Cluster Analysis

A multivariate statistical technique which assesses the similarities between units or assemblages, based on the occurrence or non-occurrence of specific artifact types or other components within them. It also involves comparing the distances between points or objects, whose dimensions are measurements or scores for a number of variables. Cluster analysis results are normally plotted as a dendrogram, a treelike representation of the distances between objects in hyperspace. Items that are closer together are deemed to be more closely related. Researchers select a case by random sampling and then include contiguous cases as part of the sample.

Technique of multivariate analysis, which compares the distances between points, objects or items, distributed in a hyperspace whose dimensions are measurements or scores for a number of variables. Cluster analysis results are normally plotted as a ‘dendrogram’, a tree-like representation of the distances between objects in hyperspace. Items that are closer together in hyperspace are deemed to be more closely related, and are linked more closely in the dendrogram. Like any other multivariate technique, cluster analysis could conceivably be done by ‘pencil-and-paper’ methods; it is the number of calculations involved that usually necessitates the use of a digital computer.