One of the most useful things in the recent Assessment Commission report was the clarity with which it set out the three main purposes of assessment:
It’s quite clear that the purposes of day-to-day assessment (which I like to call ‘feedback’) and summative assessment are separate. The formative stuff is about what happens in the classroom: the information the teacher needs, the impact on the pupils. The summative stuff is simply that: a single data point that gives us an indication of outcomes.
When it comes to appraisal, many schools are struggling to work out how to replace the sorts of targets that used to be based on levels and sub-levels. Personally, given the choice, I’d scrap them altogether, but I realise that’s a long way off in most schools. So what in the interim?
Historically, teachers used their on-going assessments in the classroom to fill in things like the APP grids. Then the judgements they made fed into an overall level for each child, and summary data was thus produced, collated, and turned into an outcome on which teachers were judged. And in many cases, teachers would tick or untick as many boxes as it took to make the official judgement match the level they had already decided upon. I know it happened: I’ve seen it done, and I’ve done it plenty.
If we try to use our formative assessments to create a summative judgement against which we can judge teachers, then we have no hope of maintaining that separation of formative from summative. I don’t think that teachers are out to cheat the system in the slightest. Nor do I think that teachers will do anything for a pay rise. But I do think that the judgements made in appraisal meetings weigh heavy on teachers’ minds, and that consciously or otherwise these are likely to affect the accuracy and integrity of teacher assessment data. Not because they want the pay rise, but because they want to be thought of as good teachers.
If we insist on setting data-based targets, then surely it is much clearer and fairer to separate that judgement of the teacher from the teachers’ judgements of pupils. All the time we use teachers’ own assessments as a tool to judge them, we can only expect one to influence the other. If we just have data targets, then lets use a simple baseline and end-of-year test, where teachers know that the expectation is to improve attainment in a key area, but aren’t then expected also to produce the data that proves the impact they’ve had.
Otherwise, no matter how good your assessment system’s design, it’s destined to fail.