Assessment after levels
This was published under the 2010 to 2015 Conservative and Liberal Democrat coalition government
Nick Gibb outlines the support given to schools as they move away from level descriptors.
Reform is an important think tank because of its clear commitment to public service reform and the fact that its reports are always based on strong evidence.
When this government came into office in 2010, there was no public service area in greater need of reform than our education system. Our ambitious programme of reform, based on increased academic standards, high aspirations for every pupil and professional autonomy for schools, is now beginning to bear fruit.
The proportion of 6-year-olds able to decode simple words and pass the phonics screening check at the end of year 1 increased from 58% in 2012 to 74% in 2014. That means 102,000 more 6-year-olds are on track to be reading more effectively as a direct result of this policy. The number of pupils attending good or outstanding schools has increased by more than a million since 2009 to 2010.
Since 2010 we have effected a major, evidence-driven renewal of the content of the national curriculum, drawing on the study of the curriculum content of high-performing countries. These changes have restored the vital content lost in the 2007 changes to the national curriculum, and are the basis for a more secure assessment system that no longer uses the notion of ‘levels’.
Levels were introduced with the new national curriculum in 1988. They were devised with the intention of delivering an assessment system which measured pupils’ progress against a national framework.
But international comparisons tell us that fast-improving countries around the world do not use levels - Singapore does not, Finland did not during its time of rapid improvement, Hong Kong does not, nor does Massachusetts. And it’s not that they are missing something. Wroxham Primary School in Potter’s Bar - with a demanding, mixed intake - doesn’t and hasn’t used levels as it has made its journey from ‘special measures’ to repeated judgments of ‘outstanding’.
What these other nations do, and what effective schools in England do, is focus on the specifics of key areas of the curriculum, and ensure deep, secure knowledge and understanding in these specifics. Levels have been a distracting, over-generalised label, giving misleading signals about the genuine attainment of pupils. They have driven undue pace as Ofsted insisted on ‘progress against levels’. They have resulted in a lack of trust between primary and secondary schools and they have clogged up the education system with undependable data on pupil attainment.
The lightbulb moment for me came when visiting a primary school I saw a poster on the wall with 16 words labelled as level 5 words, the implication being that just learning these words would result in a pupil becoming level 5.
As Tim Oates of Cambridge Assessment - and who chaired our curriculum review - points out, any assessment is a claim. A claim that a person can do ‘a’ or ‘b’; that they know and understand ‘x’ and ‘y’. These are claims which have consequences - that a person is ready to study medicine or engineering, that they are ready to progress into advanced study in languages; that they know things. These claims need to be well grounded. If they are not, pupils will be misled about their own progress and we will all be misled about the quality of schools.
But what things does someone with the label ‘level 3’ actually know; what things can they do? In the drive to use levels for school improvement, for measures of national standards, for school inspection, we lost sight of what kind of claim we were making with levels and how reliable they are.
We lost sight of the purpose of the national curriculum as a clear statement of necessary content. The revised national curriculum reinstates the distinction between the national curriculum - the content of the core academic curriculum; and the school curriculum - the broader curriculum and activities of a school which should be left to the discretion of teachers.
We lost sight of the original aims of the national curriculum - enshrined in law - that all children should have access to all of the content of the national curriculum. The models of progress, ability and assessment which came to dominate the system were focused not on ensuring access and attainment for all, but on getting a small annual increase in the numbers gaining level 4 at 11, and grade C and above at GCSE.
The education system lost sight of the need for a genuine conversation between parents and schools: to help parents support their children on this part of reading, or that part of maths; not focus on a blanket judgment that ‘your child is level 3, and that’s fine for a child of their age.’
The curriculum review - resulting in the revised curriculum implemented in September 2014 - returned both the national curriculum and its assessment to a focus on necessary detail in key subjects. The new primary curriculum delivers ‘fewer things in greater depth’ which we see in the most effective primary curricula around the world. The year-by-year structure of the revised primary curriculum makes clear what needs to be taught in each subject. The secondary curriculum also focuses on key concepts, key principles, fundamental operations and core knowledge - developed in conjunction with leading academic authorities in each subject.
