Mild-restricted confirmatory factor analysis: a new approach to assessing population factor models

Identifier:  imarina:9435520

Handle: https://hdl.handle.net/20.500.11797/imarina9435520

Authors:  Lorenzo-Seva U

Abstract:
Mild-restricted confirmatory factor analysis is a new method for computing confirmatory factor analysis (CFA). In CFA, each variable in the model is usually expected to be a pure indicator of a factor at the population level. However, even if a variable is strongly related to a single factor, it often cannot be considered a pure indicator because the other loading values of the variable are not exactly zero, although they are close to zero. This kind of variable can be said to be a close indicator of a factor. The CFA based on mild-restricted factor analysis helps to assess whether a factor model can be expected to exist at population level even if the variables are close indicators or none of the inter-factor correlations or the correlated errors are exactly zero.