Assessment and learning loss in England: never let a good crisis go to waste

ABSTRACT

Disruption of the status quo caused by the pandemic in England provides a lens to view the priorities embedded in the qualifications system. The Government’s initial priority was to avert grade inflation, followed by a populist turn to teacher assessment after a widespread backlash. Subsequent Government arguments for the return to examinations as the fairest way to assess re-introduced policies which emphasised excellence agenda. Learning loss, a consequence of the pandemic, compounded pre-existing patterns of inequality. Inclusive policies to address this in the qualification system would require fundamentally different ways of thinking compared to, for example, the current adaptations that are made for young people with special educational needs because the modern inclusion agenda is broader. The excellence agenda assumes a competitive system in which educational resources must be rationed; that there will be winners and losers. We question this logic, arguing for principles that would underpin a more inclusive qualification system. Since 2015, pupils must stay in education or training until they are 18. Thus, at very least, rationing educational access before this is unwarranted. A modern approach would be more flexible, putting learners first and embrace diversity rather than standardisation as the main principle for fair assessment.

KEYWORDS:

• Qualifications
• assessment
• social justice
• learning loss
• special educational needs
• disadvantage

Introduction

The Warnock Report (Department of Education and Science 1978) signalled a shift of systems-thinking in the support of young people, from segregation to integration and towards inclusion. Forty years on, Daniels, Thompson, and Tawell (2019) argued that the efficacy of actions identified by Warnock had been interrupted by the perverse incentives in the 1998 Education Act in England. This Act led to a tension between a concern to promote excellence through judgement and comparison of schools and implicitly, the identification of deficit and ideas of inclusive education that focused on engagement, learning and progression. These tensions lie at the heart of debates about assessment and are again emerging during Covid in the rhetoric around learning loss. This article explores how ideas of learning loss are emerging in the educational discourse in England and questions how well pre-Covid systems served those young people who are identified as being most at risk of learning loss – those who are socially and economically disadvantaged and those with special educational needs (SEND). It explores how assessment is positioned within these debates in England and identifies emerging tensions likely to interfere with the development of education systems that serve all young people well. The article concludes by setting out a series of principles that could be used to inform assessment systems that challenge ideas of learning loss and seek to be both excellent and equitable.

Learning loss in the pandemic

Physical exclusion from school rapidly became the global norm during the pandemic. OECD (2021) estimated that 1.5 billion school-age pupils in 188 countries were out of school in 2020. School closures affect many aspects of young people’s experiences, since for the most vulnerable children they may provide the child’s only nutritious meal each day and a safe, nurturing environment. Whilst schools were physically closed to pupils, remote education was provided in many countries, such as on the radio, TV or the internet (UNICEF 2021). Additionally, many schools shifted to on-screen lessons. Home schooling was prevalent. To the extent that formal education in classrooms is important, it would be anticipated that there would be a loss in learning through these closures. Learning loss has been defined as,

a reduction in the overall level of attainment that a student achieves by the end of their course of study … which is attributable to both direct and indirect impacts from COVID-19.

(Newton 2021)

Learning loss is therefore more than the number of schooldays missed; it is also the indirect effect of the pandemic on learning through a multiplicity of factors such as absence of social interaction, physical and mental health decline, additional caring responsibilities and digital exclusion. Learning loss also incorporates the effectiveness of learning during the pandemic, through whatever means that learning was gained.

In this article, we use England’s qualification system as a looking glass through which to view policy in relation to social justice in educational outcomes, especially for young people who are disadvantaged. We look at the disruptive effect of the pandemic on qualifications policy. After considering the effects of learning loss on disadvantaged groups, we look at how assessments were adapted during the pandemic, as well as considering the regular adaptations that are made for special educational needs students. Ultimately, following Hayward et al. (2022), we lay out principles that would be needed to underpin an inclusive qualification system.

