ABSTRACT
This chapter introduces readers to configural frequency analysis (CFA) and its application in clinical neuropsychology. CFA is a method for the detection of types and antitypes in cross-classifications. Types are described by patterns of categories of discrete variables that occur more often than expected from chance. Antitypes are described by patterns that occur less often than expected from chance. Application of CFA is of interest when the focus is on people rather than variables
What Goes Together and What Does Not Go Together-The Person Perspective
Empirical research has a number of preferred perspectives. Most prominent is the causal perspective where researchers conduct investigations to identify the causes of the phenomena under study. For example, a number
1 Please address correspondence to Alexander von Eye, Michigan State University, Department of Psychology, 119 Snyder Hall, East Lansing, MI 48824-1117,
of genetic predispositions have been identified as sufficient causes for the development of Alzheimer’s disease. It is well known that the empirical identification of causes poses major challenges to research. Bollen (1989) stated that the following three conditions must be met for researchers to be able to label variables or events as causes of other events: (a) isolation, (b) association, and (c) directionality. Isolation is typically taken care of using experimental methodology. Directionality has proven largely elusive (see, e.g., von Eye & Schuster, 1999), Association is the most frequently applied concept in empirical research. The present chapter is concerned with a particular facet of association, the local association (see later discussion; Havránek & Lienert, 1984; von Eye & Brandtstädter, 1982). There are associations that are present in a particular segment of the data space only. In other segments, the variables are independent.
