Waves of educational reform have come and gone over the past 20 years, with only modest success. We have had prescription, informed professionalism and the more recent enthusiasm for good schools helping weaker schools, the federation of multiple schools and so-called superheads. Last week's announcement about policies for low-achieving schools shows the continued dominance of this thinking.
But one school feeding another its ideas is a blunt instrument for change. Research and experience tells us that there are bigger differences within schools than between them. A school is an aggregate of good and bad practice. And children do not learn in schools - they learn in classrooms.
There is an alternative: to use each school's internal variation as the engine of its own improvement - the within school variation approach. This has been tried by only a small number, largely because it is a difficult road to travel. Differences between departments in secondary schools - as between individual teachers in primaries - are harder to address than lumping all staff together. If differences between schools' effectiveness have been difficult for education professionals to handle, differences within schools are even more difficult.
There has been some acknowledgement of the extent of variation within schools - which is larger in the UK than most other societies - and hints about its causes, ranging from the historic structural looseness of the component parts of British education to the effects of social class. But there has been no road map for schools wanting to travel this road less travelled. Until now.
The within school variation project, based at the National College for School Leadership, has produced such a map for schools. In our sample of 15 secondaries, we found that as variation within the school reduced, value-added achievement increased.
From the outset, it was vital to establish a culture of openness and collegiality, so we used exemplary individual teachers and departments to set the tone. Training was off site to free people from the psychological baggage of their regular environment.
We used rich, relevant and reliable data systems to identify each school's best practice. It is essential when sharing data to develop clarity about its use. We used it to track pupils to provide the ammunition for change and the best practice against which to benchmark.
We trained middle managers in data analysis and the transfer of good practice through classroom observation. By making them buddies, we improved practice and got them to see themselves as middle leaders, not middle managers. We focused on teaching and learning through specifying core teacher behaviours - high-quality questioning, good time and classroom management and behaviour control - and encouraged consistency within and between departments.
In our sample schools, we made sure everybody was doing things the same way to minimise variation and standardise "the best practice" for all teachers. And we listened to the consumers of education, that is the pupils and their parents, who assess education policies as experienced, not simply as intended.
So what did we find were the advantages of schools looking within themselves, rather than having practices and people imposed from outside? Obviously, encouraging professionals to learn from their colleagues carries a positive emotional tone, although it may lack the appeal - for politicians, at least - of dramatic closure plans. In addition, focusing on departments or individual teachers means that policy can get closer to influencing the really important determinants of pupil achievement.
The within school variation approach starts from the premise that every school has, within itself, practice that is relatively more effective than the school's "average", and that every school can learn from itself. In fact, many have their own world-class practice to learn from.
The advantages are manifold. Units such as departments - or individual teachers in primary schools - are small and therefore easier to change than whole-school staff groups. Teachers are also more influenced by policies that are closer to their classrooms than the school-level solutions beloved of policymakers.
Encouraging schools to learn from their own best practice means no one need wait for help from the school down the road. It is also a sustainable strategy, likely to long outlive the effects of flying visits from superheads. And there are no excuses for poor results, since departments in a school should have the same kinds of pupils.
Another thing in its favour is that the performance data needed is already routinely collected, and the national performance data could quickly have a within school variation measure based on core subject variation. Since schools are fully in charge of their own performance, everyone should be pleased at this addition.
It is right that education policies challenge us; so they should. But we need to ask serious questions about the usefulness of those policies. Just what is the point of encouraging teachers to learn only from someone else's school, when they could learn much more, far quicker and with greater profit from the very best practice in their own? And if this is true, why are we not introducing policies to help our schools learn from their own very best practice?
l'Schools Learning from Their Best' by David Reynolds: www.ncsl.org.ukmedia539AEwithin-school-variation-2008.pdf. This article by David Reynolds is based on research by Ray Tarleton and Peter Kent
Don't be modest, page 30
David Reynolds Professor of education at the University of Plymouth
Ray Tarleton Principal of South Dartmoor Community College, Ashburton, Devon
Peter Kent Headteacher of Lawrence Sheriff School, Rugby, Warwickshire.