Structural Equation Modeling - Introduction 2500-PL-PS-SP15-17

Psychological constructs such as intelligence, neuroticism or trust cannot be observed directly – instead, their presence is inferred from manifest variables, such as item responses. Structural equation modeling (SEM) is a statistical technique that is designed to test the relationships between observed and latent variables. Combining the features of factor analysis (the measurement part of model) and linear regression (the structural part of a model), SEM allows for separating measurement error, as well as testing complex relationships between latent variables.

This course aims to introduce students to SEM – its underlying logic, assumptions, and application. Throughout the consecutive classes, participants will be presented with different kinds of research questions that may be addressed with SEM. We will start with an overview of SEM applications. Next, we will discuss Confirmatory Factor Analysis (CFA), path analysis, and full SEM. The classes will involve a combination of lectures and lab sessions focusing on specification, estimation, and interpretation of structural equation models. All analyses would be performed in R lavaan.