Research shows up difficulties of measuring success.
How well pupils succeed in their GCSE exams can be largely predicted from their home background and their school attainment at the age of 11.
But more effective or less effective secondary schools can make a difference of as much as 10 per cent - equivalent to the difference between seven grade Cs at GCSE and seven Es in the most and least effective schools according to research at the London Institute of Education on 11,000 pupils in 87 Lancashire schools.
The research also shows schools can be more or less effective with different groups of pupils. Both the Lancashire study and further work over three years in 94 inner London schools reveals that schools can be more effective with high attainers than low attainers - or vice versa. Schools may also be more effective with pupils from high or low income groups; girls or boys and with different ethnic groups.
Overall measures of school effectiveness such as numbers obtaining five GCSEs grades A to C may thus be of limited value, the research findings suggest, since they obscure these differences and the fact that different subject departments can also be more or less effective at overcoming disadvantages of background and prior attainment.
Schools can even appear to be more effective in some years than others, prompting the researchers to suggest schools' judgements about their effectiveness should specify: what outcomes they were successful at promoting? Over what period? And for whom?
The researchers suggest schools use a range of value-added indicators such as total GCSE scores; scores for each subject; and scores of pupil groups such as pupils who were high or low attainers on entry; boys and girls; different ethnic groups; pupils on free school meals.
Leadership at the whole-school and departmental level plays an important part in good exam performance. Comparing schools that seemed to be equally effective in all subjects and those with more mixed departmental successes led the researchers to conclude: "While the departmental level was undoubtedly very important, in some schools it was apparently easier for all departments to function effectively, thanks to a more supportive context, shared whole-school emphasis on the importance of pupil learning and achievement, and the apparently beneficial impact of successful departments spurring each other on to further success.
"Conversely, in other schools it was harder for departments to be effective due to lack of overall leadership, shared goals and vision, poor expectations and inconsistent approaches."