Estimation of learning loss

A number of research syntheses have been published, investigating the impact of learning loss (Donnelly and Patrinos 2022; EEF 2022; Storey and Zhang 2021). Generally, research has shown weaker progress of pupils over the pandemic, though the size of this learning loss finding differs (EEF 2022; Hammerstein et al. 2021; Stringer and Keys 2021). Learning loss was generally estimated to be a few months where it was found. Effect sizes are small to moderate: Storey and Zhang’s (2021) meta-analysis found an overall effect of 0.15 standard deviation, which was statistically significant. Variability in findings between studies is to be expected, as they used different measures, included young people of different ages, were conducted in different countries, subjects and at different times during the pandemic. Missing data from those who did not submit assessments will also have affected the results of the research. Assessments taken at home, teacher grading and missing data are likely to have led to an under-estimate of the effects of the pandemic compared with the use of standardised tests. Effects of poor environments or resources on the test-taking conditions themselves (such as distractions or poor internet connection), however, would over-estimate the effects of the pandemic (Stringer and Keys 2021).

Most studies focused upon the core subjects. Mathematics learning was generally affected to a larger extent than language skills (EEF 2022; Kuhfeld et al. 2020; Storey and Zhang 2021), which is not entirely surprising because language is learned outside the classroom, in everyday life, to a greater extent than mathematics. Results of the National Reference Test, England’s test used for monitoring English and mathematics performance of 15-year-olds over successive years’ cohorts, found an increase in mathematics achievement in 2020, in keeping with the trend over previous years. In 2020 the test was taken before in late February and early March, before schools were closed. However, mathematics performance declined in 2021 when school closures affected performances (Ofqual 2021a). English results went up in 2020 and stayed stable in 2021. A separate Department for Education report using Star Mathematics and Star Reading adaptive tests was conducted (Renaissance Learning and Education Policy Institute 2022). Primary-age pupils were estimated to be 0.8 months behind in reading, whilst secondary pupils (aged up to 14 years) were estimated to be 2.4 months behind in reading and 1.9 months behind in maths. Thus, these findings contradict the pattern of mathematics being affected by learning loss more than language skills and represent larger differences than those found in the National Reference Test.

Learning loss and attainment gaps

There is also evidence for widening of attainment gaps. Learning loss is uneven, as some students were particularly affected by school closures. Here we consider evidence relating to students from disadvantaged backgrounds and students with SEND but other kinds of students, such as students from Black, Asian and Minority Ethnic backgrounds, were also likely to have been disproportionately affected.

Research found that children from disadvantaged families, such as economically disadvantaged, single-parent households, and less-educated parents, spent less time on learning activities at home than their peers during school closures due to limited access to learning resources, less parental involvement, lack of access to a suitable learning space or digital inequalities (EEF 2022). For a detailed account of the realities of digital exclusion in Italy, see Vigevano and Mattei (Vigevano and Mattei 2023, this issue): they argued that the lack of attention to the plight of young people in disadvantaged circumstances in the turn to digital education was a failure of government, since it is the legal role of the state to educate young people. Learning loss for disadvantaged students has been shown to be larger than in the general population (EEF 2022; Contini et al. 2021; Engzell, Frey, and Verhagen 2021), exacerbating the pre-existing inequalities in educational systems. In England, compared to 2019, in 2021 the gap in the percentage of students achieving grade A and C at A-level (examinations taken at age 18) between those in receipt of free school meals and those who were not, widened by more than a percentage point, when prior academic achievement was taken into account. At GCSE (taken at age 16) the gap widened by two percentage points at grade 7 and just over a percentage point at grade 4 (Lee 2021). It is likely that the approach to grading taken in 2021 masked the true extent of the learning gaps between relatively advantaged and disadvantaged students, as teachers were instructed to judge their students only on the subject content they had been taught.

