CONTEXTUALIZATION OF SCHOOL EXAMINATION RESULTS 1992

Early in 1992, a group of nine secondary schools within one local educ ation authority (LEA) decided to investigate ways of improving on the 'league table' approach to looking at GCSE results for their schools. As a result of LEA decisions taken several years earlier, almost all t he 1,400 pupils due to take these examinations in summer 1999 had take n the same test at or near the stage when they entered secondary schoo l. The availability of these test scores, combined with the willingnes s of the schools and the LEA to collate other pupil and school informa tion, enabled the study described here to be set up. Multilevel modell ing techniques were used to try to derive measures of relative effecti veness for these schools, taking into account some of the characterist ics of their pupils. The results show not only that 'league tables' ba sed solely on GCSE results are misleading, but also that the ranking o f schools differs markedly depending on the measure of outcome being c onsidered.