A level guide to ability

Two of the four letters responding to my article (Personal View, THES, November 17) on the national scandal of disparate A-level grades in different subjects, referred with great confidence to "convincing evidence" submitted by the Curriculum, Evaluation and Management Centre to the School Curriculum Authority that "demonstrates" that the hard sciences and languages are more difficult (implying that this justifies the higher percentages awarded grades A and B at A level).

But the statistical findings reported are open to alternative interpretations. CEMC's methodology is premised on the assumption that differences in exam performance can only reflect variations in candidates' "general ability". But the interpretative subject such as A-level English or history really is not directly comparable with - and certainly is not the same as - the kind of ability that solves maths and physics problems effectively. Some fortunate and well-taught children can do very well in both; but most really excel only in one or other.

What about systematic differences in teaching input? CEMC's report shows students with superior average grades at GCSE taking hard science A levels and then failing to reproduce the same degree of superiority at A level. The report claims this reflects greater "difficulty" of the subjects. But these are the students already selected for their abilities in these subjects, as the letter from sixth-former Gareth Mines pointed out. It more probably reflects lower value added at A level in these subjects in a systematically underfunded and undervalued state education system. They are the subjects whose quality of teaching input is more affected by shortages of equipment, and by good graduates taking their marketable skills of numeracy and modern languages elsewhere.

I therefore repeat my proposal for the principle of equal treatment of all subjects. School exam results at A level are not giving us an absolute index of "general ability". They are only showing us how well an individual performed in a particular kind of test of a certain kind of ability. Employers, university admissions offices, etc, remain free to have their own opinions about whether they are more impressed with high performance in one kind of test or another; but the comparability of the grades should not be compromised as they currently are through the practice of awarding different percentages of As and Bs.

Simon Szreter

Fellow of St Johns

University of Cambridge