This clarity and precision in the national curriculum provides the basis for sound and dependable assessment. Clarity in the things being assessed is key to validity in those assessments. One headteacher in an outstanding primary school said, ‘now we can get back to assessing what is really important: the things which the children understand and can do.’ In short, levels were just too vague and imprecise. They were misleading as to what pupils knew and could do. The use of levels was pushing pupils on to new material - in the name of pace - when they had not adequately understood vital content, and had serious gaps in their knowledge. We had a system swimming in defective data on attainment and failed to see that our legal commitment to giving all children access to all of the national curriculum had been compromised.
In any assessment system, if the same claim - such as ‘this child is level 4’ - means something different on different occasions and different people mean something different by it, then one of the first principles of assessment is being infringed.
‘This child is level 4’ could come from getting a certain number of marks on a national test at key stage 2. But different children could gain these marks in very different ways. Levels are not trusted by secondary schools, since pupils described as level 4 were supposed to meet a common descriptor, but frequently had serious gaps in their reading, or in their capability in maths. Using APP - Assessing Pupil Progress - a level 4 could come from ‘best fit with the level 4 descriptor.’ This was very different from getting the label level 4 from the score on a national curriculum test and very different from those pupils who just scraped over the level 4 threshold.
Levels were too vague to enable parents to understand what they needed to do to support their children’s education. Level 3 may have appeared reassuring to parents, but it failed to communicate that parents might encourage their child to undertake wider reading, or practise an aspect of maths, or discuss with them a particular topic.
Also, ‘levels’ were originally designed to avoid children self-labelling - a situation where a child says ‘I am grade E; I always have been and always will be.’ This capping of aspiration reinforced dysfunctional ideas of ‘fixed ability’. But levels provided exactly the same kind of problematic labelling - ‘I am level 3b and all my friends are level 4.’ And being level 4 at age 14 is not the same as being level 4 at age 9.
For the last 2 decades, schools have argued that central prescription and formal assessment have impacted negatively on standards. We have responded to this - removing levels decreases central prescription in the way schools teach and assess.
We have moved the system away from misleading general labels, to assessment which is educationally well grounded and of value. We have listened to experts and, as the review panel recommended, have shifted the purpose of assessment away from ‘assigning a “best-fit” level to each pupil, to tracking which elements of the curriculum they have adequately achieved and those which require more attention.’
I recognise that removing levels requires schools to develop their own assessment schemes, matched to the timing and content of their school curriculum. This requires effort and expertise. But we have taken heed of the profession’s concerns that too much of the school curriculum and its assessment previously has been determined centrally.
The workload survey we conducted recently showed that data collection was the single biggest contributor to excess teacher workload, reported by some 56% of respondents.
We have reduced prescription of assessment, focusing instead on a limited number of tests in primary and on public examinations in secondary education. There needs to be more assessment, not less - but not centrally determined and not high stakes. Schools need to develop their own assessments which provide clear evidence of attainment and progression, focused on real things: the reading of the pupil, the specifics of what they know and can do in maths, their understanding of key concepts in science and events in history. We need more assessment, but of a different kind.
Schools know when and through which contexts they are teaching particular ideas. They can then choose questions to put to pupils which match the school curriculum, measure progress, and feed back into their teaching. But those internal, formative assessments do not need to be benchmarked to some national standard. And if there is only one message you take away from this today, it should be this one. The national standard will be measured at the end of the key stages. It is for the professionalism of schools and their teachers that drives how they ensure their pupils reach those standards. The steps towards those goals will be varied and will depend on the pupils and the curriculum and the approach of the school.
I have seen very well-designed recap quizzes in schools such as in David Perks’s East London Science School - questions carefully chosen by teachers to show how well a group is progressing through a half-term’s work, and whether misconceptions or gaps are present. These are low-stakes, formative tests to inform pupils about how well they are doing and to influence the teaching. For those teachers who fear that the removal of levels means no data to feed into management systems, it is clear that such tests produce useful and informative data on attainment and progress. Such approaches are far more robust than systems recording levels.
