Clusterwise three-way component models to identify quantitative and qualitative differences.

Three-way data are encountered when, for instance, a psychologist measures the intensity of behavioral reactions to various situations for different persons. To reveal why people react differently upon a specific situation, one may unravel the underlying mechanisms using three-way component methods, which may give insight into how people appraise situations (i.e., which situation features they generally take into account and which feature applies to which situation), in how their behavioral reactions are organized in response systems, and in dispositional (quantitative) differences between people in how perceptions of situations are linked to response systems. Three-way component methods, however, assume that the same mechanisms are at stake for all persons, which may be an unrealistic assumption. For instance, qualitatively distinct response systems may exist (e.g., some people may only distinguish between verbal and physical channels, whereas others only differentiate betwe en approach and avoidance reactions), and participants may appraise the same situations in a different way, because they focus on different features (e.g., some people may focus on the perception others have from them whereas other people may concentrate on their self-perception). To capture such qualitative differences, we will develop clusterwise three-way component models in which participants are grouped into clusters on the basis of their underlying mechanisms.

Keywords:Clustering, Driewegs data, Analyse, componenten, Verschillen, Kwalitatieve, Kwantitatieve

Disciplines:Biological and physiological psychology, General psychology, Other psychology and cognitive sciences, Applied psychology, Psychological methods, Mathematical and quantitative methods, General pedagogical and educational sciences, Social theory and sociological methods, Political theory and methodology