Children and young people with SEND can be more vulnerable in times of crisis (ECW 2020). Although government guidance in England acknowledged that vulnerable children should attend school during national lockdowns, more than one million children and young people with SEND in England were not able to attend school between March and September 2020 (Ofsted 2021). Reasons for this related to health and care plans, parents’ concerns, and school risk assessments. By transferring learning from schools to home, students with SEND had limited access to necessary learning resources and specialist equipment, which caused difficulties for parents and children (Crane et al. 2021; DfE 2021). For children with complex needs (e.g. autism), online learning might not be a good substitute for face-to-face interaction. Additionally, the absence of school and home boundaries disrupted well-established routines and challenged some children’s social and emotional wellbeing (Canning and Robinson 2021). In a survey of parents across seven European countries, including the UK, parents reported receiving no, or insufficient, support for home-schooling children with SEND (Thorell et al. 2022).

Substantial qualitative studies have reported the psychological and mental health impact of pandemic on children with SEND (e.g. Latzer, Leitner, and Karnieli-Miller 2021; Tso et al. 2020). In general, students faced challenges in adapting to remote learning during the pandemic and most experienced worsened emotional symptoms, such as anxiety, frustration, loneliness, depression or temper (Asbury et al. 2021; Dvorsky et al. 2022; Hai et al. 2021; Ofsted 2021). Special school headteachers in the UK reported that students at special schools and colleges were on average five months behind with their emotional wellbeing and mental health (Skipp et al. 2021).

There is also evidence on the learning loss of students with SEND. At A-level, students with SEND performed better in 2019 than those without SEND when prior academic achievement was taken into account in the modelling. However, this difference declined by 2 percentage points in 2021, bringing about a deficit in progress between attainments taken at ages 16 and 18 for SEND students compared with their counterparts (Lee 2021). At GCSE (assessments taken at age 16), the gap in the percentage of students achieving grade 7 between those with SEND and those without, widened by two percentage points, when prior academic achievement was taken into account. A similar change was not observed at grade 4 or in the average grade achieved. However, the approach to grading in 2021 complicates interpretation of grade outcomes for SEND students, not only because teachers were asked to judge students on only what they had been taught but because of the risk that teachers might underestimate the attainment of SEND students (Lee and Newton 2021). For a wider sample of age ranges, teachers estimated that SEND pupils were about four months behind where they would normally have expected them to be at the end of the 2021 academic year (Sharp and Skipp 2022). This effect was far more pronounced in disadvantaged schools. Note also that the attainment gaps discussed above involved controls for prior attainment, when the disadvantage gaps were already present in those controls.

On the other hand, in the US, Lupas et al. (2021) compared students with ADHD who experienced school closures and engaged in remote learning for 3–4 months to grade-based norms on a standardised achievement test (Weschler Individual Achievement Test), and did not find significant declines in their cognitive performances. Clearly, the kind of special educational needs, the type of assessment used in the research and the circumstances of the learners are important causal factors for the results of these empirical studies. We turn next to the ways in which large-scale assessments were adapted due to Covid.

Large-scale assessment during the pandemic

Assessment results were delivered in a large number of countries, despite the pandemic. Examination boards responded to the pandemic in one of five ways (Opposs 2020; 2021), with postponement apparently being the most prevalent response in 2020 (UNESCO 2020a):

1. Final exams were cancelled, with results being based upon earlier completed exams and/or school-based assessments (e.g. France, Italy,¹ the Netherlands, Norway),

2. Final exams were cancelled, with results being based upon school and college-generated estimates (UK, Ireland),

3. Exams were replaced with on-line assessments taken at home (e.g. Columbia, Egypt, Mexico, Advanced Placement Tests, USA, Venezuela),

4. Exams were postponed (e.g. Azerbaiijan, Carribean Examinations Council, China, Estonia, Hong Kong, Ghana, New South Wales, Australia, Spain, Vietnam, Zambia) or

5. Exams were held on schedule (e.g. Finland, Germany, Hungary, Luxembourg, Poland, Slovenia).

Large-scale, high-stakes assessments are governed in a variety of ways, often being state-controlled (Opposs et al. 2020). Thus, governments had a large degree of control over how the assessments were adapted in most cases. However, national qualifications are highly controversial and embedded in cultural and historical practices (Isaacs and Gorgen 2018). Adapting high-stakes assessments rapidly is therefore a precarious political endeavour, which may explain postponement as the major strategy.

