ABSTRACT
Cluster analytic methods can be classified as being either model based or non-model based. This chapter focuses primarily on non-model-based cluster analytic methods as these are the most commonly used. Model-based clustering methods, such as finite mixture modeling, utilize probability models, whereas non-model-based cluster analytic methods do not utilize such statistical models. Because a statistical model is not utilized, non-model-based methods are not formally considered to be inferential statistics and are more appropriately classified as numerical algorithms. Readers interested in model-based methods are referred to, for example, Everitt and Hand (1981).
