When will they ever learn?

Making children work harder can often lead to worse results, reports Carol Taylor Fitz-Gibbon

Years ago, a Conservative government launched a scheme to improve what was claimed to be the abysmal level of technical education in English schools. It was called, not very poetically, the Technical and Vocational Education Initiative. The aim was to "drive up standards"; it was hoped that, motivated by this scheme, children would leave school with better qualifications. Hundreds of thousands of pounds were invested.

The result? Those who were involved in the scheme got worse qualifications than those who were not. I was one of the evaluators and, as the results came in from one area after another, my dismay grew. We searched the data carefully, but there was no getting away from it. (We used a technique called "value-added", so that it was possible to tell how two pupils of equal ability and equal previous attainment, one in the scheme and one not, had fared.) A civil servant dreamt up an explanation. The scheme involved pupils doing lots of projects outside school; the weaker pupils, she said, would have fallen behind on basic work in the classroom. Her guess was the opposite of the truth. The weaker pupils showed no detriment from taking part in the scheme; it was the more able ones who got fewer and poorer qualifications.

Undaunted, Tory ministers quickly extended the scheme, which at that stage had been confined to a few selected areas, to all schools. The only change was a reduced level of funding.

This episode ought to have been instructive to civil servants and politicians, but it was not. They fail to see that schemes dreamt up in Whitehall will often founder on classroom realities. They persist in their claim, for example, that the past decade's barrage of mentoring, harassment and paperwork has driven up school standards. It is true that pupils of modest abilities now get quite outstanding exam results. Since 1988, my colleagues and I have given a general aptitude test yearly to thousands of A-level pupils. We have found that, for the same aptitudes, higher and higher grades are awarded in almost all subjects.

Has pressure worked, and forced everybody, teachers and pupils, to raise their game? Or is it just easier marking? I fear the latter. In summer 2000, the Engineering Council asked 60 universities about the level of competence among students entering degrees that required knowledge of maths. Many universities had given the same test over many years to their entrants. Even among those who had got the top A-level maths grade, competence had declined year on year in the 1990s.

I am not blaming the teachers and pupils. They say they are working harder, and they probably are. But the sobering truth is that trying hard to get students working may actually make matters worse.

You don't believe me? Here is some evidence that might convince you. My colleagues and I test the aptitudes of thousands of pupils in their final GCSE year and also ask about their intentions regarding A-level and university. We put those with high aptitudes and low aspirations on a list of "under-aspiring" students, and send the names to their schools. The schools go to great lengths: they send letters to the pupils' parents, check their homework, mentor them in the lunch hours. But I always doubted that, if I were an under-aspiring pupil, I would have welcomed being mentored in my lunch hour.

So (with the teachers' permission) we tried what is called a randomised controlled trial, of the sort that provides the gold standard of evidence in medicine. In 15 schools, we withheld the names of half the under-aspirers, giving the schools only a random half of the list. We analysed the GCSE results. We found that those who had been named as under-aspirers, and blessed with lots of counselling, made less progress than those who had not been named. For mentoring, read harassment.

This is an example of the kind of research that education needs. Big money is thrown around on bright Whitehall ideas without an experimental design to find out if they work. I used to believe that, as civil servants were often Oxbridge graduates, they would be clever. But clever at what? They don't listen and, when they do, they perhaps don't understand.

Teachers understand the need for trials. They don't want to waste their lunchtimes doing things that make pupils worse. Politicians and civil servants couldn't be bothered to master the kind of detail used in our research; teachers took it all on board, outshining the mandarins when it came to the intelligent use of data.

Many pupils are bored, neglected and resentful in classes that concentrate on drilling them for the sake of league tables of test scores and exam results. England does, in fact, head one of the European league tables: the one for teenage drunkenness and violence. Is that unrelated to the miserable Gradgrind philosophy, the managerialism, the naming, shaming and denigrating of teachers? I don't think so. There's no ill-will among the politicians; it's just simple incompetence and lack of experience. Most of the people making policy have never run a classroom, never mind a whole school. They are amateurs in a highly complex world, and they do not have the humility of good scientists.

Given incompetence at the highest level, we need teachers, with the help of parents and concerned citizens, to take on the task of regulating their supposed masters. The Office for Standards in Inspection (Ofstin), set up as a professional forum in response to Ofsted, aims to do just that. It meets on Tuesday 6 November, 7 pm, at Friends House, opposite Euston Station, London. Do join us.

Carol Taylor Fitz-Gibbon is professor of education at Durham University. Ofstin is at www.dur.ac.uk/ofstin.forum