Measurement invariance of non-cognitive measures in TIMSS across countries and across time. An application and comparison of Multigroup Confirmatory Factor Analysis, Bayesian approximate measurement invariance and alignment optimization approach

Abstract

Measurement invariance is a key topic in international large-scale assessments (ILSA). The study investigates measurement invariance across countries and over time of noncognitive measures in TIMSS, as well as comparing different approaches of measurement invariance analyses. Multigroup Confirmatory Factor Analysis (MGCFA) is the standard procedure to test exact measurement invariance. Due to criticism of this approach in the context of ILSAs, the novel approaches alignment optimization and Bayesian approximate measurement invariance method were applied. Principal, teacher, and student data of the TIMSS cycles 2007, 2011 and 2015 of 26 countries was used. MGCFA and Bayesian approximate measurement invariance analyses indicates configural invariance in all analyses across countries and a variety of invariance levels (mostly metric) over time. The results of the alignment optimization method show for most scales that latent means are comparable across countries - under the assumption of approximate measurement invariance