Schools straitjacket themselves into levels; SLTs should cast off the shackles.
Rebecca: There is prodigious danger in the seeking of loose spirits. I fear it, I fear it. Let us rather blame ourselves and-
Putnam: How may we blame ourselves?
Arthur Miller’s The Crucible, Act One
Are rumours of the death of national levels greatly exaggerated? Since 2010, the DfE has announced the intention to remove and not replace them. They have now done so at Key Stage 3. As long ago as 2008, national tests at the end of Key Stage 3 (ages 11-14) were scrapped. But as education blogger and school leader Keven Bartle graphically points out in his magnificent post Spirit Levels: Exorcising The Ghost of Assessment Past, old regimes die hard:
‘National Curriculum Levels are dead. In secondary schools, at KS3, they have been dead for years now. Nobody much will miss NC Levels. They were a curriculum assassin and an accountability albatross rooted in suspect reliability and minimal validity. For teachers they gradually telescoped from being an end of key stage annoyance (their original intention) to being a half-termly assessment burden to being a daily lesson-by-lesson imposition to finally being a bitesize, Ofsted-friendly, 20 minute nugget of learning soundbite surrender.
And yet!!! Like Hamlet’s father or Banquo the ghost of levels walks among us still, torturing us with the feeling that we should be doing something about it. They sit on our shoulders as we write schemes of work. They whisper in our ears as we construct lesson plans. They speak through us at Parents’ Evenings like Patrick Swayze communicates through Whoopi Goldberg. They are dead and yet they will not go away.’
The reasons why levels loom on to haunt us still are understandable. To borrow from Kev, the DfE are caught between the Scylla of dependency on accountability and the Charybdis of mistrust in teacher judgement. Many SLT and teachers appear unable to think outside of leveled assessment. Kev again illuminates:
‘SLT, particularly at KS3, have failed to rise up to the challenge presented by the absence of KS3 tests. This is hardly surprising given the retention of the annual end-of-KS data collection phantasm and the eternally spectral nature of the Ofsted inspection regime (and here we have to remember that Ofsted are as reliant upon these ghosts as we are, given the limited nature of their inspection processes). But perhaps, at the very least, we ought to have exorcised a reliance on levels at years 7 and 8 where they have always been an unwanted visitation in many respects (remember the days before the Optional Tests anyone?).
‘Teachers, sometimes ghostly themselves in their removal from the decision-making processes nationally and even locally, have seemed to find solace and company in the presence of the eerily familiar. How many of our brightest and best teachers have grown up in a profession dominated by levels and grades, unable to conceptualise a time when they simply didn’t exist, and where students learned qualitatively rather than progressed quantitatively.’
Another great blogger, Caroline Osborne, says that though levels are ‘unwieldy, untrustworthy, unfathomable and, ultimately, unworkable’, ‘like heroin addicts sitting in a side-room of Lloyds Chemist’s, awaiting their Methadone dose, some school leaders and teachers seem to be struggling to imagine the horrible cold-turkey-effect of a life without levels’.
She shares a revealing conversation:
Me: Of course, levels no longer exist so a lot of that data is now irrelevant.
Boss: We’re still using the levels next year: we must have some way of assessing the kids…
Me: Even the sub-levels?! But they’ve all been scrapped!
Boss: Yes. (slightly irked) We haven’t been given anything else…
‘How may we blame ourselves?’ This conversation is now being repeated in schools up and down the country. Life after levels is where SLT fear to tread. An analogy from a Guardian Secret Teacher strikes me as apt here:
“We are like prisoners abandoned in a cell for so long that they no longer need to lock the door. Rattle the cage and we might discover that the door swings open and the warders have left.”
Schools and SLTs feel they lack the time and capacity to reinvent the entire assessment system. Frankly, it is also a matter of willpower: little matters more than how pupils’ learning is assessed.
In this climate, in which even the greatest schools I have visited or taught as, such as Burlington Danes, Hammersmith, King Soloman, Westminster, and Dunraven, Streatham, remain thoroughly wedded to their zombie brides of national levels, it is worth exploring why levels are so fatally flawed and damaging for teaching and learning. With this in mind, we might have the courage of conviction required to reinvent assessment.
Tom Bennett is scathing on this:
1. The descriptors are so vague they suggest Alzheimer’s.
2. Different teachers can give the same piece of work a different level.
3. Different schools can give the same piece a different level.
4. They invite subjectivity of analysis so advanced that it makes the scoring on Come Dine With Me look impartial.
5. Allowing teachers to assess levels and then have them assessed themselves on reaching those targets, is an invitation to inflate levels. You have created an incentive to cheat, exaggerate, or merely massage. Even without such obviously diabolic motivations, bias creeps in as preferred children are given the benefit of the doubt, and outcasts are not.
