The explanation of social differences in reading attainment: An inspection of the PIRLS New Zealand data

Abstract

The newly released PIRLS data allow secondary analyses of group disparities in reading attainment to be conducted with a representative national sample of 10-year-old students. It is found that reading attainment is markedly associated with social class and ethnic origin. A standard regression model suggests that the most powerful indicators are those pointing to the family rather than the school. The argument is, however, critical of the positivist assumptions of statistical modelling and gives considerable space to a discussion of how such empirical findings can contribute to a realist explanation of social differences in reading.