‘[It] isn’t designed to be assessed how we assess’: rethinking assessment for qualification in the context of the implementation of the Curriculum for Wales

This paper reports teacher and learner perspectives on how assessment and reform influences pedagogical practices and behaviours. The research was conducted in a context of policy reform, at a time when Wales’ revised General Certificate of Secondary Education (GCSE) specifications had been implemented, and learners were preparing for their assessments; but, also during the period of debate on the development of Wales’ new curriculum, which has taken a distinct and contrasting position on assessment to the assumptions underlying the reform of Welsh GCSEs implemented from 2015. These data, therefore, offer unique insights into the affordances and limitations of two sharply contrasting systems at a time of considerable change, offering reflections on the current curriculum and its attendant assessment practices, and also a prospective analysis of how the principles embedded in the new curriculum could challenge these existing assumptions and conventions. Findings suggest that teachers and learners currently inhabit an assessment-driven system, which encourages performative practices in pedagogy and is governed by external accountability; and that these practices are at odds with the principles of assessment articulated in Successful Futures . Consequently, teachers in this study expressed uncertainty about how assessment for certification purposes at GCSE could be compatible with the principles of the Curriculum for Wales.


Introduction
It has recently been argued that the history of Welsh education since devolution in 1999 can be characterised by three distinct policy 'movements' (Egan, 2017;Davies et al, 2018;Connolly et al, 2018), each of which have taken specific positions on the linked issues of agency, assessment and accountability. Davies et al. (2018) provide an overview of these three movements, noting that the early years of the National Assembly for Wales witnessed an 'experimental' approach in the newly created policy 'laboratory' of Welsh devolution (Reynolds, 2008;Moon, 2012); this period being marked by a loosening of external accountability, with the abolishment of SATs and school league tables Connolly et al., 2018). Power (2016) further suggests that the Welsh approach since devolution has been based on a 'high trust' rather than 'mistrust' approach to the teaching profession ). Yet, from 2010 onwards a second phase of policymaking was initiated: poor PISA results in 2009 prompted a reappraisal, by the then Minister for Education Leighton Andrews, of the approach taken during the first decade of devolution. In his much-cited speech of February 2011, Andrews described education in Wales as a 'complacent system' and signalled a 'new approach to accountability' (Andrews, 2011;Davies et al., 2018). Davies et al. (2018) have argued that this period leveraged greater external accountability and challenge into the Welsh system, citing key developments such as the introduction of national testing for 7 to 11-year old pupils, the consolidation and increased centralisation of school improvement services, and the introduction of performance-related 'banding' (subsequently 'categorisation') of schools from 2011 onwards.
This period also saw a review and radical reform of qualifications, with Wales moving away from the 'three-country' regulatory settlement on General Certificate of Secondary Education (GCSE)s which had existed between Wales, England and Northern Ireland since 1988 (Andrews, 2014, p. 169; Barrance & Elwood, 2018, p. 252;Barrance, 2019). At this time of heightened external accountability, in Welsh education, the reform to GCSE specifications which followed this review also arguably injected greater externality into assessment procedures for some subjects. This was achieved through limiting controlled assessment, and aligning pupil outcomes with skills measured by PISA (Estyn, 2018), which has itself been described as a key measure of system-wide external accountability (Egan, 2017, p. 4). Since 2016, the third policy 'movement' identified by Davies et al. (2018) and Egan (2017) has signalled a shift back towards 'trust' (Power, 2016), and a proposed relaxation of some of these measures of external accountability (National Assembly for Wales, 2018). There has also been a commitment that assessment policies and practices should be 'uncoupled' from high-stakes accountability and external monitoring (Welsh Government, 2019, pp. 38-39), in the context of the implementation of the new curriculum, which has been under development since a landmark review of the Welsh curriculum conducted by Professor Graham Donaldson in 2015. The extent to which assessment for certification can realistically be 'uncoupled' from curriculum development a delivery remains a major concern.
The development of the new Curriculum for Wales is broadly line with a wider shift to what Shapira and Priestley (2018, p. 75) term a new 'genericism' in curriculum design, where curricula are increasingly being designed to be outcome-and competency-focussed (see Donaldson, 2015, pp. 5-6). Donaldson argues that, going forward, articulations of progression in the Welsh curriculum should encompass a broader range of outcomes than just 'narrower measures of attainment', noting that outcomes should be described from the learner's point of view, with assessment conducted using a 'a wide repertoire of assessment techniques matched to the different outcomes' (Donaldson, 2015, pp. 56, 79). He further recommends a move towards long-term reference points, developed as achievement outcomes for each of the progression steps in the new curriculum: progression is to be measured at three yearly intervals, allowing teachers' scope to plan learning and assessments. The introduction of progression steps will remove reference to the Foundation Phase and Key Stages as part of the new curriculum (Welsh Government, 2019). Progression steps will be introduced at ages 5, 8, 11, 14 and 16, and while statutory guidance has not yet been published, research conducted as part of the CAMAU project (Hayward et al. 2017) notes that in each Area of Learning and Experience achievement outcomes will be developed as broad expectations, rather than a prescriptive list of knowledge and skills. The CAMAU project also suggest that progression of learning should be used in a more formative manner and incorporate assessment for learning to support pupils, and their teachers, to make pedagogical decisions on how they can be supported to improve. Achievement outcomes will, therefore, encompass a series of statements describing achievements by pupils at each progression step, and teachers will decide which of the statements captures pupil performance most appropriately (Welsh Government, 2019). The Welsh Government (2019) note that this approach will empower schools to develop their own approaches to assessment and learning. The Welsh Government (2019) has further suggested that, under the new curriculum arrangements, Head teachers will be empowered to set these achievement outcomes against each of the new progression steps.
As noted above, Donaldson's review is also vocal about the impact that highstakes and external assessment, focussed on small range of outcomes, can have on the breadth of the curriculum, most notably when it bears a close relationship with measures of external accountability (Donaldson, 2015, pp. 6, 75 and 79). The review also calls for teacher assessment to be the 'main vehicle' for assessment before qualification (Donaldson, 2015, p. 80), although it acknowledges the issues of validity, reliability and alignment with external benchmarks, which have been a cause of concern in Wales for some time (Andrews, 2011), and calls for support for teachers and robust moderation to address these.

