A Value added’ analysis of the 1996 primary school performance tables

Summary In March 1997 the British Government published the first primary school performance tables giving school‐by‐school results in the 1996 end of key stage 2 (KS2) assessments. The tables were widely criticized for failing to provide a measure of a school's effectiveness since ‘good’ KS2 results might simply reflect a ‘high’ ability pupil intake rather than any characteristics of the teaching or management of the school. This paper uses multi‐level modelling techniques to compare the results from the performance tables in one local education authority with a ‘fair’ analysis as recommended by the School Curriculum and Assessment Authority's National Value Added Project. The results reveal significant differences between schools’ raw results as given in the performance tables and ‘fair’ comparisons of their effectiveness in relation to the ability of their pupils. The analysis also indicates significant variation between schools in their effectiveness in different subject areas (English, mathematics and science), suggesting important information is lost by aggregating results across subjects. Finally, the analysis also provides evidence of under‐achievement among some groups of pupils, with pupils with statements of special educational need, bilingual pupils needing English language support, pupils entitled to free school meals and pupils of Caribbean origin all achieving lower KS2 test results than would be expected from their reasoning ability.