Latent Variable Analysis with Applications in Administrative Sciences

Structural equation models and latent variable analysis is a field of data analytics that has undergone substantial developments over the past two decades. These models allow to characterize and relate some factors that are not directly observable. The range of application of such models is wide in social sciences, including marketing, management, IT and human resources. The course will be divided into several parts, including a review of the concepts of regression, correlation, causal relation, direct / indirect effects and correlation diagrams. We will then discuss some specific types of structural equation models such as exploratory/confirmatory factor analysis and we will study the general formulation of the model, characterized by a path diagram with latent variables. Finally, component-based structural equation models will also be discussed, such as partial least squares (PLS) and GSCA.