Enhancing statistical literacy – multivariate thinking in precollege social sciences courses

Introduction Social scientists theorise about issues such as inequalities in health, educational achievement, criminal behaviour or the distribution of wealth. The starting point for theory should be a clear understanding of phenomena. Teaching social sciences at precollege level is a challenge in part because social phenomena almost always involve a number of categorisations in groups, such as by social class, ethnicity or sex. Up to now the complexity of social data in schools has been a substantial obstacle, and the use of data has often been restricted to headline statistics based on aggregated data, and explorations of the behaviour of these groupings considered separately. The Nuffield Foundation has funded a project, Reasoning from Evidence, which is developing some data visualisation tools to support the teaching of GCE A-level Sociology (student age typically 16 – 19), but the tools have much wider application across many subjects and wider age ranges. We will try to articulate the rationale for engaging with the interactions between multiple factors through an example in educational achievement where there are inequalities in sex, social class and ethnic groupings.