The reform of GCSE specifications in Wales and the debate over alignment with the new curriculum
Given such a direction of travel with regard to assessment policy and practice, it is therefore inevitable, and generally acknowledged, that further reform of GCSE qualifications will be required to ensure alignment between the principles which have guided the development of the new curriculum, and those which will form the basis of certification and its associated assessment norms and practices at (what is currently) Key Stage 4 (Qualifications Wales, 2019).
By contrast, the reforms to GCSE in Wales, implemented in 2015 during the second phase of devolved policymaking centred on developing more 'rigorous' qualifications (Andrews, 2014, p. 170), leading to substantive changes in course and assessment structure (Barrance & Elwood, 2018); changes which would now appear to be significantly at odds with the principles of assessment espoused in the new curriculum. From 2015, tiering was retained only in exceptional cases, limits were placed on controlled assessment, with a greater dependence on external examinations (Barrance & Elwood, 2018). For example, the English Language GCSE specification was adapted so that only 20% of the marks are awarded for controlled assessment (a reduction from 40% in previous specifications) with the remaining 80% awarded for externally set examinations (Welsh Government, 2014).
Public-facing external accountability with regard to GCSE grades, it has been suggested, also led to a disproportionate emphasis being placed on the outcomes of narrow groups of learners on the C/D borderline, focussing much of the attention of teachers on pupils who would raise school attainment figures with marginal improvements in their performance (Welsh Government, 2019). Such an emphasis can clearly produce problematic consequences, such as failing to encourage all learners to reach their full potential (Donaldson, 2015), or in promoting surface learning, and encouraging pedagogical practices which focus on test-specific measures (Koretz, cited in Ofqual, 2016). Much of the challenge for the implementation of the Curriculum for Wales will be the move from high-stakes assessment to a system that focusses more on the needs of individual pupils through formative assessment and assessment for learning (Donaldson, 2015;Welsh Government, 2019). And, while a wide range of evidence can be called upon to suggest that moves towards foregrounding formative assessment may well benefit learners (See Dann, 2002;Bennett, 2011;Black, 2015;Hopfenbeck & Stobart, 2015), there are clearly concerns regarding the validity of such assessments (Andrews, 2011;Stobart, 2012), as well as the ways in which significant changes such as these will be implemented in practice in schools (Priestley & Minty, 2013). This process could well be further problematized through the staggered implementation of the new curriculum beginning in 2022 with learners in Year 7 who will consequently complete Year 11 studies in 2027 (Qualifications Wales, 2019). This will lead to dual systems of accountability and measurement working in tandem in schools until the new curriculum is fully implemented. Wales: pressure and performativity? Elwood, Hopfenbeck and Baird (2017) note that high-stakes summative assessments can increase pressure on young people, obscuring the 'purpose' of schooling as learning, and reframing their educational experience as being geared towards 'performing' in examinations. Consequences of such an ethos can include greater assessment-related anxiety (Sommer & Arendasy, 2015), a tendency to adopt performance goal orientations (von der Embse et al., 2018) and a trend towards surfacelevel learning rather than deep-level learning (Vandewalle et al., 2019). As regards teachers, the literature on teacher orientations and experiences in such environments points towards performative practices being routinely employed, including teaching to the test by narrowing the curriculum and prioritising assessed content (Popham, 2000;Segal et al., 2017), as well as the excessive use of past papers as an assessment preparation technique (Firestone et al., 2004;Crocker, 2005). These practices, which can be attributed to the coupling of assessment outcomes and accountability for teachers and schools as a result of standards-based reforms, have resulted in systems where assessments exercise a disproportionate influence on what is taught in the classroom (Brown, 2004). Furthermore, literature on teachers' practices has also examined how high-stakes assessment, and externally mandated systems of quality assessment, accountability and governance not only serve to encourage performative and strategic instrumentality in pedagogical practice, but also pose a very real threat to the 'identity' of teachers as agentic professionals and ethical subjects, empowered to make decisions based on their professional judgement (Ball, 2003;Holloway & Brass, 2018).

