Consider how these grades are arrived at. Students start their A-level courses not as PhD students in embryo but as GCSE pupils who have had a summer holiday. GCSE and A-level are totally different examinations, rooted in different educational philosophies. Some students adapt quickly from the one to the other, but most take a bit of time.
By the end of the first year of A-level, many students who will eventually get good grades have still not made the breakthrough, and summer exam results at this stage are notoriously disappointing. Nevertheless, it is at this point that we must estimate the students' final grades.
Is this a student who will blossom later in the course (many do)? Or has this student come as far as he or she is likely to get (often the way)? This grade will depend on coursework to be completed next year, that grade will depend on the terminal exam, with all the elements of chance and circumstance for which A-level exams are only too well known. Yet estimate we must.
Does it matter? Some voices encourage us to relax. Colleagues report from liaison meetings with higher- education counterparts that estimated grades are taken as a guide only, to within a grade either way of what the candidate should get. It is an estimation, not a prediction, we are repeatedly told.
But the reality could hardly be more different. Students return from interviews or open days and report that their chances of a place depend entirely on the difference of an estimated grade. Any lingering doubts are soon dispelled when the offers start coming through. I have even known rejected candidates receive offers because a split estimated grade has been reversed, from CB to BC. Our students' futures depend on what can only ever be a provisional, frequently unsatisfactory judgment.
No wonder parents put the pressure on us. Few in higher education will know what it is like to face an angry parent, one moment accusing you of blighting their child's future, the next appealing to your sense of common humanity, because you have estimated a C rather than a B. Some teachers might be tempted to overpredict on principle, and who can blame them?
As long as we are stuck with exams in the summer and university years starting in the autumn, we are going to have to live with estimated grades. But why insist on such precision? Could we not at least work with split grades? At present they tend to be seen as rather a cop-out, putting a question mark both over the students who get them and the dithering teachers who give them, but in fact they are probably as precise as it is legitimate for us to be.
Sean Lang is head of history at Hills Road Sixth Form College, Cambridge If you have a strong opinion on a curriculum issue, write to Brendan O'Malley, secondary curriculum editor, TES, Admiral House, 66-68 East Smithfield, London E1 9XY