In England, A level and GCSE qualification results were produced by teacher grading in 2020 and 2021. There were fewer school closure days in 2020, but less prior warning for the change in assessment. Statistical standardisation (‘the algorithm’) was applied initially in 2020. The policy rationale for standardisation was to curb grade inflation, which highlighted the Conservative Government priority. Ensuring comparability of results with previous years is more in keeping with the excellence than the inclusion agenda discussed by Daniels, Thompson, and Tawell (2019). However, the algorithm was quickly abandoned in the face of a public backlash and unstandardised teacher grades awarded. Reasons for the backlash included the fact that already privileged independent schools benefited from a lack of statistical standardisation because their small sample sizes making statistical interventions unreliable (Ofqual 2020a).

In 2021, the Government rationale for abandoning examinations was not based only on public health concerns, it was about the unfairness of holding examinations when access to education had been so uneven (Hansard 2021 col 1173). There was little attempt to standardise teachers’ grades and no attempt to maintain standards over time. Indeed, in an attempt to make grading fairer given the uneven impact of the pandemic on learning, teachers were instructed to judge students on only what they had been taught, rather than on the whole qualification subject curriculum content. At its extreme, this policy would result in each student being assessed in a different way. Critics of the policy suggested that this approach would compound advantage, as pupils with more access to educational resources would be able to excel in the narrower domain. They also raised the risk that teachers might fall prey to unconscious bias and unwittingly widen the achievement gap between poorer pupils and their more privileged peers (see for example, Elloit Major 2021). The policy might be characterised as a populist response to the pandemic and the grading problems experienced in 2020 rather than a more philosophical shift.

As outlined above, equality analyses of the outcomes in 2021 were produced by Ofqual (the examinations regulator) for protected characteristics (Lee 2021). Making the assessments accessible to all young people over the pandemic was crucial and since so many of them were unable to access the educational resources they needed, use of teacher assessment was not likely in itself to address their needs.

Inclusion in large-scale assessments

As in many countries, there are serious attempts to make large-scale assessments in the UK accessible in normal years. The exams regulator in England, Ofqual, publishes rules on the accessibility of qualifications which exam boards must follow (Ofqual 2022). These require examination boards to comply with equalities law, to evaluate whether qualifications have features which could disadvantage groups of learners who share protected characteristics and to remove these features if they are unjustifiable. Justifiable features are those which are considered essential to the validity of the assessment. For example, assessing pupils’ understandings of complex language may be relevant in GCSE English but unjustifiable in GCSE maths. These rules are supported by recently published guidance (Ofqual 2022), which provides advice about how to create accessible assessments, for example by setting out best practice in item writing.

Further, examination boards have a duty, under the Equality Act (2010), to make reasonable adjustments for students with disabilities taking their qualifications. These are known as access arrangements. They are intended to allow candidates with special educational needs, disabilities or temporary injuries to access the assessment and show what they know and can do without changing the demands of the assessment (JCQ 2022). Access arrangements take a wide variety of forms; for example, 25% extra time may be granted, a modified paper may be provided or a reader or scribe employed. Ofqual specifies the limits of such adjustments, where necessary, to protect the validity of the qualifications (Ofqual 2017). For example, learners with disabilities cannot be exempt for more than 40% of the assessment for a qualification.