Materials and methodology
This paper reports on qualitative data gathered during the second phase of a wider, explanatory sequential mixed-methods study (Creswell & Plano Clark, 2011), the aims of the which were to examine stakeholders' experiences and perspectives on the reformed GCSE and A-Level specifications in Wales, and to evaluate the impact of these reforms on classroom practices. The study also specifically sought perspectives on how the new curriculum under development, and its position on assessment, would likely influence further reform of GCSE qualifications, and consequently teachers' and pupils' assessment practices and learning behaviours-namely, the focus of this paper. Data were collected from four curriculum pioneer secondary schools in Wales, over a period of 12 months between 2017 and 2018. The schools which participated were, therefore, highly attuned the complexities and challenges of developing the new curriculum, and embedding its principles in their arrangements for delivery, pedagogy and assessment. Schools involved in this phase of the research were selected using a purposive sampling method (Cohen et al., 2018), with the sample generated to gather perspectives from a diverse range of such settings across all regions of Wales, encompassing both Welsh-and English-medium perspectives. The data presented in this paper are part of a wider research project that included questionnaires, interviews and focus groups with a range of key stakeholders, namely pupils, parents and teachers. Overall, 45 teachers, 244 pupils and 107 parents provided data as part of the study. To ensure the validity of research instruments a process of rigorous piloting was conducted. The pilot study involved a sample of 25 pupils and 5 teachers, who participated in pilot interviews/focus groups and questionnaires. The responses to the pilot study were evaluated for reliability and validity and research instruments were adapted prior to the completion of the main study.
The data presented in this paper were collected via 10 focus groups of GCSE and A-Level pupils (a total of 33 pupils), and 15 semi-structured interviews with classroom teachers. Pupils and teachers were also invited to participate based on a purposive sampling strategy, with the intention of obtaining a range of perspectives from both GCSE and A-Level pupils in different year groups, and from teachers of a range of subjects. Table 1 identifies the number of participants from each of the four data collection settings. The data presented in this paper are drawn from a relatively small subset of the main data set, and while we do not claim that the data can be generalizable, the data are offered as an exploration of situated insights that may have a wider application.
Focus groups were deemed appropriate for data collection with pupils as they allowed for the research agenda to be driven by issues that pupils perceived to be important to their experiences (Cohen et al., 2018), and to allow for dynamic social interactions (Morgan, 1996) where perspectives could be co-constructed, challenged, explored or refined through interaction. Focus groups were also considered to have the affordance of being a comfortable and informal context for pupils to articulate their views (Darbyshire et al., 2005). By contrast, semi-structured interviews allowed teachers to explore, in a professional discursive space, issues relating to assessment that they considered to be important, and allowed for individualised, in-depth accounts of their own experiences (Newby, 2013). The qualitative data were analysed using a progressive coding procedure, starting with basic in vivo and descriptive codes, which identified commonalities in the original words and descriptions used by participants. The second stage of the coding procedure was to generate composite codes, where these commonalities were systematically combined and categorised. In these stages, pupil data and teacher data were analysed and coded separately, allowing the opportunity for distinct perspectives to be offered by these two groups and to encourage authenticity in the representation of differing perspectives. The final stage of the coding procedure was to develop more abstract codes designed to build theory and draw inferences (Braun & Clarke, 2006;Saldaña 2009). It was, at this stage, that the teacher data and pupil data were directly compared, to allow for a more complete and theoretical understanding of the experiences of both teachers and pupils in relation to the reform experience.
As noted above, the data gathered during this study were collected at a very specific juncture of reform, when the post-2015 GCSE specifications had been implemented for the first time, and when the schools involved in the study were also deeply involved in planning their future provision in response to the principles of new curriculum. As such, the data offer a unique 'bifurcated' insight on two highly contrasting systems: reflections in situ on the current curriculum and its attendant assessment conventions, and the impact these have on pedagogy; and, also an ex ante analysis of how the principles of the new curriculum could pose challenges to assessment practice.

