ABSTRACT
Classification-pattern finding and pattern matching-is at the heart of all human knowledge, philosophical, scientific, or commonsense knowledge (Everitt 1993:1, Kaufman and Rousseeuw 1990:1, Spiggle and Sanders 1984:337; also Lévi-Strauss 1966: chapter 2). Cluster analysis is a quantitative classification technique used for descriptive and explorative purposes. It is known as an interdependence method, where no variables are singled out as dependent or independent variables. As a result of a cluster analysis, a total sample of subjects is divided into groups or clusters of subjects. A cluster can be defined as ‘a continuous region of a p-dimensional space (with p being the number of variables included) containing a relatively high density of points, separated from other such regions by regions containing a relatively low density of points’ (Everitt 1974:44).
