Logistic regression and Prediction Configural Frequency Analysis - a comparison

Logistic regression (LR) and Prediction Configural Frequency Analysis (PCFA) are compared. First, the underlying statistical models are presented. Second, sample design matrices are created. Third, data are analyzed using both methods. Two data examples are analyzed. The first is artificial, the second uses data from a project on domestic violence. Fourth, the goals of LR, a variable-oriented approach, and PCFA, a person-oriented approach are discussed. One conclusion of the comparisons is that, for researchers who wish to enrich results by employing both methods, the standard model of LR needs to be extended so that it becomes parallel to the base model of PCFA.