ABSTRACT
In this chapter, we present sample global models of CFA. As was explained in Section 2.4, global models of CFA are similar in that all variables have the same status. Thus, there are no predictors or criterion variables, no dependent or independent variables, and no mediator or moderator variables. All variables have the same status. It is the goal of exploratory global CFA to identify configurations (profiles, cells, category patterns, etc.) that stand out because they were observed more often (types) or less often (antitypes) than expected based on chance. The chance concept is expressed using log-linear or other CFA base models that serve to (a) translate the chance concept into some numerically tractable form and (b) estimate the expected cell frequencies.
