Measurement invariance of the action competence in sustainable development questionnaire: can we compare between groups?

Abstract

There is a need for valid and reliable instruments to assess learning outcomes in education for sustainable development (ESD). Measurement invariance (MI) needs to be established before results of these instruments can be validly compared between groups. Despite its importance, establishing MI is an often overlooked validation step. To provide an example of testing for MI in the field of ESD, thereby setting a precedent for subsequent studies, this study tested the MI of the Action Competence in Sustainable Development Questionnaire (ACiSD-Q). First, Confirmatory Factor Analyses (CFAs) were carried out to make refinements to the original model. Subsequently, nested multiple-group CFAs were performed to test for MI of the ACiSD-Q for the grouping variables of educational level (primary versus secondary), language spoken at home (instructional language versus other languages), and gender. Model fit indices of the refined model and the ACiSD-Q’s proneness to MI are discussed.