA’s polling found that 81% of headteachers and 64% of teachers surveyed felt that “digitisation of examinations is inevitable”. Respondents were persuaded that this was necessary given the growing importance of digital skills in higher education and employment, and the centrality of digital experiences to young people’s lives. AQA noted that “many teaching leaders felt examinations could not remain static amidst such fundamental change and to do so would have a negative effect on the ability of students to demonstrate their abilities in summative assessment.”395
204.Dr Saxton described a move towards greater use of on-screen assessment as a question of “when, not if”.396 She noted that Ofqual is conducting “an ongoing feasibility study to look at … what would be needed for high-stakes qualifications such as GCSEs to be delivered on-screen.”397 In a recent speech, she argued that England should move towards having “a mixed approach” that combines on-screen assessment with more traditional, paper-based exams.398 Dr Saxton also noted that there are “obstacles and pitfalls” and highlighted the importance of ensuring that any transition towards alternative forms of assessment “is safe for students”.399
205.Ofqual published a review of on-screen assessment at the end of 2020 which found “no evidence in the literature of regulation as a barrier to greater adoption of online and on-screen assessment in its own right”.400 However, it did identify other barriers including:
206.Similar concerns were raised by NASEN and ASCL; the latter told us that “many schools are not sufficiently resourced to deliver large-scale assessments electronically”.402 Pearson noted that “only 2% of schools in England offer a 1:1 device to student ratio”, a far lower percentage than in the United States, and stressed that sufficient funding would be needed to address practical barriers such as access to equipment.403 Pearson also warned that many pupils in the UK may not have the required digital skills to complete on-screen assessments, citing polling among teachers: “When surveyed, 59% of secondary teachers feel the current curriculum does not provide learners with the necessary digital skills to take on-screen exams”. They argued that “we need to ensure there is a level playing field for children from different backgrounds with different degrees of digital literacy”.404 ASCL similarly highlighted that “many young people still don’t have everyday access to devices or unlimited access to internet data at home”.405
207.We heard that pupils and parents have expressed reservations about on-screen assessment. A recent survey by Ofqual found that just one in five students and parents thought that all GCSE exams should be taken on a computer, with 48% of students and 54% of parents preferring a mixture of on-screen and handwritten assessments.406 In contrast, Pearson drew attention to positive feedback received from candidates following on-screen assessment trials they conducted in Bahrain, UAE, Qatar and Spain in 2022. They noted that 77% of the 600 students who completed their final English language GCSE assessments on-screen said afterwards that they preferred the on-screen format to more typical pen and paper exams.407
208.Reflecting on transitions to on-screen assessment in other countries, witnesses suggested that the obstacles outlined above are not insurmountable. Ms Hague shared examples of how the issues of access to devices and data had been addressed in Egypt and the United States.408 Dr Meadows highlighted the importance of “massive communication efforts … not just with schools but with parents and stakeholders beyond the education system”. She also stressed the need for “political will” to drive change. 409 These factors, along with others such as thorough piloting and “a high degree of student input during transition”, were similarly highlighted by Ofqual as being evident in jurisdictions that had “successfully implemented” a move to on-screen assessment.410
209.In this context, the Committee was interested to hear evidence on the introduction of on-screen assessments to measure literacy and numeracy skills across all maintained schools in Wales. Mr Busuttil-Reynaud, who has been closely involved in the rollout of these ‘personalised assessments’, emphasised the role of the Welsh Government in addressing the practical barriers faced by schools:
“Taking lessons from Wales, the single most important thing is that the state has to do things that have to happen at a systemic level. The Welsh Government organised for the high-speed digital connection of all of their schools, funded that and made sure it happened. They also provided funding through the local authorities for devices and it was basically mandated that there would be on-screen assessment.”
