Assessment 'sampling' will damage learners' health
These single-level progress tests would bring even narrower "teaching to the test" than we already have in Years 6 and 9. Worse, this will happen in every year group in key stages 2 and 3, and twice in each school year.
These are essentially high-stakes accountability tests in English and maths, with progress targets and financial rewards for meeting them. Such tests invariably distort teaching and learning. As the same test will be taken by both KS2 and KS3 pupils, they will also distort the curriculum.
Progress measures will be very unreliable as well.
So what is the Department for Education and Skills proposing? Single-level progress tests will be offered twice a year (December and MayJune). They will be "shorter, better focused", "when ready" and "lighter touch" and be accompanied by progression targets based on schools increasing the proportion of pupils who progress by two levels during a key stage. Schools will be rewarded through progression premiums for boosting the number of pupils gaining two levels (5 per cent of per-pupil funding for each of maths and English). Money will be made available during the pilot, starting in September, to fund 10 one-hour individual tuition sessions for those below their expected level. The teacher's role will be restricted to selecting the pupils who are ready. The two-year pilot will involve about 500 "keen" schools, but there is no information about how it will be evaluated.
The levels gained on these tests will count towards national targets. "The system would be a one-way ratchet: once a pupil has passed a level, they will never go back, only forward", so getting a lower level on the end-of-key-stage test, which will be retained, will not affect the level they have reached.
But this system should carry a health warning. Despite the sugar-coated "personalised learning" rhetoric, the proposed tests are toxic. If, as in Scotland, these were to be "when ready" tests to validate a teacher's judgment about a pupil's progress, I would be in favour. But they are not; they are about school accountability with a panoply of targets and payment by results.
Imagine you are the head of a school that has been given its progression targets. You need to increase the number of pupils progressing through two levels during a key stage. Is not the pressure on you to ask teachers in each year group to select as many pupils as possible to try the tests? Because the tests are shorter and "more focused", teachers will soon hone their teach-to-the-test skills, especially as the curriculum will be restricted. (If they are for both key stages, some of the KS3curriculum will have to go. Goodbye Shakespeare and algebra?) So you will designate May and December to preparing this group. Because the tests are "short and light touch" they will also be unreliable, so worth repeatedly entering marginal pupils as they may strike lucky. As a head, how would you use the 10 hours of funded individual tuition? At regular intervals during the year or for intensive test coaching (the antithesis of formative assessment) in the test weeks?
These are just the school-level concerns. There are also serious technical ones. How do you set a fair test for both nine and 13-year-olds on different programmes of study? How is such a level comparable with the end-of-key stage test, which is based on the full curriculum? Is the "ratchet effect" a recipe for level inflation? How dependable are progression targets based on low reliability tests? These progress tests will damage schools' and learners' health.
Dr Gordon Stobart is Reader in Education at London university's Institute of Education and a member of the Assessment Reform Group