What is structural equation modeling?

Structural Equation Modeling (SEM) is a statistical technique used to test and estimate causal relationships using a combination of statistical data and qualitative causal assumptions. This technique is commonly used in the social and behavioral sciences to test complex hypotheses about the relationships between multiple variables, including both observed variables (measured directly) and latent variables (constructed from observed variables). It combines aspects of factor analysis and multiple regression, allowing researchers to model complex relationships and assess the fit of a proposed model to the observed data. SEM is used for path analysis, confirmatory factor analysis, and mediation analysis.