John Bald says that despite short term rough justice the goal of a more rigorous exam system must be pursued
The current spat over GCSE English – the error lies in the grading of the January papers, and so cannot now be corrected – has reignited conflict dating from the early days of comprehensive schools. The Left's immediate target, grammar schools, was just a staging post in the long struggle against the idea that the purpose of education was to develop each person's potential as far as could be.
This principle led, in the words of Michael Young, to "The Rise of the Meritocracy", in which the intellectually successful took the best of everything, with crumbs for the rest. The Left hate non-selective academies just as much as grammar schools, because they are based on individual achievement and responsibility rather than collectivism.
The key question is whether tests and examinations are a touchstone for the knowledge, skills and understanding that are needed for success in life and intellectual progress, or are a rationing system, a "narrow gate" (Shirley Williams) a "racket" (Michael Rosen), designed to regulate, control and suppress?
Most Conservatives would, I think, take the former view. The first competency examination, for naval lieutenants, was a major factor in the success of the Royal Navy as it ensured that a person in charge of a ship had to know how to sail it. Whittle, inventor of the jet engine, and Turing were products of an elite examination system – Whittle in particular attached strong importance to his First – while R J Mitchell, designer of the Spitfire, began as an apprentice. One way or another, their achievements were individual, though the application of their discoveries depended on teamwork.
To command respect, an examination system must be fair and accurate. None of our current tests and examinations achieves this, and I wonder how far it has ever been the case. Nelson's lieutenant's examination was presided over by his uncle, who later felt obliged to state that he had not asked any questions. The 11+, which set out to be fair by including an IQ test that was thought to test ability rather than background, turned out to be unfair because of factors that were not understood when it was developed.
For example, learned vocabulary is a huge advantage in verbal tests, and you can pump up your (or your child's) score in other areas by learning simple numerical progressions and square numbers, and by studying the ways in which two-dimensional shapes can be manipulated. These have led to a heavily-coached, private 11+ curriculum, which can put 30 or more points on an individual score – I know this from coaching a neighbour some years ago. And, of course, in the heyday of the 11+, girls had to achieve higher scores than boys in order to pass.
O and A level grades were allocated on a quota system, making it a competitive examination, with around 5% of the entry awarded an A, with a backup of an archive of graded scripts to give an idea of comparability between successive years. If everyone knows where they stand, a competitive system can be fair, if at times hard. For example, university degrees are now of very little value in France, while the higher educational qualifications and Napoleonic "grandes écoles" are competitive. The French moan, but most accept the system.
However, awarding grades on the basis of a normal distribution curve, aka Bell Curve, does not reflect the qualities of the candidates' work. The Bell Curve has never been validated in an educational context, and I'm sure I'm not the only one to have thought back over their exam results and found one or two of them inexplicable, either positively or negatively.
Criterion referencing, in which work had to meet specified qualities, sounded better when it was introduced, but it was soon subverted. At A level, the criteria were set out in so much detail that they became a crib, and replaced genuine thinking with mechanical conformity. The grading of the highest-attaining pupils was often grotesquely unfair, and boards found themselves giving out so many As that they tried to keep the situation in check by giving dishonestly low grades to candidates' final pieces of work.
At GCSE and in national tests, they were turned into a mark scheme whose boundaries could be moved up and down like a trombone slide to produce the result examiners wanted. National Tests were wrecked by this, as the goal of Labour's regulators was to keep the pass rates up, and so disguise the real lack of achievement in many schools, particularly at L4 in English.
When an external examiner was brought in to take over the tests for 11 year olds, around 2005, it introduced accurate grading and the pass rate fell so much that messenger was duly sacked. The full extent of this fiasco has yet to be revealed.
So, will the reported plan to have one board setting GCSE in core subjects result in a fair and accurate system, by which I mean one that gives to each candidate the grade he or she has actually earned?
It will reduce what we now know to be the malign effects of competition between boards, and the move to a final examination will eliminate the vices of coursework, which also seemed a good idea when it was introduced, but has been destroyed by inconsistent practice between schools and, at worst, by downright fraud.
Without further safeguards, such as an archive of graded scripts to ensure
comparability from year to year, and a body of competent examiners, we could just as easily slip back to the days of the QCA, when the main factor determining a grade boundary was its likely reception in the popular press. In the end, the success of the move will depend on the ability of the regulator to deliver something more than rough justice.