Journal of Statistics: Advances in Theory and ApplicationsAuthor's: André Beauducel
Pages: [111] - [135]
Received Date: May 10, 2017
Submitted by:
DOI: http://dx.doi.org/10.18642/jsata_7100121829
In the context of confirmatory factor analysis, the independent clusters model (ICM), a model with non-salient loadings that are fixed to zero, is often tested. However, the ICM has been found to be overly restrictive in several research contexts. The aim of the present study was therefore to propose a less restrictive criterion for parsimony of non-salient loadings. A promising criterion for this purpose is based on ‘buffered scales’, which have been proposed by Cattell and Tsujioka [1] as optimal indicators of corresponding factors. Variables with positive and negative loadings on an unwanted factor are balanced out in a buffered scale, so that the variance of the unwanted factor is at minimum. Model constraints are proposed in order to specify the balance of positive and negative loadings on unwanted factors as a basis for loading parsimony.
confirmatory factor analysis, independent clusters model, model constraints.
