I’m a bit slow on the uptake. I’ve now moved on from the EARLI/Northumbria Assessment Conference in Northumberland to ALT-C 2010 in Nottingham, with a day of walking, a day of writing and a day of interviewing in between. Before my thinking gets clouded with lots of things from ALT-C, I want to return to some of the issues raised at the EARLI/Northumbria Assessment Conference. In fact I want to return right to the beginning, to Royce Sadler’s keynote in which he suggested that all summative assessment should be at the end of a period of teaching, not during it i.e. that it should be terminal not continuous.
I’d read this suggestion before the Conference and had put it to one side as being too troublesome. The logic is impeccable – students are demonstrating what they have achieved by the end of the unit of study. You’re assessing the final outcome, not the path used to get there. You’re not penalising the late developers. Summative and formative assessment become completely separate things.
But if formative assessment is completely divorced from summative assessment, is it assessment at all?
More seriously, especially in a distance-learning context, how do you get students to engage in formative assessment? Professor Sadler suggested giving them an incentive other than marks. I think he talked about slices of pizza. Slices of virtual pizza perhaps. I’m not sure. Or we could tell students that people who do the formative assessment do better in the summative assessment. But is that honest? OK, students who do better in the formative assessment dodo better in the summative, but is the link causal? (or just that keen students do everything and do well).
We’re getting into dangerous territory here; I can see the conclusion ‘why bother with formative assessment at all’ looming in front of me. However if we assume that formative assessment is a ‘good thing’ (and a lot of what I do depends on that premise) then Royce Sadler’s argument points towards just letting students take it or leave it – if they use formative assessment in preparing for the summative assessment then fine and good; if they don’t – well it’s their loss. I can see logic in that, but at some point in the journey, many students are going to get a dreadful shock – will they recover from it?
Royce Sadler emphasised the point that if we assess during the unit of study then we are disadvantaging those who develop late. But what of the student who is ill at the end of the course, or who panics because ‘all their eggs are in one basket’? Also (although Royce emphasised that in talking about terminal assessment he was not advocating to a return to exams as the only means of assessment) – is terminal assessment authentic?
So there are issues and huge practical concerns, but I have to say that I agree with Royce’s basic thesis. Perhaps the question should be, ‘the end of what?’ Perhaps we should just assess at the end of a unit of study, but make the unit of study short.
Another thought is that perhaps we should get students to choose when they are ready to be assessed, as with a driving test. A driving test is an ‘end of unit of study’ assessment and we never query that – it would not be reasonable to take a test to show that you know how to start a car engine; you need to demonstrate a range of related skills in order to pass. But the learner is in control of the decision of when to take the test.
Continuing with the discussion of a driving test, it is also the case that in the UK you need to have passed a theory test before taking the practical test. Now I don’t like the theory test (in my book it’s the worst sort of multiple-choice text), but it does act as a sort of ‘gate’ that has to be negotiated before moving on. I think I’m working towards the idea of thresholded assessment (where you have to negotiate a ‘gate’ i.e. to demonstrate a certain level of understanding, but your mark doesn’t count towards your overall score). This, for somewhat different reasons, is the approach now being taken on level 3 Physics and Astronomy modules at the Open University. Interesting.