Using structural equation modeling (SEM) in educational research: Practices and Issues

Abstract

Structural equation modeling (SEM) is a method for analyzing multivariate data from both nonexperimental and experimental research. This method combines a measurement model linking observed variables with latent variables and a structural model linking latent variables. The use of SEM in social science and educational research has grown since the early days in 1980s and offers promise as a method that is useful in theory-based research. Using a data set for analysis, this paper presents a step-by-step, non-technical introduction on how the SEM technique is used. Throughout the paper, practical suggestions are given in order that researchers could use the SEM technique quickly. The paper concludes with information on relevant internet resources for further study and research. [PUBLICATION ABSTRACT]