ABSTRACT
Graphic Similarity Analysis, proposed by Waern, is one of the simplest methods for analyzing similarities among the members of a set of objects. In this method, the magnitudes of the similarity measures are utilized in stepwise fashion. Elementary linkage analysis by McQuitty is a simple method of clustering. It can be used to cluster any objects that have distinctive characteristics of similarity. Advantages of elementary linkage analysis are speed, objectivity and its provision for investigating a particular theoretical position. A 15-variable matrix can be analyzed into objectively determined types in 5 to 10 minutes. Furthermore, all elementary linkage analysis operations require only pencil and paper. Graphic methods can be quickly applied to any form of proximity matrix. Proximities can be similarities like correlations or distances. Although McQuitty's method is done easily by hand, a computer solution is available in SAS PROC Cluster Mcquitty.
