Is it not a scandal that sixth-form students have such widely varying chances of achieving the top grades of A or B at A level, depending simply on their subjects of study, rather than on their capacity to achieve a comparable degree of exceptional performance in the exam? I have analysed the statistics of A-level results in different major subjects for the whole of the United Kingdom for 1995. The statistics show very clearly that there are two very different kinds of examination in terms of proportions of grade As and grade Bs awarded at A level.
First, in both the "hard sciences" of maths, physics and chemistry, and in all the languages, including classics and music (which some might prefer to classify with maths), at least 19 per cent of candidates achieve a grade A; and 17-24 per cent a grade B. Second, in the two "softer" sciences of biology and psychology as well as in the whole range of mainstream social sciences and humanities subjects (business studies, economics, English, general studies, geography, history, religious studies, sociology and political studies) only 10-15 per cent of candidates receive a grade A and only a further 15-20 per cent a grade B.
Thus it is the enormous discrepancies between subjects in the percentages awarded a grade A which primarily accounts for the differences; but in virtually every case this is exacerbated by a similar if smaller percentage awarded a grade B. It is therefore the case that while at least 37 per cent (and up to 45 per cent in maths and classics) obtain grade A or B in the first set of subjects, by contrast, no more than 33 per cent (and less than 30 per cent in many cases) obtain either of the top grades in the second set of subjects.
It really is absurd that while 26.4 per cent of the candidates sitting maths A level (52,188 candidates) received a grade A in 1995, in another large subject like sociology (30,371 candidates in 1955) only 24.8 per cent received either a grade A or B. That means that one in every 17 holders of a grade A in maths was actually ranked lower in their examination than some of their colleagues who may think of themselves as a "failure" by comparison because they "only" achieved a grade C in sociology.
It seems most plausible that these substantial differences between subjects principally reflect the varying nature of the examination and marking process. The subjects with high proportions of top grades have in common the fact that the examination process relates most directly to the examiner's assessment of whether answers contain material that is right or wrong.
Good candidates can therefore often score 80 per cent or more on different sections of the exam and performance close to 100 per cent is not in principle inadmissible nor that uncommon. The subjects with low proportions of high grades are those where the examiner is primarily asked to judge the quality of an answer which demonstrates the candidates' powers of judgement and interpretation in mobilising the knowledge they have learned: it is not a question of right or wrong answers. There is an unwritten rule in many of these subjects, for example in history, that essay answers scoring more than 80 per cent are freak occurrences; certainly a 100 per cent performance on any essay, let alone a whole "perfect" paper, is considered an impossibility. Candidates are in practice marked out of 80, not out of 100.
This situation is particularly scandalous inasmuch as there is no technical difficulty at all in achieving equity, at least across the larger subjects where there are many thousands of candidates. This could be done by observing the simple convention that the same agreed percentage of candidates should be given each of the grades in each subject by each examining board. The top grade, A, should be awarded in each subject to exactly the same percentage of candidates (which could be anything from 5 to 25 per cent of those taking the exam, as long as it is the same for all subjects). Similarly the same percentage in each subject should be awarded a B and so on.
If universities' admissions officers, employers or society more generally, suspect that certain subject attract a generally weaker pool of candidates or that it is more employable or more impressive to perform well in certain A level subjects than others, then there should be some rigorous research on the matter to test such views, rather than tolerance for the continuing rule of prejudice and hunch.
Does the person who is prejudiced to be more impressed by someone with a grade A in maths or physics realise that it was statistically easier for them to achieve their grade than someone with a similar grade in history or English? There is a kind of cultural double-counting here, as common prejudices reinforce each other.
Simon Szreter is lecturer in history and a fellow of St John's College, Cambridge.