skip to main content skip to footer

Cluster Analysis by Linear Contrasts

Shafto, Michael.
Publication Year:
Report Number:
ETS Research Bulletin
Document Type:
Page Count:
Subject/Key Words:
National Science Foundation (NSF), Cluster Analysis, Correlation, Research Methodology


The purpose of this paper is to suggest a technique of cluster analysis which is similar in aim to the Interactive Intercolumnar Correlation Analysis (IICA), though different in detail. Two methods are proposed for extracting a single bipolar factor (a "contrast compenent") directly from the initial similarities matrix. The advantages of this general approach are that: (a) It helps avoid certain misclassification problems inherent in IICA; (b) It is related in a straightforward way to conventional techniques o multidimensional scaling and therefore allows a unified treatment of dimensional and "typal" structures; and (c) It provides an interesting solution to the problem of relations among linear contrasts based on different subsets of the stimuli. (Author/DJ) (26pp.)

Read More