Concluding remarks

Abstract

In these concluding remarks, it is asserted that cluster analysis is a method of studying heterogeneity, in our case as it applies to behavioral and medical disorders. Several of these disorders are characterized by extensive variability such that while the core disorder may be present, there is extensive variability with regard to its specific features. It is also noted that cluster analysis is a method of discovery that should be accomplished at the preliminary stages of an investigation preceding formal hypothesis testing utilizing the information the cluster analysis provides. The major point of a cluster analysis can be said that of providing evidence that the noted heterogeneity is not random, but within it are meaningful subgroups that potentially may evolve into established subtypes. Following an introduction to the mathematical and statistical bases of cluster analyses, examples are provided of its application in several areas including schizophrenia, learning disability, memory development, homelessness, and traumatic brain injury.