He concluded that “that has to happen as an enabler. The reason why we all use computing devices in business is because it is mission critical. It has to be made mission critical and mandated for schools”.411
210.The on-screen numeracy and literacy assessments in Wales are a positive example of the use of ‘adaptive’ testing. In an adaptive test,
“an algorithm controls the difficulty level of the questions. Each student faces different questions according to how well they are doing during the assessment. If the student is doing well and answering questions correctly, the difficulty of the questions increases. However, if the student struggles to answer questions correctly, the algorithm presents slightly easier questions.”412
211.According to the e-Assessment Association and the Cambridge Centre for Evaluation and Monitoring, adaptive testing has a number of advantages. These include:
Limitations to adaptive tests are that they can only realistically be delivered digitally and can therefore include only computer markable questions with a clear right/wrong answer.414 The Times Education Commission saw adaptive testing as ‘blowing apart’ the secondary assessment system, arguing that it “would remove the need for all teenagers to sit the same paper at the same time across the country.”415
Box 2: The introduction of ‘personalised assessments’ in Wales
Pupils in maintained schools in Wales are required to take national ‘personalised assessments’ in years 2 to 7 (ages 6 to 13 at the start of the year) in procedural numeracy (number, measuring and data skills), numerical reasoning (solving problems) and reading. The assessments can be taken at any point during the academic year, at a time decided by the school. They must be taken at least once, and schools have the option to use them once more during the academic year.416 These assessments were previously taken on paper but since 2018 have moved to an online, adaptive format. As the assessment is personalised for each learner, there is no need for a whole class to take the test at the same time, and schools can choose to test classes, smaller groups or individual pupils as preferred, and according to their facilities. Following an assessment, feedback is provided in class by the teacher, in reports to parents and via Hwb, the Welsh Government’s digital learning platform.417 The introduction of personalised assessments is part of a move in Wales towards the greater use of assessment for learning (or formative assessment) rather than assessment for school accountability purposes.418 Prior to their introduction, the Welsh Government reiterated that the results of these assessments are not used to make judgements on school performance.419 |
212.Mr Busuttil-Reynaud emphasised the benefits of adaptive testing for formative assessment, noting that by testing pupils “on exactly the same scale” every year, the personalised assessments in Wales enable schools to see “their learning trajectory and overall performance” in literacy and numeracy.420 He suggested this could help to improve pupil performance in GCSE English and maths, particularly for those who do not achieve a grade 4 or above at 16, by identifying those who need additional support sooner. He stressed that: “The seeds of that are planted long before GCSE.”421
213.Another potential use of adaptive testing was highlighted by Dr Saxton. Currently some GCSE subjects, maths in particular, include the option to take the exam at either a higher or lower tier depending on the presumed ability level of the pupil. A lower (or foundation) tier paper features easier questions to give lower-ability pupils greater opportunity to demonstrate their knowledge.422 These lower-tier papers are capped at grade 5, but theoretically increase the chances of lower-ability pupils achieving a grade 4 or 5 than if they sat the higher-tier exam. However, Mathematics in Education and Industry argued that there are weaknesses in this model:
“The current two-tier GCSE mathematics has low grade boundaries for the lower grades at each tier so students can achieve those grades by picking up odd marks throughout the paper without demonstrating a thorough understanding of anything in particular.”423
214.Requiring every pupil to complete the same adaptive test would remove the “tricky” decision for teachers, highlighted by Dr Saxton,424 of deciding which tier paper a pupil should be entered for. Mr Busuttil-Reynaud suggested that if adaptive tests were adopted as the method of measuring literacy and numeracy skills at 16, they could provide “a straightforward assessment of [a pupil’s] capability and whether they meet that grade 4/5 threshold boundary and remove the backwash pressure from GCSEs.”425
215.Witnesses raised other important considerations regarding adaptive testing. Mr Glanville told us that as it can mark reliably only “a narrow right/wrong answer”, we must avoid “a race to the bottom” in assessment: “We must not allow it to go to narrow multiple choice or drag and drop.”426 Dr Saxton also stressed that, due to the need for an extensive question bank to ensure adaptive tests are not predictable and therefore open to malpractice, “they are incredibly resource-intensive to develop.” However, Sir Ian Bauckham identified the possibility of using AI to generate test questions in future.427
216.Paper-based exams that require pupils to write for extended periods are increasingly out of alignment with the experiences and tasks young people will encounter in their education, life and work. On-screen testing represents a more modern approach and offers huge potential to enhance the assessment experience for learners. We welcome Ofqual’s initial investigatory work in this area and support its future vision of a mixed model combining on-screen and paper-based assessment.
