ERIC Number: ED418126
Record Type: RIE
Publication Date: 1998-Apr-13
Structural Equation Modeling versus Ordinary Least Squares Canonical Analysis: Some Heuristic Comparisons.
Dawson, Thomas E.
This paper describes structural equation modeling (SEM) in comparison with another overarching analysis within the general linear model (GLM) analytic family: canonical correlation analysis. The uninitiated reader can gain an understanding of SEM's basic tenets and applications. Latent constructs discovered via a measurement model are explored and the structural models that "connect" latent constructs are described. In addition to reviewing SEM constructs, the paper uses the analysis of a heuristic data set to show how SEM as the most general analytic approach subsumes canonical correlation analysis (CCA) as a special case, even though canonical correlation analysis in turn itself subsumes other parametric methods (e.g., t-tests, analysis of variance, regression, multivariate analysis of variance, and discriminant analysis) as special cases. An appendix contains an illustration that SEM subsumes CCA. (Contains 1 table, 4 figures, and 21 references.) (Author/SLD)
Publication Type: Reports - Evaluative; Speeches/Meeting Papers
Education Level: N/A
Authoring Institution: N/A
Note: Paper presented at the Annual Meeting of the American Educational Research Association (San Diego, CA, April 1998).