A frequent concern has been over what Ofsted will be looking for when they inspect. I can assure all teachers, heads and governors that the Ofsted inspection framework has been changed to reflect ‘life without levels’. Training for inspectors has been revised, and they will inspect schools’ approaches to continuous assessment of pupils’ attainment and progress in the key elements of the national curriculum. There will be no expectation that formative assessment will need to be benchmarked against some national standard other than the national curriculum for maintained schools and against the school’s own curriculum in the case of academies. And I rang Ofsted yesterday just to double check it reflects their current thinking and it does.
It is early days, but I have visited schools where the changes have been enthusiastically embraced, where rich question and answer permeates class time, where workbooks encourage children to practise and demonstrate both to themselves and others their thinking, where progress is seen clearly against the revised detail in the national curriculum. In these schools there is no nostalgia for ‘levels’. In these schools there is no notion of ‘fixed ability’, but a model akin to that in schools in Singapore and Finland; namely, that a child is capable of anything, depending on how it is presented to them, and the effort which they put into learning it.
As one former teacher, Henry Fletcher-Wood, wrote recently on his blog:
I’m not arguing for less assessment, I’m arguing for frequent, useful assessment. Levels do not help a teacher or a head of department (if they did, it would be possible to explain how to support a student at ‘level 5’ in history without any further information).
What teachers need to assess on a regular basis is students’ knowledge of individual concepts and ideas, and their capacity to use that knowledge; this kind of analysis must happen at a departmental level. Question-level analysis in departments provides usable insights: if 80% of students in class A answered a question about the Blitz well, but only 40% in class B did so, it seems highly likely that the teacher of class A has done something her colleague would benefit from learning about.
Practice is an important component of both the learning process and assessment which is why we need to increase the amount of practice pupils undertake. In the same way we see the need to practise music or sport, this also applies to maths and writing. In Singapore, for example, children enjoy engaging with carefully-designed maths problems, including outside of contact time. Children who quickly grasp ideas or operations have a chance to consolidate and secure their fluency and understanding. Children who find a particular idea difficult are presented with a range of contexts, with a greater chance that one of these will be the trigger to understanding.
This carefully designed variation in problems is something which Debbie Morgan continues to research at the National Centre for Excellence in the teaching of Mathematics. It is a feature of the work of the newly established 34 maths hubs which are helping to identify the best approaches to teaching maths. The approach challenges the notion of ‘demonstrate once and move on’ - related to the idea of pace which I mentioned earlier. Practice is through well-designed activities in high-quality textbooks and student workbooks. These books reveal pupils’ levels of understanding and misconceptions and they can show how well a pupil is progressing.
Rich question and answer is a form of assessment, linked intimately with the learning process. But there is another key idea which we would wish to see embedded in assessment practice - the idea of ‘production’, a term coined by Tim Oates who has been exploring this with Dame Alison Peacock at Wroxham Primary School. He noticed that the pupils’ exercise books at Wroxham were far more substantial than in many other primary schools. The chidren wrote, drew, and produced more. What this research emphasises is that the things which pupils produce are a key to the nature of their thought, and to the way that their understanding is building.
The exercise books at Wroxham give the teachers a clear idea of where they need to direct support and of the need to present ideas in varied ways. Such books are there for Ofsted to see, to observe what rate of progress children are making, and what kind of questions they can answer.
At Westminster Academy in London, teachers have broken down the curriculum into 15 topics which are each independently assessed via an in-class quiz, homework and an end-of-term exam. A score is produced for each topic and then used to provide an average score. Teachers then use topic scores to provide support where needed.
To help schools as they develop effective and valuable assessment schemes, and to help us to identify model approaches we are today announcing the formation of a commission on assessment without levels. This commission will continue the evidence-based approach to assessment which we have put in place, and will support primary and secondary schools with the transition to assessment without levels, identifying and sharing good practice in assessment. I am delighted that John McIntosh CBE will chair the commission and look forward to sharing more details on the commission’s members in due course.
The leading US academic E. D. Hirsch wrote that:
It is the duty of schools to provide each child with the knowledge and skills requisite for academic progress - regardless of home background.
In government we have made it our aim to deliver a curriculum and assessment framework that will ensure that all pupils, regardless of background or ability, will be taught the national curriculum. We are determined to close the attainment gap between those from richer and poorer backgrounds, preparing all young people for life in modern Britain, and today’s announcement will, I believe, help schools to deliver that key objective.