217.We recognise that there are numerous barriers to delivering a greater proportion of on-screen assessment within national exams at the end of key stage 4, including the need to develop school infrastructure and ensure pupils are equipped with the necessary digital skills. It is imperative that the transition towards on-screen assessment at GCSE is managed in a way that ensures fair treatment of all learners.
218.The Government should lead on ensuring that the transition towards on-screen assessment at GCSE is implemented successfully. In its response to this report, the Government should summarise the steps it is taking to support progress towards a greater proportion of GCSE assessments being undertaken on-screen in future.
219.Some have suggested that the fact that around a third of pupils do not achieve a grade 4 or ‘standard pass’ in their English and maths GCSEs each year is unavoidable due to the way GCSE grade boundaries are set. It has been suggested that grades are awarded according to a normal distribution, with fixed percentages of pupils receiving each grade each year, to ensure comparable outcomes between cohorts. 428 However, Dr Saxton confirmed that “there is absolutely no quota” for particular grades and that GCSE results “are a direct consequence of the marks that students achieve”.429 She noted that:
“It is true that before the pandemic, grade distributions were very stable from one year to the next. It does not follow, however, that this is because there is a quota of available grades. Students are not fitted to a bell curve or normal distribution. Rather, the stability in grading reflects the stability in attainment from year to year.”430
220.Sir Ian Bauckham also clarified that ensuring comparable outcomes from one year to the next does not rely on a fixed number of pupils achieving each grade:
“It is about holding the relationship between the grade and the underlying performance standard year on year. You get debasement or inflation when that relationship is not held and a grade indicates progressively lower levels of performance. Provided that a particular grade—let us say a grade B at A-level—relates to the same standard of performance year on year, not the same number of students, you hold the value of grades.”431
221.Grade boundaries are determined each year by the exam boards under the supervision and close monitoring of Ofqual. For GCSE English and maths, three factors feed into this process. First, each year 11 cohort’s performance in national maths and English exams taken at the end of key stage 2 (SATs) is compared against that of a previous cohort. The performance of the previous cohort at GCSE is then used to predict what proportion of the current cohort will achieve each grade.432 If one cohort’s English SATs results were slightly higher than those achieved by the previous cohort, for example, then the GCSE English grade boundaries for that new cohort might be adjusted accordingly.
222.Ofqual then looks at two additional factors to assess how the actual performance of pupils in the cohort compares to what would have been expected based on SATs results. The primary method is consultation with senior examiners from each of the exam boards to see whether there is evidence of an improvement or decline in performance based on exam scripts received.433 For GCSE English language and maths, Ofqual can also refer to how pupils from the cohort performed in the National Reference Test (NRT).434
223.The NRT was introduced in 2017 and aims to provide an objective, longitudinal measure of how year 11 pupils in England are performing in English language and maths. The test is effectively identical each year and is taken by a representative sample of year 11 students.435 Dr Saxton suggested that it “helps us to determine whether the underlying performance standard of 16 year-olds in this country in English and maths is changing”.436 She highlighted that the 2020 NRT recorded “a statistically significant improvement in performance in maths compared with 2017” and emphasised that this “would have been reflected in results that summer” had GCSE exams not been cancelled during the pandemic.437
224.For most GCSE subjects there is no NRT equivalent, which means evidence of improved performance can be more difficult to demonstrate.438 Grade boundaries may therefore be less likely to be adjusted. Dr Saxton suggested that being able to refer to NRT data in other subjects “would be absolutely wonderful”, but that funding for this “would not necessarily” be available from the DfE.439
225.It has been suggested that the system for determining GCSE grade boundaries requires a fixed proportion of pupils to ‘fail’ their English and maths GCSEs each year. However, we are persuaded that it does not set quotas for the number of pupils who can be awarded each grade.
