A fair few months ago the government announced the end of KS3 levels. Only there has been little real change – to my knowledge – in the status quo because schools are seemingly still beholden to accountability measures that link to the quantifiable measures of NC levels. Schools are still manoeuvring through the murk of change. Of course, change is uncertain and uncertainty hampers innovation. But we cannot pass up this opportunity to improve assessment. It is a great chance for each school to define their own standards and create an assessment model that best suits our students in our context.
Each subject area and each school can set their own high standards. A national one-size-fits-all approach invariably doesn’t work. In an attempt to spread assessment criteria so thinly, as to cover every assessment eventuality, NC levels (at least in English) have to be generic and vague to the point of being ineffective. We can create something better can’t we?
In our English department we are designing a draft KS3 curriculum rooted in the ‘big ideas‘ of English Literature – see here. We are also looking to identify the crucial ‘threshold concepts‘ for our students – see here. As Tim Oates advises, “Don’t assess everything that moves, just the key concepts”.
You can start identifying these crucial ‘threshold concepts‘ by working backwards. For us, we have had 6th form students lacking the basic knowledge about language, grammar and our literary tradition that they need to succeed. By thinking deeply about what our students need to flourish at KS5, and beyond, we can begin to form an apt assessment model for KS3 that is better than NC levels.
We need to have a clear thread for any assessment model that guides teaching and learning. Here is my suggested sequence for redefining each step:
Begin with the students
What tools does the ‘ideal’ English student at [insert school] possess?
Move to the ‘big ideas’
What are the key ‘threshold concepts’ in our subject discipline they must master for success?
Decide what concepts matter most and drop any extraneous content
What do we need to drop and what is essential that we deepen?
Make the outcomes match the ‘threshold concepts’
What outcomes will best ensure that students learnt the subject knowledge and ‘threshold concepts’ most deeply?
Spend time refining the criteria of each outcome
What refined assessment criteria will best focus students on the essential knowledge and skills of the subject area?
Design a formative assessment model to communicate effectively
How best do we communicate progress to students, to parents and to wider accountability systems?
Our curriculum model requires specific, precise assessment criteria for each different unit, not some vague attempt to assess everything under the sun. Therefore we need to create our own standards model. We don’t need any nationally standardised student work.
We can make our own standard models in the very first instance and then use exemplars from our own students. By creating exemplar work ourselves, as part of devising our schemes of learning, we create models and worked examples that are ideal for our teaching our new curriculum. If we need to clearly recognise an exemplar that is ‘inadequate‘ we can create such a model, which we can use that as an example that students can use to improve.
Assessment becomes rooting in pedagogy. Even creating our own standards for assessment creates opportunities for improved teaching and learning.
What about the numbers and accountability?
Firstly, Paul Black never intended NC levels to be used for formative, ongoing learning. They existed for summative key stage assessments. At some stage, a desire for accountability and endless data reporting, to recognise progress, meant NC levels and sub-levels became ubiquitous. APP became a Frankenstein’s monster of assessment: with students highlighting endless assessment grids with strands and sub-levels. Assessment for learning had ate itself and some people simply gorged on the profits.
If we stand back, we can see that the vague NC levels are something of a ‘collective myth‘. We all use these vague descriptors and present a show of apparent perpetual progress. Secondary teachers see the KS2 level and ensure little Jonny in year 7 is making sub-level progress at each data input.
The levels distort real learning.
The reality of learning in English – with different areas of knowledge and skills of oracy, reading and writing all combining in a complex web – is that any vague and generic assessment criteria is insufficient. Students go up and down any true scale of assessment depending upon the nature of the outcome. Many students can write excellent analytical essays, then subsequently struggle with writing to narrate and describe in English. An endless array of reported numbers are problematic as they wouldn’t fit the expectation of continual improvement.
If we make our assessment specific and relative to the outcome being assessed, we get nearer to something like an accurate judgement of learning. We can even move towards making our feedback relative to the expected progress of our students.
So – can we chuck out numerical levels when giving feedback to students and parents?
Undoubtedly, there will be anxiety from parents who feel they can understand the simplicity of numerical values. Only, if you ask a parent what a 3a in French or History really means they would struggle to tell you. Levels can actually inhibit home learning. If we give students quality formative feedback parents can help with home learning. If young Barbara is struggling specifically with French verbs and has a strategy to improve then that makes sense to both Barbara and her parents. A level 5a and some generic criteria can prove pretty ineffective and, at best, a distraction from the specific feedback.
Teachers too will feel anxious moving away from a numerical level model that we know (but don’t love!). We need to ensure teachers are involved in creating an assessment model from the bottom up. This will make it more likely to be understood and enacted successfully. Any lasting change in the classroom is rooted in this bottom up, inclusive approach.
We know from decades of assessment for learning research that numerical levelling can have a distracting and corrosive effect on learning. So what do we use instead? I think we need a language based formative response that relates to our own criteria and expected standards, keeping the feedback straight forward for students to act upon. We know that if students have a ‘fixed mindset‘ about their ability in a given subject then they are less likely to make good progress than if they had a ‘growth mindset‘ regarding their potential in a subject. Bald, numerical assessments in many subjects can create a detrimental ‘fixed mindset‘ because they are a blunt tool. We can make our reporting of progress to students with a ‘growth mindset‘ approach. Our language of assessment can help foster motivation in our students.