However, many of the over-arching qualification design features in England, for example, modular or end of course examinations, or the amount of teacher assessment versus examination, are not under the control of the examination boards, or even Ofqual. These are matters of government policy. Government provides a policy steer to Ofqual, which then implements the policy by creating qualification design rules. Before doing so, to comply with the law, Ofqual must conduct an equality impact assessment of any proposals; the aim being to ensure that the needs of specific groups are considered in policy making. Proposals may be modified in light of evidence of likely adverse impact(s) on these groups. Unfortunately, generating compelling evidence of likely negative impacts is often difficult. For example, research evidence as to the impact of modular versus end of course examinations on outcomes is complex (Pinot de Moira, Meadows, and Baird 2020) and requires careful interpretation since any individual design may have positive or negative affects depending on the wider context in which it operates. Hence, modifications to government policy due to equalities analyses tend to be minor.

Moreover, while there is an association between certain protected characteristics and socio-economic status, socio-economic status itself is not a protected characteristic. This means it is not directly considered in equality impact assessments. The system therefore relies on government consideration of the effects of any design on the outcomes of less socially and economically advantaged learners in the initial policy development. If at this stage, consideration of their needs is crowded out by a focus on rewarding excellence, later attempts to ensure accessibility may only have limited impact.

Further, the accessibility and inclusion agendas that emerged during the pandemic were wider than could possibly be addressed by traditional adaptations of assessments. There are sound economic, social and moral reasons for increased inclusion but, now more than ever, it is needed as the ‘excellent’ students are those who are likely to have weathered the pandemic relatively well. An improving system is one that can respond to the learning gaps that will have grown. To continue with only adaptations of assessments of the traditional kind leaves the field open to the allegation of being tone deaf.

Principles to inform equitable assessment systems

Assessment has served as the backbone for the illusion of meritocracy (Sandel 2021). Distribution of educational and labour market opportunities flow from individuals’ talents and efforts under this model and structural inequalities are swept aside. Notwithstanding the radical discrepancies in individuals’ opportunities to access educational goods, the function of assessment has been to selectively ration scarce opportunities in England.

Over the pandemic, disadvantage, exclusion and alienation were felt sharply, bringing about protests such as Black Lives Matter and against the algorithm used to standardise teacher grades in England. In a US context, Cook-Sather and Mawr (2021) referred to a ‘twin pandemic’ of Covid and anti-black racism. Assessment has been viewed as both a problem and a potential policy lever to bring about inclusion and social justice (e.g. McArthur 2016; Flórez Petour and Rozas Assael 2020). Yet, assessment alone cannot solve the underlying structural inequalities in a society.

Nonetheless, assessment has had a role in perpetuation of the status quo through representation of the worldview of the powerful in their design (Cook-Sather and Mawr 2021). Assessment practices are so culturally embedded that they are hidden in plain sight; they are so accepted as the natural and logical way to operate that it can be hard for those within a system to envisage alternatives. Walton (this issue) argued that the classification logic which infuses England’s education system reinforces notions of difference and exclusion, squeezing out alternative inclusive practices. Against this backdrop though, the four nations of the UK produced national qualification results during the pandemic on the basis of teacher grades, rather than the examinations that would have been sat in a normal year. Radical and rapid change was feasible in response to the crisis, even if the form it took was hugely problematic in its first year.

Echoing Machiavelli, Winston Churchill when working to form the United Nations after WWII is reported to have argued, Never let a good crisis go to waste (see Gruère 2019). The radical change in the approach taken to qualifications raises a number of interesting issues in the context of learning loss that are worthy of exploration. For example, the changes to the qualifications process that occurred during the pandemic offer a unique evidence base to explore assessment practices and their impact on social justice.

In 2020, originally the grades submitted as evidence of student achievement were standardised to reflect patterns of attainment in previous years. This process was undertaken to promote comparability of the pattern of results over time and between schools and colleges, which in normal years would also give rise to public confidence in the consistency of standards. However, evidence emerged that students from privileged, independent schools had been advantaged by the standardisation process (Ofqual 2020b), a finding that evokes Cook-Sather and Mawrs’ (2021) argument that assessment advantages the powerful. By the 17th of August newspapers in England were reporting public outcry.

