Go Back Research Article February, 1980

Multivariate analysis with latent variables: Causal modeling

Author Peter M. Bentler

Abstract

Examines multivariate analysis with latent variables and, more narrowly, linear structural equation models (simultaneous equations, path analysis, structural relations, and covariance structures) with unobserved or unmeasured variables. Basic concepts are discussed: model specification, research design model testing, and model reconfirmation and modification; and research on attitude–behavior relations is reviewed within this context. An overview of the field of multivariate analysis is given, including a guide to the literature, a discussion of professional issues, and recent technological developments in mathematical structures and statistical problems.

Keywords

Multivariate Analysis Latent Variables Causal Modeling Structural Equation Modeling (SEM) Path Analysis Simultaneous Equations Covariance Structures Unobserved Variables Model Specification Model Testing Model Modification Attitude-Behavior Relations Research Design Statistical Methods Mathematical Structures Technological Developments
Details
Volume 31
Pages 419–456
DOI 10.1146/annurev.ps.31.020180.002223
ISSN 1545-2085
Impact Metrics