Abstract:
Structural equation modeling (SEM) is a method for analysis of multivariate data from both non-experimental and experimental research. This method combines a structural model linking latent variables and a measurement model linking observed variables with latent variables. The use of SEM in social science and educational research has grown over the years 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. The paper concludes with a discussion of pertinent issues that potential SEM users should attend to.