226.Dr Meadows, then Ofqual’s Executive Director for Strategy, Risk and Research, told the House of Commons Education Committee in 2020 that 96% of GCSE grades in each subject are accurate “plus or minus one grade”.440 The then Chief Regulator of Ofqual, Dame Glenys Stacey, told the Committee in the same session that “they are reliable to one grade either way”.441 This analysis was based on a research paper Ofqual published in 2018. Ofqual outlined that whereas the probability of being awarded the “definitive” grade in maths at GCSE or A-level is 96%, this falls to 52% for qualifications in English language and literature.442
227.This led one witness to argue that, on average, one GCSE grade in four is wrong.443 However, this notion was challenged by Pearson:
“The probability of receiving a definitive grade varies depending on which mark you have (and how close you are to another grade boundary) and between subjects. We know that for subjects with essay questions, history and English, for example, it is reasonable for two experts to award two different marks to the same essay (25 or 26 out of 40).”444
They drew attention to comments made by Dr Saxton “in which she confirms that this does not mean that one in four grades are wrong, but that the method of assessing these subjects means that there is some expert judgement required.”445 Sir Jon Coles, Chief Executive of United Learning, challenged the 2018 Ofqual analysis itself, arguing that “it exaggerates the extent to which there is marker inaccuracy overall”.446
228.Witnesses stressed that grade reliability can be a significant issue at the level of the individual pupil, given the weight that is placed on GCSE results.447 Receiving a grade that is lower than expected in a particular subject could affect the post-16 routes available to them, or commit them to resitting their English and maths exams. Dr Meadows, now Associate Professor of Educational Assessment at the University of Oxford, commented that:
“unfortunately … a lot of weight is placed on particular GCSEs for progression, maths and English being the obvious ones. In maths that is less problematic because the assessment in maths is generally highly reliable. In English that is problematic.”448
229.Dr Meadows also argued that reaching a completely reliable measure of attainment at GCSE “in a complex area such as English literature” would be “pretty much impossible without the most extraordinarily long assessments and perhaps multiple assessments”. She implied that this would not be desirable: “It is how we use the grades that needs to change rather than creating a system of lengthy assessments.”449 Pearson highlighted that an alternative way to improve reliability is to base assessments on multiple choice or short answer questions only. They noted, however, that for subjects such as English, this “would be of detriment to the learner experience of the subject and would not allow [pupils] to express their analysis or conclusions in long form.”450
230.Dr Saxton told us that quality assurance of grading in England is “as good as it can be” and “up there with the best standards around the world”. She highlighted that Ofqual has taken steps to improve transparency. For example, pupils are allowed to see their marks in a subject as well as their overall grade, and schools can now request to review a pupil’s GCSE exam script if they are considering appealing a grading decision.451 She also noted that technology has enabled advancements in marking, with scripts now being “divided up into questions and marked on-screen”. She argued that this process “protects the student from the particular approach of one marker versus another”.452 Ms Hague raised similar points.453
231.Ms Hague also suggested that AI could play a role in the marking of GCSEs in the future, explaining that in the United States “automated marking is used extensively.” She noted that the technology can assist with identifying anomalies in marks awarded: “You might have it marked twice by AI and then identify differences and use human scorers to check.”454 Dr Saxton suggested in a recent speech that AI “has a place” and could support “quality assurance of human marking and spotting errors”.455 Ms Fletcher was also positive about this potential use of AI, but stressed the need for regulation of the technology. This would be necessary to ensure that “the algorithms at its root and the way it is developed are genuinely equitable” and that it can effectively assess the skills and knowledge being measured. In particular, she noted that while existing AI tools can successfully mark shorter answers, they are less effective in assessing long-form or creative work.456
232.Where external assessment is used, it is vital that GCSE grades are, and are perceived to be, accurate reflections of a pupil’s performance. This is particularly important given that, at present, a pupil’s GCSE results can have a direct impact on the post-16 options they are able to pursue. Robust processes must be in place to assure the reliability of grading. These should be kept under review, including in light of the potential enhancements that emerging technologies such as AI may bring to marking.
233.We heard that taking steps to increase the reliability of GCSE assessment would be likely to entail more extensive testing of pupils, or moving to a more limited set of question types. Such changes could have a detrimental impact on learners and would run counter to efforts to transition to a more varied and less onerous assessment system at key stage 4.
