A Necessary and Sufficient Identification Rule for Structural Models Estimated in Practice
- 1 July 1995
- journal article
- Published by Taylor & Francis Ltd in Multivariate Behavioral Research
- Vol. 30 (3), 359-383
- https://doi.org/10.1207/s15327906mbr3003_4
Abstract
This article presents a straightforward classification system that is a necessary and sufficient condition for identification of the structural component of structural equation models of the type most commonly estimated - block-recursive models with no more than two equations per block. The technique is graphical and thus does not require actual estimation of the model in question. The article also reviews the limitations of other identification techniques, including econometric and information matrix techniques, when applied to these models.Keywords
This publication has 14 references indexed in Scilit:
- The FC1 Rule of Identification for Confirmatory Factor AnalysisSociological Methods & Research, 1993
- Confirmatory Tetrad AnalysisSociological Methodology, 1993
- Conventional wisdom on measurement: A structural equation perspective.Psychological Bulletin, 1991
- Some Theory and Applications of Confirmatory Second-Order Factor AnalysisMultivariate Behavioral Research, 1988
- Changing a Causal Hypothesis without Changing the Fit: some Rules for Generating Equivalent Path ModelsMultivariate Behavioral Research, 1986
- Some algebraic properties of the Reticular Action Model for moment structuresBritish Journal of Mathematical and Statistical Psychology, 1984
- A note on the investigation of local and global identifiabilityPsychometrika, 1982
- Causal modeling applied to psychonomic systems simulationBehavior Research Methods, 1980
- Causal Models in Educational Research: Nonrecursive ModelsAmerican Educational Research Journal, 1978
- Uncorrelated Disturbances and Identifiability CriteriaInternational Economic Review, 1963