GNVQs fail to make the grade on three counts, Alan Smithers argues. England and Wales, indeed the whole of the United Kingdom, vitally need a good applied education route for educational reasons. We know from the high failure rates and worrying truancy figures that a lot of young people do not see much point in education for its own sake.

We also know that countries such as Germany, Switzerland and France are able to reach much higher levels of achievement in mathematics and languages through applied education than we are by concentrating on them as subjects. General National Vocational Qualifications could be that applied education route, but not without some radical changes.

What are the main problems? I would identify three: confusion over purposes, inappropriate assessment and failure to specify essential content.

The main weakness of GNVQs is that it is not clear what they are for. They derive from the old technician training route which was good as far as it went but not large enough to provide for all our needs. Compared with other countries, while we were producing about as many graduates as they were, we were grossly under-educated and under-qualified at the technician level.

É«ºÐÖ±²¥

GNVQs could have enlarged and strengthened career development paths, aiming particularly at technician and supervisory posts, but instead, at the advanced level, the award has been declared to be equivalent to two A levels, with the implication that it opens up the whole of higher education and employment.

We therefore have the odd situation whereby people who do well at GCSE go on to do GCE A levels and those who do less well go on to do GNVQs, yet by 18 they are supposed to reach the same level. Sometimes there are A levels and GNVQs in the same subjects, as with art and design, business, technology and science.

É«ºÐÖ±²¥

To make GNVQs work, their role must be clarified. It must be decided whether GNVQs are for people of lower ability in the same subjects as A levels, or whether they are to be distinctively about practical and applied learning.

If it is to be the former we must be clear that the GNVQs in for example technology and science are aimed primarily at technician posts. There could be degrees for master technicians. If GNVQs are to be the applied education route then the ambiguities between A level and GNVQ must be resolved. It must be understood and accepted that A levels are to be about disciplines and subjects, the fundamental ways of making sense of the world, and GNVQs are to be practical organisation of skills, knowledge and understanding.

On this basis, business studies, media studies and sports studies would belong with the GNVQs, science with A levels. There is no reason why the current examination boards and awarding bodies should not be involved in both.

This leads us on to assessment in GNVQs, which really is a joke. There are some short external tests but they do not contribute to the grading into distinction and merit. Nevertheless they are necessary to pass and on occasions have been so inappropriate that more than 90 per cent of the students have failed.

They attempt to encapsulate wide areas such as the whole of European art history in single multiple-choice questions and, since there is no syllabus, the students have no foreknowledge of what they will be asked. In any case, short multiple-choice tests would seem a highly questionable way of testing the practical application of knowledge and understanding, particularly in areas such as art and design, manufacturing and engineering.

The grading of performance into distinction and merit depends on the inspection of a portfolio of course work. This is assessed against broad criteria such as information gathering and evaluation. Quality was only introduced as a criterion as an afterthought. Assessors in different parts of the country and different institutions make very different judgements, so the grading is hopelessly unreliable.

É«ºÐÖ±²¥

The obvious way forward is to have externally set and marked practical assignments. With the introduction of GNVQ Part 1s into schools, the School Curriculum and Assessment Authority has become involved and I believe would like as much. However, even Ron Dearing's supreme simplifying skills have so far not worked their magic on GNVQ assessment.

With the Part 1s, though not necessarily the whole award, the external tests are to play a part in the grading, but there will be a pass-fail test to decide who passes and a further multiple-choice test to decide who passes at merit or distinction level. The results of the tests are to be brought together with course work according to some complicated combination rules. If GNVQs are about people being able to do things, let us see what they can do in as fair, authentic and practicable a way as possible. I repeat this should involve externally set and marked practical assignments.

É«ºÐÖ±²¥

But there is also the problem of content. GNVQs have been developed along the same lines as National Vocational Qualifications. In reaction against the time-serving of the old apprenticeship system the NVQs deliberately eschew all reference to time. This is considered liberating because students and trainees can work at their own pace.

It also avoids, however, having to prioritise and distinguish the essential from the merely desirable. It was failure to face up to taking hard decisions of this kind that almost sank the 1988 National Curriculum.

With several hundred "performance criteria" GNVQs have the same problem: too much to cover with no clear indication of essentials. The solution for the National Curriculum was to identify core programmes of study. This would also improve GNVQs. The core programmes should include general as well as vocational education as a basis for progression and flexibility.

Introducing GNVQs was a bold step. It opens up an avenue of possibility. To capitalise on that we must clarify their role, introduce appropriate assessment and specify their core content.

With these reforms we could hope to sustain the interest of the large numbers of students who are attracted to these awards but drop out. Getting the GNVQ right is vital to increasing educational opportunity and enhancing the economy.

Alan Smithers is director for the centre of education and employment research at Manchester University.

É«ºÐÖ±²¥