234.The Government should instead prioritise lowering the stakes of assessment at 16, to ease the pressure for testing at this age to meet such high reliability standards, and reduce the present emphasis on exam-based assessment at the end of key stage 4.
303 See paras 46–50.
304 Roundtable discussion with pupils (20 September 2023): https://committees.parliament.uk/publications/42297/documents/210208/default/
309 Roundtable discussion with pupils (20 September 2023): https://committees.parliament.uk/publications/42297/documents/210208/default/
315 Institute for Government, The exam question: changing the model of assessment reform (August 2022), p 6: https://www.instituteforgovernment.org.uk/sites/default/files/publications/exam-question.pdf [accessed 1 December 2023]
316 See paras 238–43.
321 Institute for Government, The exam question: changing the model of assessment reform (August 2022), p 7: https://www.instituteforgovernment.org.uk/sites/default/files/publications/exam-question.pdf [accessed 1 December 2023]
322 Ibid., p 4
327 The Times Education Commission called for the introduction of “a personal online portfolio for every student. It would include academic qualifications alongside a record of other achievements: video footage of a pupil playing a musical instrument, photographs of projects they have worked on or details of expeditions, volunteering and work experience.” Times Education Commission, Bringing out the best (June 2022), p 40: https://nuk-tnl-editorialprod-staticassets.s3.amazonaws.com/2022/education-commission/Times%20Education%20Commission%20final%20report.pdf [accessed 1 December 2023]. A similar recommendation from the Independent Review of Qualifications and Assessment in Scotland, which reported in June 2023, was that “all learners should have a digital profile to allow them to record achievements”. In evidence to the review “the idea of a technology-based profile … was perceived to be the natural solution for future learners to present their achievements.” Scottish Government, It’s our future: report of the independent review of qualifications and assessment, (June 2023), pp 37 and 67: https://www.gov.scot/binaries/content/documents/govscot/publications/independent-report/2023/06/future-report-independent-review-qualifications-assessment/documents/future-report-independent-review-qualifications-assessment/future-report-independent-review-qualifications-assessment/govscot%3Adocument/future-report-independent-review-qualifications-assessment.pdf [accessed 29 November 2023]
333 ‘Private school to move further away from ‘outdated’ GCSEs, headteacher says’, The Independent (11 September 2023): https://www.independent.co.uk/news/uk/gcses-english-private-school-daniel-daylewis-geoff-barton-b2408894.html [accessed 29 November 2023]
335 ‘Latymer Upper School ditches GCSEs for its own qualification’, The Times (12 September 2023): https://www.thetimes.co.uk/article/latymer-upper-school-ditches-gcses-for-its-own-qualification-zwdqn3z08 [accessed 29 November 2023]
337 Ibid.
338 Times Education Commission, Bringing out the best (June 2022), p 39: https://nuk-tnl-editorial-prod-staticassets.s3.amazonaws.com/2022/education-commission/Times Education Commission final report.pdf [accessed 1 December 2023]
339 The International Baccalaureate (IB) offers four educational programmes: the IB Diploma Programme for students aged 16 to 19; the IB Middle Years Programme (MYP) for students aged 11 to 16; the IB Primary Years Programme for children aged 3 to 12; and the Career-related Programme, also for students aged 16 to 19. An IB programme is offered in 137 schools across the UK, including 58 state schools. However, only 30 schools offer the Middle Years Programme (MYP), 18 of which are state schools. Written evidence from International Baccalaureate Organization (EDU0079). Data available at International Baccalaureate Organization, ‘Find an IB World School’: https://ibo.org/programmes/find-an-ib-school/ [accessed 30 November 2023]
