≪Organizer Abstract≫
Recently, more and more researchers are applying the structural equation modeling (SEM) technique in various fields of psychology. SEM is an integrative framework of a multivariate analysis that combines the functions of confirmatory factor analysis and path analysis. This technique has many advantages over traditional approaches, such as simultaneous testing of measurement and causal models, correction of correlation attenuation caused by measurement errors, model comparison based on fit indices, flexible handling ability for irregular data (e.g., categorical data, missing data, and non-normal data), and testing of measurement invariance in multiple population data. On the other hand, SEM is often misused owing to insufficient knowledge. In this symposium, we will report the current status of the contributions and misuses of SEM in various fields of psychology and discuss the technique's promises and pitfalls.
Advanced Search