ACCUSATIONS about stingy educational research, usually levelled by researchers themselves because they had too little time and money, do not apply to the school effectiveness project (page three). It cost Pounds 500,000 and allowed a longitudinal study, debarred to most research because of the cost. The results ought to be examined by proponents of school targets, by “behaviourists” who say that pupil performance is entirely a matter of the socio-economic make-up of the school and by teachers who believe they know what their pupils think of them.

For almost every group in education there are messages, often uncomfortable. If research has something to tell the system (and if not, why is public money spent on it?) then the work by John MacBeath, Peter Mortimore and their colleagues deserves more than the cursory glance practitioners usually give academic studies.

A main lesson is that schools, even primaries, are complex institutions where the temptation to define performance and targets numerically and in comparison with other schools has to be tested, if not resisted outright. On the other hand, standards vary among schools and subject areas within schools. By definition as well as desire, performance can be enhanced.

The question is how to isolate aspects that need attention and to arrive at sensible, agreed measures for improvement. Clearly, the Government’s target-setting does not yet satisfy the conditions found in the research.