What do exam results mean?

The argument over academic standards ignores the issue of fair assessment, says Alan Smithers

For young people of a certain age, August is not silly: it is the season of great joy or huge disappointment, depending on how they do in their exams. This year's batch of results has been confusing, to say the least. A-level pass rates appear to be going up and that is bad because it suggests the exams are getting easier. GCSE pass rates appear not to be going up and that is bad because it means we will not reach national targets. Three-fifths of vocational students seem to have gone missing - but that could be good because it may mean the qualifications are more rigorous than we thought. What is going on?

Education in England and Wales (Scotland does it differently) used to be about producing an elite. Tests at age 11 and tough examinations at ages 16 and 18 picked out a select few for higher education and the top jobs.

Even in 1980 only a quarter of the age group were successfully negotiating the step at 16, and 14 per cent at 18, with 11 per cent obtaining degrees. Since then successive Conservative administrations have sought to move towards developing the talents of all. They have introduced a national curriculum. But post-16, instead of designing afresh, they have adapted what was already there and bolted on some new vocational qualifications.

To underline the Government's seriousness, however, it has adopted from the Confederation of British Industry ambitious targets for 2000: that 85 per cent should mount the step at 16-plus, and 60 per cent at 18-plus.

It is against this background that the seemingly contradictory reports of the past week begin to fall into shape. A-levels are the descendants of university entrance exams designed to ration out scarce and valuable places. As we have moved from an elite towards a mass system, so A-levels have changed, with new, and some say softer subjects introduced, and syllabuses and examinations adapting to the changed clientele.

Those who claim that there has been grade inflation are probably right, but if the examinations are sound and if they enable fair transactions to take place between school and university, it matters not that they are different from yesteryear. What is important is that the examinations should be fair, consistent and valid in today's conditions. We must be confident that the grades awarded accurately reflect a person's achievements.

At GCSE we are beginning to see some of the consequences of stronger regulation. Limits have now been placed on how much of the examination can be by coursework, and last year some inconsistencies were noted in the grades awarded by different boards. This led to tightening of the marking schemes.

This year's GCSE results have a more authentic feel. As regulation has begun to bite so passes at English and maths have fallen back a bit. The smaller and more highly selected group taking physics, chemistry and biology separately rather than as combined science have got higher grades, as might have been expected. The overall pass rate is less than last year.

This has led to fears that the Government's national training and education targets will not be met. These targets represent aspirations, however, rather than plans, and the hopes of young people will be better served through qualifications that record real gains than through cavalier credentialism.

In any case, the national targets are expected to be met not only by GCSE but also through the new vocational qualifications. The extraordinary thing about the results of the General National Vocational Qualifications announced this week is that while 262,000 students are shown as having been registered only 62,000 have qualified, with a further 41,000 getting some way towards qualifying. What has happened to the rest is a mystery. Has there been some statistical mix-up, are the registrations not what they seem, or have a lot of students tried out the courses and not liked what they found?

It would not be surprising if they were dropping out in huge numbers. GNVQs, introduced at three levels ranging from sub-GCSE to A-level in 1992, are a bizarre concoction. The course content is not specified, neither does it distinguish the essential from the desirable; and the assessment process is convoluted. GNVQs also have external tests which, while they do not contribute to the grade achieved, are essential to gaining a pass. Yet some of these tests are so inappropriate that failure rates of more than 90 per cent and in one case - a unit in manufacturing - 100 per cent have been recorded. In three years, there have been only 47 awards in the A-level-rated GNVQ in manufacturing.

All this must bring a wry smile to the faces of our Continental neighbours, who have no difficulty in providing good general and vocational education. In Germany and France about twice as many young people reach the equivalent of GCSE grades A-C in maths, the national language and science. The secret seems to be that many who are not turned on by the subjects themselves are able to master them through good applied education, which demonstrates the relevance of learning to work and life.

We should learn from this and get the new vocational qualifications right. At present they are ill-thought out. Some are in the same subjects as A-levels. Most 16 year-olds who do well at GCSE go on to A-levels; those who do less well go on to GNVQs. Yet a GNVQ has been declared equivalent to two A-levels. From different starting points the students are expected to reach the same standard.

GNVQs have developed from the old technician training route. It is at this level that international comparisons show us to be particularly under- qualified. Yet GNVQs are marketed as opening up higher education. They should be developed as worthwhile qualifications in their own right with clear destinations in mind and clearly differentiated from A-levels.

We need, therefore, three types of education to follow on from the national curriculum: academic, applied and occupational. These could be met through the present qualifications of GCSE/A-level, GNVQ and their equivalents for training in the workplace, NVQs. The GCSE/A-level route is the best developed, though this year's results suggest they may still need some fine tuning. The idiosyncratic choices of some students across just three A-level subjects suggest there is a case for expanding this to five to achieve genuine breadth. But the GNVQ results merit a major investigation and there are also doubts about the fairness and consistency of NVQ assessment.

Effective assessment is the key to providing opportunities, developing talents and having a sound economy. The hand-wringing each August will only end when we move on from sterile arguments about academic standards to securing genuine improvements in vocational education.

The writer is director of the centre for education and employment research at Manchester University.