While everyone expects a clear answer to this question, the truth is that it is impossible to give one. First, comparing results from year to year is surprisingly complex. Even with identical exam formats it is impossible to guarantee that questions are of exactly the same difficulty from one year to another. Second, there probably is no single trend - up or down.

Some pupils, in some schools, are performing considerably better that their earlier counterparts, other pupils, in less effective schools, are doing worse. The variation in reading scores identified in the NFER study Reading in recession supports such a view, as does close analysis of last year’s school-by-school GCSE results. The existence of contradictory trends should not surprise us. Both crime statistics and economic indicators frequently puzzle experts and make valid overall conclusions impossible, why should educational trends be easier to interpret?

Whether standards are slipping, remaining the same or improving is not the point. What matters - according to educationists as well as their critics - is that pupils in English schools start to reach considerably higher levels so that they can compete successfully with young people educated elsewhere. I also believe it is imperative that the long tail of underachievers, currently failing in or dropping out of, school be reduced if we are to have any hope of avoiding a disaffected underclass.

The recent changes can be seen as part of a world-wide pattern of school improvement initiated by governments of differing political persuasions. School improvement is a generic term for a broad range of approaches which seek to improve the learning opportunities of pupils and to raise the standards of their achievement. It is based on the application of knowledge gleaned, over the past 20 years, from the research into school effectiveness.

School improvement combines concern about the achievement of individual pupils with a recognition of the need for school change. It does not favour the creation of a recipe book because it recognises that most schools are too complex for such simplistic approaches, but school improvement programmes commonly include: * charting the school’s strengths and weaknesses; * listening to the views of pupils, parents and employers; * agreeing ambitious, yet realistic, targets; * planning and implementing changes in school management; * experimenting with changes in classroom teaching; * systematically checking on how well the changes have worked; For an individual school, such initiatives bring staff and governors together in a task of continuous improvement ideally long before - but sometimes after - an inspection by the Office for Standards in Education. In such cases schools can now receive money from the Grants for Education Support and Training fund to implement the inspectors’ main recommendations.

Increasingly help is sought from outside. Some schools look to their local education authorities to provide it, others turn to university or independent consultants. Many schools have benefited from the data analyses provided by researchers at various British universities (John Gray, David Jesson and Carol Fitzgibbon) and many of my colleagues at London University’s Institute of Education. The school improvement centre at Bath University, the new Sheffield Centre and the institute are all working on a host of improvement projects. For the few schools found at inspection to be failing to deliver an acceptable standard, the special measures regime is proving effective with more than two-thirds recovering fairly rapidly.

Of course school improvement offers no magic. Its efficacy depends on the knowledge, skills and enthusiasm of those supporting schools and on the quality of the vision and the judgments of those managing them. Both parties need to engage in hard work, systematic attention to detail and careful evaluation. A basic tenet of school improvement is that all schools - no matter how good they are - need to improve. As the Canadian guru of school change, Michael Fullan, says “you don’t have to be ill to get better”.

An interesting challenge exists for schools which routinely admit less advantaged pupils. The National Commission on Education devoted its last project, “Success against the odds” to test this issue. In the case study of a school in south London carried out by Howard Davies (then of the Confederation of British Industry), Sarah Portway of IBM and I, we were particularly struck by: * the quality of leadership; * the clarity of aims; * the common commitment to learning by pupils and staff; * the high expectations - of both pupils and of staff about each other; * the ceaseless search for better administrative systems.

Researchers such as David Reynolds, Michael Barber and Louise Stoll have also asked why some schools fail to improve. Reynolds found staff blamed the conditions of the school for its failure. Barber differentiates between struggling (where there are problems and something is being done about them) and failing schools (where there is a lack of awareness about the poor quality). Stoll identifies a lack of vision and dysfunctional staff relationships. These findings illustrate the need for concerted and committed action.

Colleagues in our International School Effectiveness and Improvement Centre have been collaborating with the Department for Education and Employment and the other education agencies in activities emanating from the Improving Schools programme. This is an innovative attempt to promote collaboration between central and local government, researchers and practitioners. The institute has thus been at the heart of a series of initiatives designed to avoid failure and to build on good practice. These include involvement in seminars on post-OFSTED action planning, on teaching and, with the Organisation for Economic Cooperation and Development, on Combating Failure. With the Banking Information Service we produced a broadsheet on governing bodies and effective schools.

Now, with the support of the Government Office for London, we are establishing a regional leadership centre for current and aspiring headteachers.

Another group is compiling case studies of schools which had been found to be in need of special measures and are now showing steady and - in some cases - spectacular improvement. With colleagues from Keele University we are also studying the relationship between extra-curricular provision and school performance. The aim of collecting data about schools which, despite difficulties, are known to be successful and about those which have emerged as improving is to learn more about the signs of impending failure and how best effective help can be given before the total breakdown of the school.

During the past 10 years numerous top-down changes have been imposed on schools. Some have failed to live up to expectations, but a number are now proving their worth. The rehabilitation of failing schools is perhaps the most dramatic. There is a limit, however, to what any top-down initiative can achieve. The latest evidence from school case studies is revealing new ways of raising standards that have been pioneered - often in adversity - by school staff and governors.

Researchers in British universities have an important role in teasing out and disseminating the practice, and the emerging theory, of successful school improvement.

Professor Peter Mortimore is director of the Institute of Education, University of London. A conference on how weak schools recover is to be held at the institute on July 1. Details: 0171 612 6017