Inhabiting an assessment-driven system
Throughout the research, pupils showed a keen awareness of the assessment-driven climate within which they were learning, and of the performative practices that this engendered. They spoke extensively of their experiences of a system, which many felt reduced 'learning' to a narrow suite of circumscribed assessed outcomes, particularly during the later years of their education as they progressed towards assessment for certification. Pupils also spoke critically of what they deemed to be the 'manufactured' nature of some exams. Many felt that the full range of their subject-specific knowledge or skills was not being appropriately assessed; rather, they felt that they were being assessed on their ability to replicate contrived answers and re-enact answering strategies they had rehearsed previously in class: Reflections from teachers also supported a view that the Welsh education system had moved towards engendering an assessment-driven ethos. Teachers identified increasing 'top-down' pressure to meet and exceed attainment targets, and accountability measures and inspection procedures as increasingly influential in prioritising assessment outcomes, above other indicators and outcomes of pupil learning.

Assessment, performativity and preparation
Furthermore, pupils also elaborated on what many saw as the excessive use of assessment-specific preparation methods used in schools, most notably teaching to the test and the frequent use of past papers. They spoke of, in some cases, years of strategic preparation for GCSE assessments, starting as early as Year 8 for some pupils.
Pupil G: when it comes to the actual [science] assessment, I would say it would probably be more or less every lesson.

Pupil I: I feel like English give you a lot of time… and they started preparing us for maths I would say as soon as we went back [in the new academic year in September].
Pupils in every focus group cited past papers as a primary method used by schools to prepare pupils for assessments. They spoke of an excessive cycle of 're-doing [and] re-doing' 'past paper after past paper after past paper', and had begun to identify that they felt the use of past papers were 'the only way [they] get through' their assessments.

