ABSTRACT
The importance of classification in science and in general and behavioral science in particular has already been remarked upon in Chapter 12, in which techniques were described for examining multivariate data to discover whether the data consisted of a number of relatively distinct groups or clusters of observations. In this chapter, a further aspect of classification will be discussed, namely that when the groups are known a priori. Such data arises when investigators collect samples of multivariate observations from several different populations, for example, observations on a number of symptoms for patients from different diagnostic categories.