6. The ghastly culture of sublevels, and levelled homeworks, and everything that its architects Wiliams and Black didn’t intend.
‘Levels are a damn mess; a pulpy wet love letter to wooly ideology. They assess nothing. They are pseudo assessment. They are cargo cult assessment. They are runes. I would take a hammer to the whole rotten cabinet. I would pull the lever myself, and dance as it dangled. Levels have been one of the most harmful ways that education has been metrified and commodified. This yardstick of progress and achievement is so unrelated to anything real that it becomes a perverse, surreal straitjacket within which we lash ourselves.
‘Paralysed by freedom, by inertia, until schools take the lead in this, they’re in limbo’.
Levels are not only broken; they never worked to begin with. On this I have been heavily influenced by Daisy Christodolou, and on whose unpublished paper I now heavily draw. She has been thinking about these ideas more than anyone else I know in the education sector, and her paper will be well worth reading on this when it comes out. Distilled from Daisy’s ideas, here are four of their most fatal flaws:
1. Imprecise: Levels can’t measure precise granularity
2. Misguided: Levels are founded on the flawed premise of generic content
3. Engulfing: Levels crowd out other forms of assessment
4. Distorting: Levels warp the curriculum
Level descriptors are deliberately vague and supposedly unprescriptive. As Tim Oates, head of Cambridge Assessment says:
‘Level descriptors in secondary Chemistry state that pupils must understand ‘that there are patterns in the reactions between substances’. Seemingly innocuous due to its generic character, this is, in fact, highly problematic. This statement essentially describes all of chemistry. So what should teachers actually teach? What are the key concepts which children should know and apply? The concept of entitlement becomes seriously eroded, if not absent, from a National Curriculum formed of such generic statements. Assessment becomes highly problematic, since a clear specification of what should be assessed becomes impossible. Assessment degrades into ‘ambush assessment’ since learners and teachers may not, in the school curriculum, have focussed on that which appears in a specific national test. Frantic search, by teachers and parents, for past test papers thus ensues, and the curriculum degrades into ‘that which will be assessed’ (Mansell W 2007; Stobart G 2008). Assessment developers, teachers, pupils and parents all are disadvantaged when this occurs.’
Teachers then get lumped with the adverb and adjective problem. Differences between levels in analytical skills are vaguely defined with words like ‘analyses clearly,’ ‘analyses precisely,’ ‘analyses with sophistication.’ It is very hard for teachers to know the difference with any real accuracy, let alone pupils or parents.
Levels are based on a fallacy: the fallacy that skills are generic and can be taught in the abstract. As we know from decades of cognitive science, this notion is mistaken. Detaching skills from content doesn’t work.
In English, level descriptors are misguidedly designed to apply to any text and any question. They supposedly assess pupil’s analysis of a range of texts, from a cereal packet to The Tempest. But of course, even the same question (‘How does the author achieve certain effects in the text?’) is easier when it is asked of a cereal packet, and harder when it is asked of The Tempest. Not only that, but one teacher could set the essay question: ‘How is tension used in Act 1 of The Tempest?’ and another could set the essay question: ‘Compare how the themes of power and authority relate to the characters of Prospero and Caliban as The Tempest develops’, and the second would clearly be harder. Yet all these essays could be marked with level 7s being awarded for analysis, which makes a nonsense of the entire conception.
In Maths, levels are misguidedly designed for linear progression. But there are glaring inconsistencies about content across both numerical levels and mathematical domains. Within the same numerical level, you would classify both these questions as being of equal difficulty: What is 50% of 100? and What is 93% of 459? Most pupils, calculator or not, would get the first question right. Not nearly as many would get the second question right. Yet both of these, according to the levels, count as level 5 questions. Across domains, it is far easier for pupils to access level 6 statistics than level 6 algebra. So Maths teachers know in observations it is better to be teaching stats than algebra to demonstrate seemingly extraordinary ‘progress’. Again, this makes a nonsense of the entire conception of levels.
Levels encourage the mistaken notion that high-level skills are unrelated to the underpinning knowledge, and they offer no help in assessing such underpinning knowledge.
Levels crowd out other useful forms of assessment. The only form of assessment most schools are using is generic level descriptors. But these level descriptors cannot possibly take into account the huge range of possible tasks. So levels are being required to produce fine detail even though they are ill-equipped for the task. If a school only uses NC levels to mark work – as most state schools do – then there is much that is of value that can never be captured by these levels.
One example of this is multiple-choice questions. Very few English, History or other Humanities departments use multiple-choice options as regular formative assessment. Everything hinges on complex, high-level tasks. But complex, high-level tasks is not the only (or even the most effective) vehicle to assess underlying knowledge and understanding. In English, it doesn’t diagnose precisely enough what pupils don’t know about grammar, plot, context or character, and targets remain generic: ‘add more punctuation’ or ‘analyse more precisely’. Fine-grain assessment is crowded out by high-level assessment.