340 Institute for Government, The exam question: changing the model of assessment reform (August 2022), p 1: https://www.instituteforgovernment.org.uk/sites/default/files/publications/exam-question.pdf [accessed 1 December 2023]
341 Q 15 (Sharon Hague, Gavin Busuttil-Reynaud), Q 137 (Dr Jo Saxton), Q 141 (Sir Ian Bauckham) and written evidence from Department for Education (EDU0085)
342 Department for Education, ‘The Advanced British Standard: Everything you need to know’ (5 October 2023): https://educationhub.blog.gov.uk/2023/10/05/the-advanced-british-standard-everything-you-need-to-know/ [accessed 6 November 2023]
343 Department for Education, A world-class education system: the Advanced British Standard, CP 945 (October 2023), p 7: https://assets.publishing.service.gov.uk/media/651d36656a6955001278b293/A_world-class_education_system_-_The_Advanced_British_Standard__accessible_.pdf [accessed 29 November 2023]
344 Ibid., p 38
345 Ibid., p 39
346 EDSK, Examining exams (April 2023), p 16: https://www.edsk.org/wp-content/uploads/2023/04/EDSK-Examining-exams.pdf [accessed 7 November 2023]
347 Ofqual, Review of controlled assessment in GCSEs (June 2013), p 11: https://assets.publishing.service.gov.uk/media/5a8092ff40f0b62305b8bf22/2013–06-11-review-of-controlled-assessment-in-GCSEs.pdf [accessed 6 November 2023]
350 Ofqual, ‘Get the facts: GCSE reform’ (updated 26 January 2018): https://www.gov.uk/government/publications/get-the-facts-gcse-and-a-level-reform/get-the-facts-gcse-reform [accessed 29 November 2023]
351 These contain a speaking assessment that makes up 25% of the overall grade.
354 Ibid.
357 Ibid.
358 Rethinking Assessment, A blueprint for change (2023), pp 11–13: https://drive.google.com/file/d/1NoyUc3P-wFi_LCJS0SlzzUrQxkLC2WsG/view [accessed 6 November 2023]
362 Roundtable discussion with pupils (20 September 2023): https://committees.parliament.uk/publications/42297/documents/210208/default/
363 Roundtable discussion with teachers (14 September 2023): https://committees.parliament.uk/publications/42296/documents/210207/default/
364 Ibid.
365 Teacher Tapp, ‘Ofsted, your colleagues and a return of GCSE coursework’: https://teachertapp.co.uk/articles/ofsted-your-colleagues-and-a-return-of-gcse-coursework/ [accessed 6 November 2023]. This survey question was answered by 5,137 teachers, with results weighted to reflect national school and teacher demographics.
367 Evidence relating to the reliability of grading in exam-based assessment is discussed in paras 226–34.
368 Data available at Joint Council for Qualifications CIC, ‘Examination results’: https://www.jcq.org.uk/examination-results/ [accessed 10 October 2023].
369 EDSK, Examining exams (April 2023), p 4: https://www.edsk.org/wp-content/uploads/2023/04/EDSK-Examining-exams.pdf [accessed 7 November 2023]
373 EDSK, Examining exams (April 2023), p 3: https://www.edsk.org/wp-content/uploads/2023/04/EDSK-Examining-exams.pdf [accessed 7 November 2023]
377 AQA, Level 2 Higher Project Qualification, p 7: https://filestore.aqa.org.uk/subjects/AQA-W-7992-SP-19.PDF [accessed 8 November 2023]
378 AQA, ‘Introduction’: https://www.aqa.org.uk/subjects/projects/project-qualifications/PQ2-7992/introduction [accessed 8 November 2023]
380 AQA, Level 2 Higher Project Qualification, p 2: https://filestore.aqa.org.uk/subjects/AQA-W-7992-SP-19.PDF [accessed 8 November 2023]
381 Department for Education, ‘Key stage 4 qualifications, discount codes and point scores’ (updated 28 September 2023): https://www.gov.uk/government/publications/key-stage-4-qualifications-discount-codes-and-point-scores [accessed 8 November 2023]
382 Data available at AQA, ‘Results statistics’: https://www.aqa.org.uk/exams-administration/results-days/results-statistics [accessed 8 November 2023] and Pearson, ‘Grade statistics’: https://qualifications.pearson.com/en/support/support-topics/results-certification/grade-statistics.html?Qualification-Family=Project-Qualification [accessed 8 November 2023].
