Measurement invariance in international large-scale assessments: Integrating theory and method

This chapter reviews the basic concept of measurement invariance and its challenges with regard to international large-scale assessments (ILSAs). It examines to integrate measurement invariance evaluations with substantive insights into measurement bias. The chapter argues beyond the existing measurement invariance evaluation that a generic latent structural and measurement modeling or the simple structure of a common-factor model is unlikely to yield fully comparable scores within the contexts of ILSAs. It provides a stepwise strategy for evaluating measurement invariance with a productive way of dealing with the unattainable ideal of scalar invariance. The chapter presents an empirical example using ILSA data and discusses some general issues in evaluating measurement invariance across many groups along with considering limitations and strategies. It also presents analyses of questionnaire data and an integration of theory and method in measurement invariance evaluation within the context of ILSAs in order to evaluate the possibility of mean comparisons across countries.