Structural Equation Modeling in Personality Research

Structural equation modeling (SEM) refers to a family of statistical techniques that test theoretical models specifying causal relationship among a number of observed and unobserved variables. The fundamental purpose of SEM is to determine whether a hypothesized model explains patterns of relationship among variables in sample data. Once a model is specified, model fit is assessed by determining the degree to which covariances in the sample data resemble those implied by the specified model, as well as by the standardized parameter estimates for each hypothesized causal path.