Pupil B: Past papers, revision sessions, revision lessons, erm, working during free lessons on revising subjects through erm past papers mainly.
Pupil H: if you have a past paper, they will say 'right do this by Monday' and then you mark it and you go over it, so whatever you do wrong, you go over it and you know how to improve, and then you just re-do it, re-do it until you get it.
One teacher outlined a relentless regime of practicing using past papers and engaging in a strategic 'guessing game' as to the likely content of each paper, immediately prior to an exam. As noted above, pupils were also very self-aware about such performative practices which involved teaching to the test, the prioritising and foregrounding of content likely to appear in the assessments, and preparation routines within which they were engaged with their teachers. In fact, some pupils were among the first cohort to take the reformed GCSEs, and many felt that their learned reliance on regimes of preparation using past papers had left them in a difficult situation, guessing the likely areas which would appear in the assessments but with limited precedents from which to take their cue: When reflecting on their own practices some teachers also felt that they had little choice but to prioritise content that they thought would be assessed because of external standards and accountability measures.

Experiences of reforms
As noted above, some pupils also articulated their perception that they were disadvantaged by the changes, owing to a lack of assessment precedents to study. And, while Barrance and Elwood (2018, p. 255) note that examination results can often dip suddenly following assessment and qualification reform, and improve gradually thereafter, our study makes no claims about the validity of this perception as expressed by participants. However, we do note that a lack of past papers for assessment did seem to make it difficult for some pupils and teachers to engage in the familiar preparation regimes outlined above: Teachers also identified the relationship between classroom teaching and assessment content, with one openly identifying that the assessments drove the curriculum, particularly in times of educational change, and acknowledged that this experience was 'miserable for the pupils'.
Teacher F: When there is a new change, there's always a tendency for teachers to teach for exams, and that is the problem, so although it might get them good grades at the end, it's incredibly miserable for the pupils themselves […] I think that is the danger, and especially for the new cohort that went through last year, it was very much taught for the exams rather than teaching them, […] it was very much exam-based driven, the lessons, so we worked through a lot of past papers, erm, and I, as I said, I think that is mainly due to teachers not perhaps being as confident with the new syllabus, they wanted to make sure they were covering everything in the exam.
The same teacher also commented that:

Perceived challenges presented by the Curriculum for Wales
Teachers participating in the study were invited to reflect on the changes that the new curriculum would bring about. The responses given by teachers throughout this section are in response to a number of interview questions which aimed specifically to elicit teachers' perspectives on how the reformed specifications at GCSE A number of general risks and challenges were identified, which included the potential for tokenistic and surface-level implementation, and a lack of time and capacity to undertake the type of involved, collaborative work curriculum development involves: Teacher B: I am a little bit worried as to how much individual schools have the capacity to change and reform and that […] possibly where opportunities might be lost for the sake of tokenism.
Teacher J: I think one of the challenges that we are facing as a school is not so much erm, teachers' attitude, but whether or not they get enough time allocated to be able to collaborate, to come up with new schemes of work and a new way of thinking.
Teacher B: So for example, I'm the head of maths, let's go down Donaldson, I'm the head of maths and numeracy, have I actually fundamentally changed anything by doing that? No, but I've hit that tick box for Donaldson. As regards assessment specifically, the prospect of a new curriculum with a contrasting ethos, and attendant assessment practices from those currently in use, did cause some concern. Teachers seemed uncertain as to how increased autonomy in relation to curriculum delivery and teacher assessment could be faithfully aggregated into system-wide measures which were consistent and mutually cognate: one teacher noted that Donaldson's (2015) call for the curriculum to assess a wide range of competences, using a variety of assessment methods, could prove difficult to transpose to the sphere of assessment for certification: Although teachers acknowledged that the structure of qualifications would require reform to align with the new curriculum, they found it difficult to conceptualise an approach based on areas of learning as conducive to deep-level learning of subject areas and preparation for formal assessments. Teachers found it difficult to envisage what a GCSE assessment system which was not standards-or exam-driven would look like in practice, and expressed scepticism that creativity and teacher agency would be sustained during the latter stages of compulsory education when the focus returned to the outcome of assessments. Beyond this, one teacher suggested that, while the principles behind the new curriculum were a welcome addition to twenty-first century schooling in Wales, the lack of measurability of the four curriculum purposes and the overarching ethos of the curriculum, would ultimately lead to a system built on tokenism and 'lip service'.

