The political context of the new centre is clear. Since 1 January, if school inspectors on a four-yearly visit discover that a school is failing, it has only eight weeks to say how it will improve. The school's plan then goes to the secretary of state for education who, if he or she does not think it will work, can call in a five-member team of managers to do the job instead.
Few schools will reach that point, argues Barbara MacGilchrist, head of in-service training at the Department for Education and one of the centre's founder directors. But all schools could be improved; the difficulty, when local education authority teams of advisers are being cut back, is for schools to find outsiders to support and prod them to change for the better.
Over the past year, pilot school-improvement programmes, led by institute staff, have been running in the London boroughs of Haringey and Lewisham. None of the schools involved has been under inspectorial threat, but most, struggling with inner-city deprivation and disruption, show plenty of room for improvement.
The first stage, according to Mrs MacGilchrist, is to understand what makes a good school. Recent research here and in the United States and Canada has produced a consistent list of characteristics - such as teachers working in teams and high parental involvement - that distinguish effective schools from poor ones.
Next comes the method for change. 'While we knew what made effective schools, we didn't know how to improve them,' Mrs MacGilchrist says. 'What we are learning now is that by focusing on small groups of pupils, by thinking big but starting small, you can help schools to improve themselves.'
The Haringey and Lewisham schools have focused on on a range of issues - from stopping Muslim girls being bullied by involving them in playground games, to revising the teaching of writing and maths. In each case, as teachers begin to develop better learning and teaching for a small group of children, they find this ripples out to change the schooling of every child.
What have yet to be found, according to Mrs MacGilchrist, are precise ways of measuring how much better a school has become - not solely from teachers' perceptions or only from test results and league tables, but with a mix of measures such as how interested pupils are in their lessons and how good relationships are among staff.
The Lewisham project, which involves all 100 schools in the borough, compared with a handful in Haringey, aims to make 18-month surveys of pupils' progress. The new centre will offer value-added analysis of test results, as well as before-and-after analysis of children's work, support and advice to teachers and heads plus a newsletter about research findings and particular successes.
It takes months to achieve measurable results and probably years to turn a failing school into an effective one. Would any secretary of state have patience with such a long-term approach? 'The past failure of improvement projects has been because they have been imposed from above,' Mrs MacGilchrist says. 'We think the key to sustained change is to change the culture of the school, so that teachers feel they own the changes, that they matter to them.'
12 characteristics of effective schools:
Good leadership, subtle but compelling
Pupils feel involved in running the school
Rewards and incentives, rather than punishments
Good behaviour among pupils
Parental and community involvement
Teachers working as a team
Carefully worked out curriculum
Good record-keeping and monitoring of progress
Good physical environment
Shared values: a sense of 'This is how we do things around here'
(Photograph omitted)Reuse content