383 Ofqual, ‘Infographics for GCSE results, 2023 (accessible)’ (24 August 2023): https://www.gov.uk/government/publications/infographic-gcse-results-2023/infographics-for-gcse-results-2023-accessible [accessed 8 November 2023]
384 BERA, ‘The Higher Project Qualification: It doesn’t matter what you’ve learned. At the end of the day, it’s the grade’: https://www.bera.ac.uk/blog/the-higher-project-qualification-it-doesnt-matter-what-youve-learned-at-the-end-of-the-day-its-the-grade [accessed 8 November 2023]
385 We understand that only three GCSE examinations currently involve an on-screen component: the GCSE computer science exams offered by the Pearson and WJEC Eduqas exam boards, in which candidates are required to use the Python 3 programming language during the exam; and the WJEC Eduqas GCSE geology exam, in which candidates complete multiple choice questions, and short, structured and extended written answers on-screen. AQA, England’s largest exam provider, announced in October 2023 that it would introduce digital assessment for some components of its GCSE Italian and Polish exams by 2026, and incorporate digital assessment into “at least one of the large entry subjects”, for example GCSE English, by 2030. AQA, ‘Click to the future: exams to go digital to better prepare the workforce of tomorrow’: https://www.aqa.org.uk/news/click-to-the-future-exams-to-go-digital-to-better-prepare-the-workforce-of-tomorrow [accessed 8 November 2023]
387 Times Education Commission, Bringing out the best (June 2022), p 40: https://nuk-tnl-editorial-prod-staticassets.s3.amazonaws.com/2022/education-commission/Times Education Commission final report.pdf [accessed 1 December 2023]
388 Institute for Government, The exam question: changing the model of assessment reform (August 2022), p 11: https://www.instituteforgovernment.org.uk/sites/default/files/publications/exam-question.pdf [accessed 1 December 2023]
392 Ibid.
393 AQA, On-screen exams: what school leaders, teachers and students think (July 2022), p 7: https://filestore.aqa.org.uk/content/news/G02404-ON-SCREEN-EXAMS-REPORT-2022.PDF [accessed 25 September 2023]
394 Ibid., p 20
395 AQA, On-screen exams: what school leaders, teachers and students think (July 2022), pp 7–8: https://filestore.aqa.org.uk/content/news/G02404-ON-SCREEN-EXAMS-REPORT-2022.PDF [accessed 25 September 2023]
398 Ofqual, ‘Dr Jo Saxton’s speech at the Wellington Festival of Education’ (7 July 2023): https://www.gov.uk/government/speeches/dr-jo-saxtons-speech-at-the-wellington-festival-of-education [accessed 8 November 2023]
400 Ofqual, Online and on-screen assessment in high stakes, sessional qualifications (December 2020), p 25: https://assets.publishing.service.gov.uk/media/5fd361b7e90e0766326f7f6e/Barriers_to_online_111220.pdf [accessed 8 November 2023]
401 Ibid., p 4
402 Written evidence from National Association of Special Educational Needs (EDU0038) and Association of School and College Leaders (EDU0029)
404 Ibid.
406 Ofqual, ‘Dr Jo Saxton’s speech at the Wellington Festival of Education’ (7 July 2023): https://www.gov.uk/government/speeches/dr-jo-saxtons-speech-at-the-wellington-festival-of-education [accessed 8 November 2023]
410 Ofqual, Online and on-screen assessment in high stakes, sessional qualifications (December 2020),
pp 5–6: https://assets.publishing.service.gov.uk/media/5fd361b7e90e0766326f7f6e/Barriers_to_online_111220.pdf [accessed 8 November 2023]
412 Cambridge University Press and Assessment, ‘5 reasons to use adaptive tests’: https://www.cem.org/blog/5-reasons-to-use-adaptive-tests [accessed 8 November 2023]
413 e-Assessment Association, ‘Adaptive Testing’: https://www.e-assessment.com/news/adaptive-testing/ [accessed 8 November 2023] and Cambridge University Press and Assessment, ‘5 reasons to use adaptive tests’: https://www.cem.org/blog/5-reasons-to-use-adaptive-tests [accessed 8 November 2023]
414 e-Assessment Association, ‘Adaptive Testing’: https://www.e-assessment.com/news/adaptive-testing/ [accessed 8 November 2023]
415 Times Education Commission, Bringing out the best (June 2022), p 40: https://nuk-tnl-editorial-prod-staticassets.s3.amazonaws.com/2022/education-commission/Times Education Commission final report.pdf [accessed 1 December 2023]
416 Welsh Government, ‘Personalised assessments: information for parents and carers’: https://hwb.gov.wales/curriculum-for-wales/reading-and-numeracy-assessments/personalised-assessments-information-for-parents-and-carers/ [accessed 8 November 2023]
