It is teaching which raises standards and assessment is a tool to drive better teaching
The science of assessment is quite sophisticated. It is not perfect – how could it be? – and not without controversy but there are many clever things we can do to find out what people know, what they can do and what they are good at. Yet we struggle with assessment all the time in schools, constantly doubting the results, bemoaning the side effects and changing the approach. The cause of this contrast is in the uses to which we put assessment. In particular, high stakes accountability puts a stress on assessment that even the most sophisticated methods struggle to bear.
It is not so much that people cheat – the integrity of the education profession is high – but that they adapt their practices inevitably and imperceptibly to performance on national tests, rather than the domain that the test is a sample of. However sophisticated, no feasible assessment captures the whole of the topic it is acting as a proxy for. This applies to tests as much, if not more so, than teacher assessment. This has two pernicious side effects on the education system.
Firstly, it narrows down the range of national assessments we can use reliably: for only certain types of assessment can bear the weight. The danger is that these forms of assessment may not be the best way to capture the skills we value. We choose them because they are robust, not because they are insightful. Secondly, and because of this, it also narrows down the curriculum, because we end up teaching only that which can assessed through the specific methods that we are restricted to. We have the whole chain of logic the wrong way round. We teach what can be assessed and we assess what can be verified; pursuing reliability at the expense of validity.
This week’s announcements of a potential return to testing at the end of KS1 are a case in point. They are driven explicitly by concern that that national tests need to be robust enough to support a measure of progress across primary school (that is, they are a baseline rather than an outcome). The conversation does not begin with a judgement about what should be taught by the age of seven or what might be the most appropriate means of assessing that in seven year olds, but by what assessment will support the weight of accountability. This is counterproductive.
It is teaching which raises standards and assessment is a tool to drive better teaching. Every time assessment is lifted out of that context and used punitively, it devalues its use for the one thing that can actually make a difference.
Russell Hobby is the General Secretary of the National Association for Head Teachers.
Follow Russell on Twitter @RussellHobby