There are key questions that every school can ask when designing an assessment system for its curriculum and pupils – but there is no one-size-fits-all approach, says Dylan Wiliam
Let me be clear. I was a huge fan of the system of 10 (later eight) levels that Paul Black’s Task Group on Assessment and Testing recommended to Kenneth Baker (then Secretary of State for Education) in December 1987, not least because it was based on the work that Margaret Brown and I had done on levels of achievement in graded assessment schemes, and Carol Dweck’s early work on mindset. I actually did my PhD on national curriculum levels. And had we stuck to reporting student achievement at the end of each key stage—which is still the only legal requirement—everything would have been fine.
But then schools started reporting levels every year, and then every term, and then on individual pieces of work, which makes no sense at all since the levels had been designed to be a summary of the totality of achievement across a key stage. And then Ofsted inspectors insisted students should make a certain number of levels of progress each year and started asking students what level they were working at, in response to which schools started training students to answer appropriately. And don’t get me started on sub-levels…
So that is why, when I was appointed as a member of the Expert Panel to advise the Secretary of State on revisions to the national curriculum, I recommended that national curriculum levels should be abolished. Not because the levels were a bad idea, but the way they were being used was getting in the way of children’s learning.
Some people are happy about this. Others are not. But levels have been abolished, and the Department for Education is not going to recommend anything to replace them. It will be up to each school to decide how to determine whether children are learning what they need to be learning. Some schools are planning to continue with national curriculum levels for the time being. That’s fine. But it is important to realise there will be no straightforward way to carry levels forward from the current national curriculum to the new curriculum.
Most importantly, Ofsted inspectors will no longer be able to walk into a school and assume they know how a school is monitoring student progress. They will have to ask. And as long as the school has a good answer, they will be OK.
Developing an assessment system will be challenging, to be sure, but primary schools now have an opportunity to develop assessment systems that fit their curriculums, rather than trying to shoehorn their curriculum to fit a predetermined assessment system. And because every school’s curriculum is different, the best assessment system for one school may be useless for another. Ultimately, each school will need to find an assessment system that meets its needs.
Companies are already falling over themselves to offer ‘assessment systems’ to primary schools, but it is important to realise that most of the solutions on offer are really recording systems, not assessment systems. They allow teachers to specify what they are recording, and then, once information has been entered into the system, nice reports will be generated at the touch of a button. However, before a school decides how it wants to keep track of student progress, it needs to decide what it’s going to keep track of. Moreover, there can be no off-the-peg solutions, because the assessment needs to match the curriculum in place in the school.
What follows, therefore, is not a blueprint for an assessment system, but rather a set of principles that schools should consider as they design, and over time, refine, their assessment system. These principles will often be in tension, so there can never be an assessment system that satisfies all the principles, but by thinking about these principles, schools can ensure that the compromises and trade-offs they are making are ones with which they feel comfortable.
A school’s assessment system could assess everything students are learning, but then teachers would spend more time assessing than teaching. The important point here is that any assessment system needs to be selective about what gets assessed and what does not, and so the assessment system needs to focus on the ‘big ideas’ in each curriculum area. For example, place value is a central concept in the understanding of our number system. Without a profound understanding of place value, most of mathematics makes little sense. Roman numerals, on the other hand, is not quite so important. As the headteacher or a parent, I would far rather know how a child is doing in terms of their understanding of place value than their knowledge of Roman numerals. You can’t assess everything – be selective.
Once a school is clear about the ‘big ideas’ on which the formal assessment system will focus, it makes sense to think about the ways in which students will get there. Not all students will follow the same routes in their learning, but the assessment system will need to collect evidence of how students are progressing towards the goals.
Once the learning progressions have been identified, it is useful to establish ‘checkpoints’ along the way, which can either be intrinsic to the subject, or driven by extrinsic demands, such as the need to report to parents. Intrinsic checkpoints might include particular issues that are known to cause some students difficulty, or significant stages in development. Extrinsic checkpoints would include end of years and key stages.
It is one thing to say that we want to know whether students can “use inference and deduction” in their reading. It is quite another to decide what evidence we need to enable us to conclude they can, or cannot, do this. Some of the evidence may come from formal tests or set-piece situations such as interviews or discussions with children. Some evidence will come from marking and some will come from just observing children. Obviously, the more formal the assessment procedure is, the easier it is to record the evidence, but schools also need to make sure that the desire to record evidence for purposes of external accountability does not result in high-quality ephemeral evidence being ignored.
In particular, schools will need to decide how much evidence is needed before a child is regarded as being able to do this. This is tricky because almost all students may be able to demonstrate a particular competence in one context, and no-one will be able to do so in all contexts. Simplistic rules of thumb like requiring a child to demonstrate something three times to prove they have ‘got it’ are unlikely to be helpful. Sometimes a single example of a child using something learned in one context in a very different context will be convincing evidence of mastery. At the other extreme, a dozen repetitions of a particular skill in similar contexts may mean very little. Here, there is no substitute for professional judgement – provided, of course, ‘professional’ means not just exercising one’s judgement, but also discussing one’s decisions with others, to establish that they too, with the same evidence, would have drawn the same conclusion.
Too many schools set ‘minimum target levels’. The problem with such targets is that given our ability to predict future performance is so poor, the targets are hopelessly undemanding for some students and really challenging for others. A related problem is that due to the use of monitoring and tracking systems that have become popular in England over the last decade or so, we tend to use current performance as a guide to future performance. This may sound innocuous enough, but it is a recipe for reproducing the status quo. For a child in Year 3, the minimum target achievement for the end of the year should be the level of achievement they need to thrive in Year 4. For many students this will be a relatively undemanding target, and higher targets should be set. But for lower achievers, the aim should always be to break the cycle of failure and do everything we can to get that student to where he or she needs to be. In response, some people might say this target may be too ambitious, but my response would be to change the question. Instead of asking “What level of achievement should we have as a target?” we should ask, “What do we need to do to make sure that this child is ready for Year 4?” As Rick DuFour says, “Don’t tell me you believe that all students can succeed. Tell me what you do when they don’t”.
Assessment is a good servant, but a terrible master. Too often, we start out with the idea of making the important measurable, and end up making the measurable important. By sticking clearly to a set of principles for the design of an assessment system, schools can ensure that the assessment system supports learning, rather than gets in its way.
For those interested in learning more about the approach outlined in this article, see Principled assessment design by Dylan Wiliam, published by the Specialist Schools and Academies Trust (ssatuk.co.uk).