As bitterly disappointed students take to the streets … .calls are issued for Williamson [the then Minister for Education] to quit … ..

and within days the tide had turned:

All A-level and GCSE results in England will be based on teacher-assessed grades, the government confirms in a spectacular U-turn.

The Guardian, 17th Aug 2020

Teachers’ submitted grades were accepted and became the qualification results.

The public outcry was reportedly driven by a concern about injustice. Nonetheless, the increase in outcomes associated with the award of unstandardised teacher grades was appreciably greater for independent schools than for comprehensive schools and other types of centre (Ofqual 2020b). The relative advantage conferred to independent schools was, in fact, greater than had standardised grades been awarded.

The situation where the procedures of examination boards were reversed in favour of teacher judgement led to a pattern of results that differed substantially from the pattern that would have emerged had the intended standardisation process been followed. Further, the pattern of results observed in 2021, the year in which substantial learning loss was experienced, was even more different from the patterns seen in the years preceding the pandemic. The trend of even higher grades being awarded to students in independent schools compared to those in comprehensive schools and other types of centre continued (Ofqual 2021b).

Nonetheless, the change in grading outcomes offers an evidence base to explore the extent to which what is described as learning loss mattered and, more fundamentally, to question whether many of the assessment assumptions built into educational systems are exclusive by design. Some described the different patterns as grade inflation, where individual learners were awarded grades by their teachers that in previous years, they would not achieved. Others argued that teacher assessment offered more individual learners opportunities to demonstrate attainment in ways better suited to them. For example, those made highly anxious by formal examinations were able to demonstrate their achievements using alternative assessment approaches in 2021.

Young people have entered further and higher education to embark on courses to which, in more traditional years, access would have been denied. There are now opportunities to track the progress of those learners whose results would have been downgraded by standardisation in 2020 but, where teacher grades opened up different opportunities for them. In a context where the learning experiences of the 2021 cohort are perceived to have led to learning loss, the pandemic has opened up questions of what matters and whether the qualification system has, in pre-pandemic years excluded young people who could have made good progress in further or higher education but were never give the chance. In essence, to investigate the extent to which current high stakes assessment practices are themselves problematic and whether changed practices could reposition high stakes assessment as a potential policy lever for inclusion and social justice (e.g. McArthur 2016; Flórez Petour and Rozas Assael 2020). Whether or not these opportunities will be taken remains an open question.

Undoubtedly, education systems are influential nation-building tools (Green 1997), with assessment setting the agenda for the programme. Using our qualification system, then, as a looking glass, what does it tell us about the kind of state England is in? Our qualification systems have, as Daniels, Thompson, and Tawell (2019) argued, focussed for a number of decades upon excellence. Attention is drawn to learners who are likely to do well. What of the forgotten 40% (DfE 2021) who are not likely to reach the same levels of attainment?

During the pandemic, through digital education, teachers were confronted with some of the lived realities of their students’ lives, seeing into their homes, which had become their physical learning spaces. In the minds of many teachers, assumptions about attitudinal, individualised causes for attainment (or lack thereof) are likely to have been displaced by structural, societal causal explanations. In Argentina (Gomez Caride 2023, this issue, this issue), ‘disconnected’ students in the Buenos Aires area who were not accessing education physically or digitally, were provided with an army of tutors, who went to students’ homes with educational resources. This policy is a reminder that governments can respond to learning loss and disadvantage with inclusive policies, rather than sticking within the parameters of the excellence agenda.

