Choosing a multivariate model: Noncentrality and goodness of fit.

Abstract

It is suggested that Akaike's information criterion cannot be used for model selection in real applications and that there are problems attending the definition of parsimonious fit indices. A normed function of the noncentrality parameter is recommended as an unbiased absolute goodness-of-fit index, and the Tucker-Lewis index and a new unbiased counterpart of the Bentler-Bonett index are recommended for those investigators who might wish to evaluate fit relative to a null model

Cites (3)

1. A new look at the statistical model identification
   Akaike, H. · 1974 · IEEE Transactions on Automatic Control
2. Model Selection and Akaike's Information Criterion (AIC): The General Theory and its Analytical Extensions
   Bozdogan, Hamparsum · 1987 · Psychometrika
3. Factor Analysis and AIC
   Akaike, Hirotugu · 1987 · Springer Series in Statistics

Show all 3 references

Cited by (77)

1. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives
   Hu, Li‐tze, Bentler, Peter M. · 1999 · Structural Equation Modeling A Multidisciplinary Journal
2. Evaluating Goodness-of-Fit Indexes for Testing Measurement Invariance
   Cheung, Gordon W., Rensvold, Roger B. · 2002 · Structural Equation Modeling: A Multidisciplinary Journal
3. Fit indices in covariance structure modeling: Sensitivity to underparameterized model misspecification
   Hu, Li-tze, Bentler, Peter M. · 1998 · Psychological Methods
4. Sensitivity of Goodness of Fit Indexes to Lack of Measurement Invariance
   Chen, Fang Fang · 2007 · Structural Equation Modeling: A Multidisciplinary Journal
5. Evaluating the Fit of Structural Equation Models: Tests of Significance and Descriptive Goodness-of-Fit Measures
   Schermelleh-Engel, Karin, Moosbrugger, Helfried, Müller, Hans · 2003 · Department of Psychology - University of Koblenz-Landau

See all 1,438