Alan Smithers: Are we heading for another A-level fiasco?

A-levels, it seems, are close to collapse. With 24 million scripts batting about the country to be marked in just a few weeks, there are fears that another fiasco like last year's is on the cards. Even the person in charge, Ken Boston, the chief executive of the Qualifications and Curriculum Authority, sees the arrangements as "virtually unsustainable" and recently admitted that he hoped "the wheels don't come off" this summer.

Clearly something has to be done. Dr Boston, himself, favours "massive technological upgrading and computerisation". Kathleen Tattersall, the outgoing head of the largest board, AQA, wants to hand most of the marking over to teachers. She is supported by Judith Norrington, the director of curriculum at the Association of Colleges, who argues that exams should be replaced by teacher-assessed credits of "bite-sized chunks". Then, there is the British bac lobby which insists that everything would be all right if only we adopted the bac.

But all of these "cures" seem to me to be worse than the malaise. Dr Boston's computer assessment and marking implies switching wholesale to multiple-choice testing and giving up the attempt to assess coherent and sustained thought through essays. Kathleen Tattersall's and Judith Norrington's proposals keep, or even increase, the amount of assessment, but render it fundamentally unreliable. The bac solves nothing but simply adds the problem of agreeing what subjects we are going to force all students to take.

The way through the present difficulties is to understand how they started. Much of the present overload comes from the splitting of A-levels into AS and A2. The new AS was a well-meaning attempt to provide for sixth-formers who left after one year with nothing to show for it. But it has been badly mishandled. The modularisation of A-levels, with the expectation that all candidates would take AS, has more than doubled the assessment load.

The straightforward way out of the morass, therefore, is to revert to the notion of AS as a stopping-off point for those content to have completed one year in a subject. From this it would follow that most students could proceed on two-year courses unfettered by any national assessment for most of that time. But, additionally, there would be a separate and distinct one-year exam available. Now I can already hear the cries that this would also mean placing more reliance on external end-of-course examining. But this, in itself, would be an improvement. The second major contributory factor to the massive assessment overload has been the search for unrealizable perfection when it comes to judging pupils' performance. As the limitations of each form of assessment have come to be recognised, more and different forms have been added. To written and practical papers we have piled on continual assessment and course work portfolios. To summative assessment we have added formative assessment and diagnostic assessment.

There is an urgent need to simplify by going back to first principles and asking: why examine? A-levels derive from university entrance examinations and their original purpose, which still remains important, was to distinguish between people in terms of what they can do so that places could be allocated fairly. Given that perfect measurement is not possible the best approximation is for performance on specified tasks under standard conditions to be assessed independently.

To urge external marking is not to question the professionalism of teachers, only to point out the near impossibility of achieving fair and reliable grades when each teacher will only encounter a small sample of the possible performance range. This is without any of the accountability pressures arising from teachers being judged on their pupils' performance which has led to some cheating in the national curriculum tests. My recipe for getting A-levels, which have served us well for over 50 years, back on track is just plain common sense. First, bud-off AS as a one-year exam. Then, concentrate on getting right for A-levels the relatively small number of end-of-course examinations that we would probably find we actually needed. Kathleen Tattersall, in her recent speech, conceded that "the public saw external exams as the only reliable indicator of pupils' progress". Teachers themselves, she said, needed to help change these perceptions. Bully for the public.

The writer is the Sydney Jones Professor and Director of the Centre for Education and Employment Research at the University of Liverpool

education@independent.co.uk