417 Ibid.
418 Welsh Parliament, ‘Personalised Assessments: Assessment for learning not accountability?’ (updated 27 May 2021): https://research.senedd.wales/research-articles/personalised-assessments-assessment-for-learning-not-accountability/ [accessed 8 November 2023]
419 Welsh Parliament, Kirsty Williams AM, para 220: https://record.senedd.wales/Plenary/4293#C16488 [accessed 8 November 2023]
421 Ibid.
428 The Times Education Commission, for example, described this as “an inevitable consequence of grade boundaries, set to ensure that a certain proportion of pupils get each mark every year.” Times Education Commission, Bringing out the best (June 2022), p 23: https://nuk-tnl-editorial-prod-staticassets.s3.amazonaws.com/2022/education-commission/Times Education Commission final report.pdf [accessed 1 December 2023]
430 Letter from Dr Jo Saxton to Lord Johnson of Marylebone Chair of the Education for 11–16 Year Olds Committee (24 July 2023): https://committees.parliament.uk/publications/41096/documents/200101/default/
432 Ofqual, ‘Levelling the playing field’ (24 March 2017): https://ofqual.blog.gov.uk/2017/03/24/levelling-the-playing-field/ [accessed 20 October 2023]. This article discusses predicting A-level grades, but Ofqual confirmed to us that the process is the same for GCSEs, using key stage 2 results.
433 Ofqual, ‘Prediction matrices explained’ (21 April 2017): https://ofqual.blog.gov.uk/2017/04/21/prediction-matrices-explained/ [accessed 20 October 2023]
434 Ofqual, ‘How do we achieve fairness in exams?’ (26 April 2019): https://ofqual.blog.gov.uk/2019/04/26/how-do-we-achieve-fairness-in-exams/ [accessed 20 October 2023]
435 Ofqual, National Reference Test information (updated August 2023), p 3: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/1179254/NRT_2019_General_Report_background_-_updated2023__1_.pdf [accessed 29 September 2023]
437 Letter from Dr Jo Saxton to Lord Johnson of Marylebone Chair of the Education for 11–16 Year Olds Committee (24 July 2023): https://committees.parliament.uk/publications/41096/documents/200101/default/
438 Ofqual, ‘How do we achieve fairness in exams?’ (26 April 2019): https://ofqual.blog.gov.uk/2019/04/26/how-do-we-achieve-fairness-in-exams/ [accessed 20 October 2023]
440 Oral evidence taken before the House of Commons Education Committee, inquiry on The Impact of COVID-19 on Education and Children’s Services, 2 September 2020 (Session 2019–21), Q 997 (Dr Michelle Meadows)
441 Oral evidence taken before the House of Commons Education Committee, inquiry on The Impact of COVID -19 on Education and Children’s Services, 2 September 2020 (Session 2019–21), Q 1059 (Dame Glenys Stacey)
442 The term ‘definitive’ is “based on terminology ordinarily used in exam boards for the mark given by the senior examiner”. Ofqual, Marking consistency metrics (November 2018), p 4 https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/759207/Marking_consistency_metrics_-_an_update_-_FINAL64492.pdf [accessed 29 September 2023]
445 Ibid.
449 Ibid.
452 Ibid.
454 Ibid.
455 Ofqual, ‘Dr Jo Saxton’s speech at the Wellington Festival of Education’ (7 July 2023): https://www.gov.uk/government/speeches/dr-jo-saxtons-speech-at-the-wellington-festival-of-education [accessed 29 September 2023]