Abstract
Multilevel CFA models (MLV CFA) modeling permits more sophisticated construct validity research by examining relationships among factor structures, factor loadings, and errors at different hierarchical levels. In the MLV CFA models, the latent variable or variables have two kinds of elements: 1) the between-group elements (Level 2 or higher level) and 2) the within-group elements (Level 1 of lower level). The between-group elements represent the general part of the model and the within-group element the individual part. The within-level variation includes an individual-level measurement error variance, which generally expands the impact of the within-level variation to the intraclass correlations. Multilevel CFA therefore generates results corresponding to those generated by perfectly reliable measures. If the same measurement model is specified across levels, by defining each item loading to be invariant with its across-level counterpart, the researcher can equate the factor scales across levels. Thus, the factor variances at different levels are directly comparable. The fit of this constrained MLV CFA model can be evaluated by comparing it with an unconstrained model specified with freely estimated factor loadings at each level. In the present work the steps of the above procedure are fully described and additional issues relevant to the use of MLV CFA are discussed in detail.
Highlights
The purpose of this study is to describe the procedure of Multilevel Confirmatory Factor Analysis Modeling (MLV CFA, Byrne, 2012), i.e., how to incorporate the multilevel approach into a CFA model
More complex data structures like cross-classifications, multiple-memberships or covariates are only few of the possible extensions of the basic CFA models developed. Another feature of the multilevel CFA modeling is the disintegration of the total variance (Ψ) of the latent variables into the part attributed to between-cluster variation (ΨB) and the part attributed to within-cluster variation (ΨW)
The MLV CFA procedure assumed that latent factors contain between- and within-group elements
Summary
Real world data are frequently structured in multiple levels Research in psychology deals with designs about individuals acting within a context, like families in the previous example, or schools (see Byrne, 2012; Geiser, 2013), organizations (see Brown, 2015; Darlington & Hayes, 2017) or neighborhoods (see Tabachnick & Fidell, 2013). Models used to analyze clustered data are called Multilevel Models, Hierarchical Linear Models, Random Coefficient Models, or Mixed Models (Geiser, 2013; Field, 2013 among many others). The purpose of this study is to describe the procedure of Multilevel Confirmatory Factor Analysis Modeling (MLV CFA, Byrne, 2012), i.e., how to incorporate the multilevel approach into a CFA model
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
