TOP-DOWN approaches to school improvement such as action zones, beacon schools and performance-related pay do not work, Professor Peter Mortimore, head of the Institute of Education at London University and president of the British Educational Research Association told the conference.
"We know it has to be bottom up and has to involve people. People have to have a sense of ownership to have the dedication that's needed," he said.
Some schools were more effective than others and all had a duty to be as effective as possible, but school improvement could not overcome all the effects of social disadvantage. "Intake is the dog that doesn't bark," he said.
The professor, who retires this year, believed school leaders were key players and had to involve staffs in the costs and benefits of improvement.
Picking up Professor Gray's finding, he repeated: "The best schools at improvement and effectiveness have involved the pupils in it, the governors and the parents too. If you do not involve them, you're heading for big trouble."
Teachers had to make the aims of learning more explicit for pupis, such as showing them what an A-pass answer looked like. Pupils had to adopt better learning strategies and take more personal responsibility.
A study of two schools in London and two in Singapore which had been turned around, partly with new heads, revealed almost identical approaches.
"They set about motivating staff, they focused on learning and teaching, they improved the environment and they all believed they had to change the culture in the school to make learning the thing that all the kids were locked into," he said.
But there was no way British schools could match the motivation of a country which had moved so rapidly from a peasant to an information society.
In Singapore, there was less concern with intelligence and more on hard work and getting help and coaching. Two-thirds of their pupils gained GCSE benchmark passes against 47 per cent in England. "It becomes much more of the norm that you do succeed in school," he said.
The conclusion was that turning schools around can take three to four years, but even then it was impossible to overcome all disadvantage.