A level guide to ability

Simon Szreter's view on A level grades is an astonishing misuse of statistics. He argues that, because a higher proportion of A-level candidates gain A and B grades in some subjects than in others, those subjects are easier. That would be true only if all subjects had the same pattern of ability among their candidates. That simply is not the case.

As Fitz-Gibbon and Vincent of the CEM Unit at Newcastle University have demonstrated, for a given level of prior academic attainment (as measured by GCSE results) the "hard" sciences of physics, chemistry and maths are indeed harder; a candidate of given prior attainment is likely to do worse (up to nearly a grade worse) than an "average" result. Dr Szreter implies that some As in maths may equate with Cs in sociology. The CEM evidence, however, suggests that a given candidate is likely to score a grade higher on sociology than on maths.

It is akin to arguing that, because 75 per cent of GCSE Greek candidates but only 10.8 per cent of English candidates gain A* or A grades, it is seven times easier for a candidate of given ability to gain an A in Greek than in English.

James Miller

Royal Grammar School

Newcastle upon Tyne