Abstract
The article employs exploratory structural equation modeling (ESEM) to evaluate constructs of economic, cultural, and social capital in international large-scale assessment (LSA) data from the Progress in International Reading Literacy Study (PIRLS) 2006 and the Programme for International Student Assessment (PISA) 2009. ESEM integrates the theory-generating approach of exploratory factor analysis (EFA) and theory-testing approach of confirmatory factor analysis (CFA). It relaxes the zero-loading restriction in CFA, allowing items to load on different factors simultaneously, and it provides measurement invariance tests across countries not available in EFA. A main criticism of international LSA studies is the extended use of indicators poorly grounded in theory, like socioeconomic status, that prevent the study of mechanisms underlying associations with student outcomes. This article contributes to addressing this criticism by providing statistical criteria to evaluate the fit of well-defined sociological constructs with the empirical data.
Acknowledgements
During the course of this investigation Daniel Caro was a research analyst at the IEA-DPC. The authors would like to thank participants of the symposium “Measuring Student Home Background in Large Cross-National Studies: Conceptual and Methodological Issues” at the Annual Meeting of the American Educational Research Association 2012 and two anonymous reviewers for their helpful comments.
Notes
1. Heuristically, values of TLI and CFI greater than 0.90 and 0.95, respectively, reflect acceptable and excellent fit to the data, while for the RMSEA, values less than 0.05 and 0.08 reflect a close fit and a reasonable fit to the data (Marsh, Hau, & Wen, Citation2004).
2. Higher interfactor correlations introduce problems for supporting discriminant validity, and it is recommended not to employ factor analysis for interfactor correlations larger than |0.5| (Gorsuch, Citation1983). Cross-loading magnitudes greater than |0.3| are indicative of complex factor structures and complicate the interpretation of the rotated solution (Sass & Schmitt, Citation2010).
Additional information
Notes on contributors
Daniel H. Caro
Daniel H. Caro is a Research Fellow at the Oxford University Centre for Educational Assessment (OUCEA), Department of Education, University of Oxford. He completed a PhD in Education at the Freie Universität Berlin and a Master’s degree in Interdisciplinary Studies at the University of New Brunswick. He is alumnus of the International Max Planck Research School on the Life Course (LIFE). His research interests include education inequality, international large-scale student assessments, quality of examination marking, mixed models in cross-sectional and longitudinal settings, and causal inference with observational data.
Andrés Sandoval-Hernández
Andrés Sandoval-Hernández is the Head of the Research and Analysis Unit at the International Association for the Evaluation of Educational Achievement in Hamburg, Germany. He earned a PhD in Education from the University of Bath, UK. Andrés has worked as a research associate for the Universidad Iberoamericana and for The Latin American Faculty of Social Sciences. His research work deals with comparative analyses of educational systems using large-scale assessment data with a focus on educational inequalities.
Oliver Lüdtke
Oliver Lüdtke is a Professor of Psychological Research Methods at the Humboldt-Universität zu Berlin. He received his PhD in Psychology from the Freie Universität Berlin and worked as a research scientist at the Center for Educational Research at the Max Planck Institute for Human Development. His main research interests include the application of multilevel modeling in psychological and educational research, international student achievement studies, and personality development in adolescence.