Suffused through the qualification system is the notion of selection, in turn prompted by a view of society as having restricted resources. The 2020 and 2021 experience outlined above offers an opportunity to ask questions of the dependability of our selection processes. But there are more fundamental questions to be posed: do all of our assessments need to have a selective function? Education of the masses has been accepted public policy for over 100 years in England (Eckstein and Noah 1993) and the school leaving age has risen to 18 in the UK, so the vast majority of national assessments are not about university entrance or employment. Why would we ration educational opportunities prior to that, within a national education system? In fact, of course, we do. Young people do not have a free choice about what they would like to study, nor where, even before the age of 18. They are barred entry to certain courses, schools and colleges across the country. Other countries, such as the Scandinavian models, have far more open access for all (Tröhler et al. 2023). Assessment may not be able to solve the problem of restricted public educational opportunities but neither does it need to scaffold them.

What would a qualification system look like if it emphasised inclusive education on the principles of engagement, learning and progression (Daniels, Thompson, and Tawell 2019)? What would more socially just national assessments look like? Drawing upon Hanson (2019) and Heiser, Prince, and Levy (2017), Montenegro and Jankowski (2020) argued that equitable assessment calls for,

1. Disregarding the objectivity myth and accepting that assessment is inherently subjective and guided by the biases and experiences of those conducting assessment;

2. Varying the types of evidence used to assess learning outcomes to not privilege specific ways of knowing or preferred ways to demonstrate knowledge;

3. Including the voices of students, especially those who belong to minoritized populations or those whose voices can often be left unheard, throughout the assessment process; and

4. Using assessment to advance the pursuit of equity across previously identified institutional parameters that demonstrate disparate outcomes across student populations.(9)

The fourth objective, in particular, envisages considerable agency for the people using assessment, since Montenegro and Jankowski (2020) were writing in the context of higher education. Assessment users in national assessments may have less clout since the structures involved in assessment data use go beyond their own institutions or spheres of influence. Overall, though, the argument was that there needs to be an understanding of the context of the assessment use, the people conducting the assessment and the learners to design and conduct equitable assessment.

The 2020 and 2021 experience of qualifications in England described above demonstrates the beginnings of characteristics reflected in the Montenegro and Jankowski model (2020). The objectivity myth that national assessment processes led to qualifications that were objective and fair was rejected. Varying the type of evidence to include teacher judgement was perceived to be fairer than statistical standardisation. The voices of students, including those not traditionally heard, drove change. And notwithstanding the higher grades of students attending independent centres, young people who would have been denied access to particular courses were able to access them. What Sandel (2021) described as the tyranny of merit has been exposed and the tender shoots of a more inclusive system have begun to emerge. How might they be encouraged to grow?

Currently, in England, there is a growing concern for the post-Covid qualification landscape. Learning loss has been assessed as an inadequate construct (Harmey and Moss 2021). It is by nature a concept of deficit and cannot be used as the basis for the design of a better, more inclusive educational landscape as envisioned by Warnock (1978). This concern for post Covid qualifications to support a better educational future has emerged across the political spectrum from the National Education Union, commonly associated with the political left, to the Times Education Review (2022), commonly associated with the political right.

In a review of national assessment practices in England, Hayward et al. (2022) concluded that qualifications and assessments should not exist solely for the purposes of selection, but should support progression for all in the next stage of their lives socially, culturally and economically; a vision consistent with Warnock (1978). They argued that assessment should be inclusive in purpose, design and development. Since designing a socially just assessment involves challenging the status quo, it requires a critical perspective. Hayward et al. (2022) similarly concluded that the context of national qualifications needs to be recognised, since they are part of a broader education system and the links with curriculum, pedagogy and accountability needed to be considered when designing an equitable assessment. The 2020(b) UNESCO Global Monitoring Report on inclusion and education pointed out that,

When used for accountability purposes, such high-stakes assessments can lead to the adoption of negative practices, such as selective admission, strict discipline policies, student reassignment and greater focus and time given to those most likely to succeed. (130)

For national assessments to support progression and inclusion in England, there remains much to be done. Documented effects of current approaches have been a narrowing of the curriculum and pedagogy (Ball 2003) and off-rolling of children to avoid their inclusion in institutional accountability data. Off-rolling has a particularly severe effect upon disadvantaged pupils, since 78% of those permanently excluded from school were young people classified with special educational needs (Done and Knowler 2020). Use of assessment data in the school accountability system underpins many of these effects.

