Comparability and Measurement Invariance

The chapter discusses the cross-national, cross-time, and cross-study comparability of quantitative research from the statistical perspective focusing on the measurement invariance assumption. The assumption of measurement invariance is elaborated. The chapter offers an overview of statistical approaches to testing and handling measurement invariance including (i) classical, and (ii) partial invariance approaches based on Multiple-Group Confirmatory Factor Analysis (MG-CFA), (iii) approximate invariance models based on Bayesian Structural Equation Modeling (BSEM), and (iv) partial approximate invariance method that combine MG-CFA with alignment optimization. The newest approaches are described and critically analyzed. The review shows potential problems and limitations of presented modeling approaches and provides directions for further studies necessary for further development of quantitative comparability studies.