Tomarken, Andrew J. and Waller, Niels G. (2005). Structural Equation Modeling: Strengths, Limitations, and Misconceptions. Annu. Rev. Clin. Psychol.1: 31-36.

Abstract

Structural Equation Modeling (SEM) is a second-generation multivariate statistical data analytic technique that is often used in current research in both natural and social sciences. This is mainly due to the believe that it can test theoretically supported linear and additive causal models. In the field of marketing, for example, with SEM marketers can virtually examine the relationships that exist among variables of interest in order to prioritize resources to better serve their customers. Additionally, the fact that unobservable, hard-to-measure latent variables can be used in SEM makes it ideal for tackling business research problems (Wong, 2013).

          In principle, SEM involves observable or manifest variables or indicators and unobservable or latent variables, causal modeling, confirmatory factor analysis,          multiple regression, and path analysis. SEM technique is used mainly to see if empirical data fit the proposed causal model by the examination of x² and indexes of goodness of fit.