NotesFAQContact Us
Collection
Advanced
Search Tips
Peer reviewed Peer reviewed
Direct linkDirect link
ERIC Number: EJ857026
Record Type: Journal
Publication Date: 2009
Pages: 25
Abstractor: As Provided
Reference Count: 36
ISBN: N/A
ISSN: ISSN-1070-5511
Causal Indicator Models: Identification, Estimation, and Testing
Bollen, Kenneth A.; Davis, Walter R.
Structural Equation Modeling: A Multidisciplinary Journal, v16 n3 p498-522 2009
We discuss the identification, estimation, and testing of structural equation models that have causal indicators. We first provide 2 rules of identification that are particularly helpful in models with causal indicators--the 2C emitted paths rule and the exogenous X rule. We demonstrate how these rules can help us distinguish identified from underidentified models. Estimation and testing of identified models with causal indicators is straightforward. For underidentified models with causal indicators we develop "partially reduced form" and "zero error variance" models that are implied by the original model and use these to estimate and test aspects of the original model. We can test the overall fit of the model, the proportionality constraints sometimes implied by a latent variable with causal indicators, and the statistical significance of parameter estimates. Two empirical examples illustrate the procedures. Our results substantiate that it is feasible to incorporate causal indicators in modeling efforts. (Contains 9 figures and 23 footnotes.)
Psychology Press. Available from: Taylor & Francis, Ltd. 325 Chestnut Street Suite 800, Philadelphia, PA 19106. Tel: 800-354-1420; Fax: 215-625-2940; Web site: http://www.tandf.co.uk/journals
Publication Type: Journal Articles; Reports - Descriptive
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A