Can we not give feedback that is wholly formative at KS3? Yes we can. Can we mediate a model that retains accountability measures, whilst not being beholden to endless levels and sub-levels? Yes we can!
The HuntingEnglish draft assessment model
It is important to create an assessment model that is complex enough to identity the specific knowledge and skills we assess as teachers, but simple enough at the other end for students and parents to understand.
We can start by defining our baseline assessment. Our students come to us with KS2 SATs levels. They also undertake reading tests and cognitive ability tests. We can then define, with some degree of accuracy, the starting point of our students. We can therefore group our students into three more distinct groups, whose progress can be tracked effectively by teachers:
It is crucial to clarify that these groupings are for teachers only. Students can be easily demotivated by being labelled an ‘Emerging‘ attainer, when their best pal is a ‘High‘ attainer. Such information is useful for teachers in terms of planning for differentiation and for analysing attainment, but it is not useful for students, nor parents. Excellence is not fixed or capped so we should not report in a way that indicates that it is.
We do need to make progression comprehensible for teachers, so we can use teacher knowledge of the current levelling system with very broad strokes in the transition stage. Of course, attainment is relative. For example, young Fiona attaining a notional level 4 attainment is exceptional progress, whereas for Freddy a notional level 7 is exceptional. We can therefore create a language based feedback system relative to students who have differing levels of attainment. Progress should be scaled from the student’s starting point.
Now, we need to create a common language to define their progress. These are the suggested phrases I have drafted:
Now, we need to use these phrases relative to the prior learning of the student. If you want I use levels to define the notion. Young Freddy who is a ‘High achiever‘ as a starting point, has a nominal level 7 as ‘Exceptional progress‘, a level 6 is ‘Good progress‘ and level 5 is ‘Expected progress‘ or if it insecure it may be even be deemed ‘Below expected progress‘. Only, we need not use levels. We can set our own standards. We can remove the crutch of levels for teachers with quality standardisation of student work. Soon enough, levels disappear altogether.
Every department needs to define its standards. It should be student centred. We can ask: what knowledge and skills are essential for an exceptional [insert subject] student at [insert school]? No matter what starting point, we are aiming for every student to make exceptional relative to their starting point. Setting the highest standards possible for our students is essential. In our department, we can moderate and standardise our student outcomes as we go, as we will be following the same structure in terms on an overarching KS3 curriculum. This will likely support good practice and collaboration.
Of course, we need to define our assessment objectives and these need to hone in specifically and precisely on the essential knowledge and skills our students need. Each literary context can be unpicked, the specific aspects of the literary text etc. to ensure students grasp a core knowledge. In the past, with levelling we separated out reading and writing assessment levels. Individual outcomes, such as essay writing, were being assessed for reading attainment only, ignoring the crucial importance of writing with clear grammar and generic discourse structures etc. With the APP process we atomised the process even further. Only students don’t read or write separately, nor is writing accuracy ever wholly separate from any assessment.
I therefore propose the same assessment objectives for each and every formative unit of work and assessment. For us, these bring a focus back upon the ‘threshold concepts‘ that we deem essential for making exceptional progress in English. The AO’s are as follows:
When we give written and oral feedback to students on their unit assessment objectives we can avoid numbers – giving them digestible, useful feedback. If we return to the exemplar of Freddy. He could be given feedback on his Shakespeare essay assessment that informs him of his progress relative to each AO and future targets for improvement can be set focusing upon each AO. No hunt for a level, just useful formative advice.
We are not moving wholly away from reporting any assessment that creates numerical data. We plan to have summative end of year tests. These give us the chance to design our own purposeful test that gives us useful information to feed forward into the school year ahead. We effectively create a new, more accurate baseline annually. The attainment for this assessment can be synchronised with the new 1-9 GCSE assessment model. It will prepare our students for a future that involves challenging external written assessments. It will also provide us with rigorous summative data to determine progress for any external accountability.
The end of year exam is based not solely upon pragmatism – but in good teaching and learning. They give students experience of a different type of learning challenge and will encourage teachers and students to revisit the knowledge and skills honed throughout the year. A judicious use of a test isn’t all bad. If they do know remember after six months and beyond then have they really learnt anything? We must make judgements, but they can be on our terms. We can define how we share the outcomes to ensure we stick to our principles of high quality feedback.
We are still in the making of course. Our model is a lot like learning: it will suffer failures and it will need refining and honing. We need to share it within school, with parents and with students, so that we can move beyond NC levels with confidence. I am confident and excited by the prospect and so are my colleauges.
The following blog posts helped shape my thinking:
Tom Sherrington’s post on setting standards of excellence: www.headguruteacher.com/2013/11/20/defining-the-butterfly-knowing-the-standards-to-set-the-standards
– A critique of NC levels by Daisy Christodoulou: www.thewingtoheaven.wordpress.com/2013/12/01/do-national-curriculum-levels-provide-us-with-a-shared-language
– A alternative model to levels, focusing on ‘mastery‘ by Joe Kirby: www.pragmaticreform.wordpress.com/2013/11/30/mastery
This following documents by Tim Oates is also very useful:
Using international models to refine the NC and assessment: http://www.cambridgeassessment.org.uk/Images/112281-could-do-better-using-international-comparisons-to-refine-the-national-curriculum-in-england.pdf