Education: Putting theory into practice: a head's story: Judith Judd visits a school in Wales where staff are trying to persuade more pupils to attend - and even to stay on after 16

HOW effective are school effectiveness theories? Glan Ely High School in Cardiff should know. Since John Bell took over as head four years ago, almost every school improvement idea in the book has been tried there.

Glan Ely, which is just half a mile from the scene of last year's Cardiff riots, may be on the edge of the city, close to green fields, but it has all the difficulties of an inner-city school. Around 10 per cent of its pupils rarely attend school. Attendance at parents' evenings is often below a quarter. Pupils' behaviour presents a constant challenge to staff.

As Tom Dyer, the school's curriculum manager, puts it: 'Some find it very difficult to get on with others in a controlled atmosphere so they blow up one of their peers.' A substantial proportion of pupils do not complete their GCSE courses, he says.

Mr Bell, an enthusiast for school effectiveness, is trying to improve the school with the help of researchers from the University of Cardiff school of education. He has organised the painting of walls and carpeting of corridors, set up committees to involve staff in planning, encouraged a friendlier atmosphere and higher expectations of pupils among staff. He has abandoned remedial streams and separate classes for children with severe learning difficulties. He has also reorganised the school into smaller classes - this year's entry will be in groups of 20 or 21 - and has led a drive to improve attendance.

The most remarkable change is in the numbers staying on after 16. Four years ago, between 15 and 20 students were staying on to do A-levels. This year 43 out of 147 have remained to take BTEC First Diploma courses, introduced last year, and 16 to do A-levels, a total increase of 40 per cent.

Barry Whittaker, who organises the BTEC courses, which offer pupils programmes that are related to work, says: 'Some pupils were doing A-level, which really wasn't appropriate for them, and dropping out. We are able to offer them another route into higher education.'

This year's returners include some pupils who dropped out of school before taking their GCSEs. All those in year 11 (formerly the fifth year) are sent a letter inviting them to return to school whatever their attendance record before the age of 16.

The school's older pupils believe the incentive to stay on has grown because the atmosphere is more relaxed. Ceri Hicks, a lower sixth-former and member of the recently formed student council, says: 'The school is much better than it was because it seems to be doing things for its pupils. You can go and talk to Mr Bell. We wouldn't have done that with the last head. People respect the building more now the school's been painted and some of the corridors carpeted.'

Improvements in attendance have been slower to materialise, despite a series of initiatives. Teachers use a 'form daily record', praised by John Patten, the Secretary of State for Education, after a visit. They also check that each pupil is present at the beginning of every lesson. During the day, the school employs a warden to stand outside the local shops, where many children go at lunchtime. She sends anyone trying to skip lessons back to school. There are rewards and cash prizes for attendance.

Absenteeism has edged down by a couple of percentage points, according to the Cardiff university researchers, but Mr Bell is not satisfied. During some weeks in this spring term only around half the pupils in their GCSE year attended school. Mr Bell says the form record has helped, but feels he has yet to make substantial inroads into non-attendance.

Exam results also remain much as they were, though both Mr Bell and Dave Reynolds, lecturer in education at the university, believe they will get better. It takes between four and seven years to turn a school round, Mr Reynolds says, and Glan Ely is already making strides.

The school's experience reveals the complexities of school improvement. Mr Reynolds suggests that ideas may have to be rethought in general. 'Improvement may depend more on the context a school finds itself in than we have realised: socio-economic circumstances, the head, the phase it is in. It will take longer in a school that is not used to debating educational issues.'

In some schools there is a rump of teachers that is unwilling to change and is suspicious of effectiveness theories. Mr Reynolds says: 'Perhaps if we talked about classroom improvement, about teaching and the curriculum, rather than school improvement, we would appeal more to this group.'

The most impressive school improvement scheme he has seen was in Israel. 'Super teachers' were brought into a school to sit in lessons and suggest new ways of teaching: the bottom-up approach. Would this work better than the Government's policy of putting in a small team of experts to sort out the school from the top? There are no easy answers.

(Photograph omitted)