Longitudinal studies at the country level as a new approach to educational effectiveness: explaining change in reading achievement (PIRLS) by change in age, socio‐economic status and class size

Abstract

This article investigates the effect of change in average age, socio‐economic status (SES) and class size upon the change in average reading achievement between the Progress in International Reading Literacy Study (PIRLS) 2001 and PIRLS 2006 at the country level. First, we consider the effect of age upon reading achievement. We use regression–discontinuity to analyze the effects of age and schooling upon reading achievement. A tentative correction for age differences yields a new ranking of countries with regard to their corrected average achievement in PIRLS 2006. Second, in order to facilitate cross‐country comparisons in respect of educational equity, we construct internationally acceptable SES indices for PIRLS 2001 and PIRLS 2006. Third, we investigate the effects of change in class size, age and SES upon the change in reading achievement through regressions. We find that the change in reading achievement at the country level can be partially explained by the change in age but not by the change in SE...