Such high-level, complex, ‘open’ tasks are easy to set, but very difficult to mark. When testing complex skills such as reading, writing and mathematical problem-solving, making judgments is hard. We compromise dependability and accuracy in assessment if we rely solely and exclusively on such a vehicle. All-engulfing levelled assessments every half-term or every fortnight compromise precise diagnosis and feedback.
Levels distort what we actually teach, and result in contorted curriculum sequences.
In English, levels mean we neglect grammar, because it is only one of twenty or so ‘assessment foci’. Grammar then only gets taught in the odd starter activity here and there, without the dedicated time and attention devoted to mastering its fundamentals. Perhaps this is part of the reason why so many sixth formers still struggle with punctuation, and children in English classrooms round the country are telling their teachers: ‘I don’t know the difference between colons and semi-colons.’ Levels warp us away from the focus on mastering the fundamentals.
In Maths, levels mean we move on to new topics too rapidly. Maths is sequenced in blocks vertically across domains so as to plough remorselessly on to a new topic every two weeks, heedless of who’s forgotten what about the last or previous topics. Pupils only spend a couple of weeks a year on fractions, and perpetually rehash the same forgotten material year after year. Perhaps this is part of the reason why children in Maths classrooms up and down the country, are telling their teachers: ‘I can’t do fractions.’ Levels combined with a much-vaunted spiral curriculum are a recipe for forgetting.
As I explained in my last post, this assessment regime is not designed with memory in mind. Remembering depends on sufficient attention, storage, usage and transfer. Insufficient attention is given to the underpinning fundamentals – level 3 calculation drills and ‘low-level’ grammatical, textual and contextual knowledge – which are neglected as a result of levels.
Flawed arguments for sticking with levels
There are a number of notions suggested for why we’d better stick with the devil we know:
‘OFSTED need to know national levels.’
No, they don’t, especially not at Key Stage 3. This is a stock response but it is mistaken. Accountability is focused at Key Stage 4 – Key Stage 3 is up to us.
‘There’s no alternative.’
Yes, there is. Many independent schools have never used levels, only percentages. I am writing about one alternative for my next blogpost.
‘We don’t want to confuse parents with different assessment systems in different subjects.’
They’re already confused. Few parents understand the difference between levels 4 and 5 across subjects, let alone between 4a and 5c, not should we expect them to: the distinctions are bureaucratic, arbitrary and distracting.
‘We don’t want to lose comparisons between departments.’
We never had meaningful comparisons. MFL started at level 1 at Key Stage 3. Maths went up to level 8 but English rarely got on to level 7. Levelled content rarely made coherent sense.
‘We don’t want to lose data we can use to benchmark our progress nationally.‘
Assessment is overloaded with purposes – to compare, to benchmark – but most important of all is to help learners improve. If we want to benchmark effectively we must separate out that purpose, just as we have separate PISA assessments for international comparisons.
‘It’s too much time and energy to develop and retrain staff into a new assessment system.’
This is what’s really behind such reluctance to budge on this: a lack of willpower. That’s why my next post will explain why those who redesign assessment will reap the rewards.
Halycon age or halcyon cage?
Teacher bloggers Kev Bartle, Caroline Osborne and Chris Hildrew have set out the challenge ahead:
- ‘Build something magnificent on top of the unvisited grave of National Curriculum Levels.
- Identify alternative methods of evaluating student learning
- Ask our departmental leaders to use the new emphasis on knowledge in the revised national curriculum to identify expected learning at each year group in KS3
- Muster all our professionalism and create a form of assessment at Key Stage 3 that is very much alive and kicking.’
‘There is plenty of mileage in Joe Kirby’s mastery model, but it needs flesh on the bones to become a viable proposition. Curriculum levels have been wrestling with the problem since 1989 through APP, end of key stage testing, teacher assessments, progress maps and sublevels – and, judging by the reaction to their demise, they have failed. I hope that whatever we design to replace them can do a better job but I, for one, am daunted by the prospect.’
As Caroline Osborne says, ‘What replaces levels must be simple and truthful. Is such a system possible?’
A reinvigorated system of assessment is possible, and we should more excited than daunted at the prospect. For it is nothing less than a chance to demonstrate the professionalism of the teaching profession. For too long schools have straitjacketed themselves into a flawed assessment system, and OFSTED and SLT’s prodigious fear of them have kept those shackles on. It is time to lay these spirit levels to rest. It’s time to let drop the albatross, rattle the cage, cast off the shackles, undo the straitjacket and leave the open jail cell. It’s time to explore an alternative assessment system, the subject of my next blogpost: a mastery model.