Chapter 21: General Structural Equation Models

Robert Cudeck & Stephen H. C. du Toit

General structural equation models Structural equation models (SEM) are a collection of statistical models for the analysis of multivariate data. The name ‘structural equation model’ describes the two major elements of the method: first an algebraic representation of latent variables that underlie manifest variables, and secondly a system of linear regressions among the latent variables. The distinction between variables which are actually measured and latent variables which are not is a key element of the enterprise. In particular, the ability of the method to separate true scores from errors of measurement is an invaluable feature of classical SEM. It also partly explains the popularity of the method in the social sciences where theory often pertains to hypothetical constructs and not to specific variables. Many popular statistical methods— multiple regression, classical path analysis, classical test theory models for psychometric problems, factor analysis in all of its versions—are special cases of SEM. ...

Users without subscription are not able to see the full content on this title. Please, subscribe or login to access all content on this website.

• Buy The Book