Unlike in Scotland and Wales, the English Government’s response to learning loss in the pandemic was to get back to assessment as normal, as soon as possible; arguing that examinations are the fairest way to assess. The focus on selection and standardisation of assessment as fairness are hallmarks of the Conservative Government policy. The English Conservative Government policy to address learning loss is currently to offer additional tuition,² but in a system with widespread teacher vacancies and supply teachers are difficult to come by, it is not clear how this can be effective. Further, there is no policy direction that would target the tuition to reduce inequalities.

Hayward et al. (2022) went further than previous literature, in their call for qualifications to serve the needs of the individuals, engaging students far more actively in the assessment process. This approach is consistent with ideas of assessment as transformability (Florian and Linklater 2010; Walton 2023, this issue, this issue), where the process itself acts as a medium to transform individuals’ perceptions of what it is possible to achieve. It is also consistent with UNESCO’s (2020b) Global Education Monitoring Report on inclusive education, in which concluded that

All students’ learning progress and achievement should be identified and valued, and all students should have the opportunity to demonstrate their progress and achievement. Second, assessment procedures should be complementary, coherent with the goal of supporting learning and teaching, and coordinated, avoiding segregation through labelling. Third, students should be entitled to reliable and valid assessment procedures that accommodate and, where possible, are modified to meet their needs. (130–1)

In conclusion

As we have seen, the disruption of the pandemic shone light on existing inequalities and compounded them in England’s education system. The Conservative Government’s focus upon control of grade inflation, followed by a populist, teacher assessment policy indicated their priorities. The post Covid concern with learning loss matters. Our analysis of the qualification system has shown that a return to business as usual will entrench the excellence agenda, rather than open the space for inclusive policies. What matters is far more than the current obsession in England with lower standardised test scores or fewer textbook pages covered. The issue of learning loss has long before Covid had a profound impact on the lives of 40% of young people whose learning has been lost as they transitioned beyond school with little to show for their time there formally, or from which they might build a positive future. Societies cannot afford socially, economically or morally to leave individuals behind, yet pre-Covid that was the position in England. Whilst there is nothing good about a pandemic, it does create an interruption that allows us to ask more fundamental questions about education systems and that might begin with a critical look at what is meant be learning loss, whose learning is being lost and what learning really matters. Only when these questions have been addressed, can any society design a qualification system that serves both students and society well and the real issues of learning loss begin to be addressed.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Jo-Anne Baird

Jo-Anne Baird is a professor of educational assessment at Oxford University, UK. Her research interests include systems issues in assessment, such as design, policy and structural features. She has taken a particular interest in the conceptualisation of assessment standards in national systems. Jo-Anne was previously a standing advisor to the UK Parliament's Education Select Committee.

Louise Hayward

Louise Hayward is an emeritus professor of Educational Assessment and Innovation at the University of Glasgow, UK. Her research interests include assessment and social justice. She recently led a review of the qualifications system in Scotland and has been involved in system-wide reviews in Wales, Portugal and Norway.

Michelle Meadows

Michelle Meadows is an associate professor of Educational Assessment at Oxford University, UK. Her research interests include high-stakes assessment and qualification design. She has recently undertaken a review of the introduction of standardised assessments in Flanders for the OECD. Michelle was previously a deputy chief regulator for assessments in England.

Zhanxin Hao

Zhanxin Hao is a postdoctoral research fellow at Tsinghua University, China. Her research interests include test preparation for high-stakes assessments. She gained her doctorate in educational assessment from the University of Oxford and her Master's degree from the UCL Institute of Education.

Notes

1 Socially distanced oral examinations took place.

2 https://www.gov.uk/government/publications/catch-up-premium-coronavirus-covid-19/catch-up-premium.