To weight or not to weight?: the case of PISA data

Abstract

International large-scale assessments (ILSA) –like e.g. PISA, TIMSS, PIRLS, etc.– have obtained a high worldwide popularity among researchers to study students’ academic achievement. Nevertheless, in spite of their recently acquired relevance, there are few studies which really account for the complex survey and test designs that they present and follow the technical procedures suggested by their developers. The current study intends to provide researchers with a comprehensive explanation on how these databases should be dealt with and the drawbacks of not employing the recommended procedures. Furthermore, we also alert about the use of methodologies which would not be adequate in the context of these studies, as the common use of student fixed effects. Hence, the main goal of this research is to provide some kind of ‘good practices guide’ which can be a quick reference to researchers.