Discussion and conclusion
The data presented in this paper from the experiences of pupils and teachers corroborate received accounts of education in Wales being in transition between two contrasting systems (Egan, 2017;Davies et al., 2018;Connolly et al., 2018): from one which prizes high levels of accountability, externality and which encourages performativity to one which will ostensibly offer greater levels of curriculum freedom, teacher autonomy and agency, and less high-stakes assessment (Donaldson, 2015;Welsh Government, 2017). Pupils in this study echoed the views of participants in Elwood et al. (2017)'s study, articulating a clear sense that their educational experiences during their final years of compulsory schooling were too keenly focussed on performance rehearsal for examinations, which they did not always feel reflected the full range of their learning. This paper offers evidence that the accountability frameworks under which schools operated, and the reforms to GCSE and A-level assessment implemented from 2015 have led to an 'assessment-driven curriculum' (Brown, 2004), which prompted teachers to engage in performance-driven classroom practices designed to improve visible and externally mandated measures of progress and school standards-often against the professional and pedagogical judgment of teachers themselves. Teachers spoke of their powerlessness to act in ways that they saw as meeting what they regarded as the 'true' learning needs of their learners, echoing Ball's (2003, p. 216) contention that systems which encourage high levels of performativity can undermine teachers 'as ethical subjects' who find 'their values challenged or displaced'. The performative practices, which teachers perceive as central to the effective implementation of the reformed specifications, are clearly misaligned with the principles of assessment articulated in Successful Futures.
Somewhat understandably, teachers in this study found it difficult to conceptualise a system of assessment, in what is now Key Stage 4, which is not only simultaneously aligned with the Curriculum for Wales, attuned to contextual progression steps and distinctive curricula, but also able to be meaningfully consistent and cognate across Wales, for the purposes of GCSE certification. This perception of an 'un-assessable' (Teacher H) curriculum will need to be considered carefully and addressed as the next stages of implementation are planned. A negotiated position will need to be articulated precisely, which clarifies the proposed link between assessment and certification, beyond the initial commitment to uncoupling assessment and external accountability. And, while some teachers clearly do seem beleaguered by a system where assessment and qualification lead pedagogy, some seem also uncertain and cautiously sceptical about the prospect of system where pedagogy and curriculum are wholly unburdened by the requirements of qualification, and where learning outcomes may not always necessarily lend themselves to cognate assessment and comparison across contexts.
In response to Welsh Government, Qualifications Wales have recently set out the timetable for development of new GCSE qualifications, to be delivered from September 2025, with the qualifications themselves, in all likelihood, being finalised at least 12 months prior to implementation (Qualifications Wales, 2019). Yet, in spite of what appears to be a consensus that greater alignment is needed to ensure that any revised qualifications do not undermine the integrity of the curriculum, there are clear risks that are posed by the timing of the changes. Priestley and Minty (2013, pp. 49-50) note that in Scotland some secondary schools operated a 'wait and see' approach to developing the senior phases of the Curriculum for Excellence, until after detailed qualification specifications had been published. One of the potential risks of the current situation in Wales is that some teachers and schools may be unwilling to invest themselves conceptually and practically in the ethos of the new curriculum until they understand fully the requirements of the revised qualifications, and their associated assessment modes. This study also suggests that there is also the potential for weak, tentative or surface-level implementation; or by contrast, we note that curriculum implementation in schools could come in two phases: one prior to the development of the qualifications where the principles of the curriculum may be fully embraced; and another following the publication of the qualification specifications, where teachers and schools may be tempted to review their assumptions and practices hitherto, to realign them strategically with assessment goals in the new qualification specifications. In any case, as we move, ostensibly at least, to a curriculum and set of assessment and certification practices that purport to take a more rounded view of learner achievement, schools and teachers in Wales will need to be reassured that they can safely move away from the types of strategic, instrumental and performative pedagogical behaviours that are described above, and which, we contend, can be understood as their response to a system which has historically encouraged such practices. Doing so will involve truly uncoupling assessment from the pressures of external accountability, in order that assessment can have at its heart the needs of Welsh learners and can be in line with the principles of the Curriculum for Wales.