Is the School Composition Effect Real?: A Discussion With Evidence From the UK PISA Data

Abstract

Abstract The school composition effect has attracted attention from researchers who believe that progress towards equality of educational opportunity is compromised by the existence of schools with "unbalanced" social rolls. Evidence for 2 central hypotheses in the "mix" theory is tested by an analysis of UK PISA data. It is suggested that the school composition effect is sometimes an artefact of statistical procedures and caused by "unmeasured" within-SES school selection for noncognitive dispositions and variable family resources within social classes. Some implications for the debate into the origins of social differences in education and the limits of "effective schooling" are discussed.