ABSTRACT
Structure equation modeling (SEM) is probably the most versatile tool to capture multivariate relationships of several variables while maintaining the sequence of relationships in a structural framework, commensuration with the multivariate nature of the problem that mimics the reality. This makes SEM a complex statistical technique involving multiple equations. SEM comprises two multivariate models, namely measurement model and structural model. The measurement model is used to quantify the latent variables (factors) and the structural model is used to evaluate the relationships among the latent variables. The confirmatory factor model (Chapter 13) and the path model (Chapter 10) provide the necessary mathematical foundation for the measurement model and structural models, respectively. In this chapter, we describe the SEM in detail, starting from prerequisites, model conceptualization, model identification, parameter estimation, model adequacy tests, model respecification, and other important issues. Finally, a case study is elaborately discussed linking all the useful concepts and modeling strategies pertinent to the SEM.
