The time had come to focus on learning rather than performance, said Barbara MacGilchrist, deputy director of London's Institute of Education, in a speech last month. Assessment to judge and enhance the way children learned was most effective in improving schools. Research on school improvement showed:
* No two schools have the same context. A school can have a different impact on different pupils. Comparing ones with similar levels of free meals, as value-added tables do, hides huge differences.
* Change is complex and there are no quick fixes. Schools need long-term strategies for developing their capacity to change and grow.
* Short-term tactics to improve exam results do not lead to sustained improvement over time.
* Continuous improvement is unusual for more than three years, and extremely rare for longer. The static level of key stage 2 English results since 2000 "masks significant volatility locally". This is known as the "wobble effect". About half the schools went up and half down, and it was not the same schools improving or declining over two years.