CLUSTER ANALYSIS

If you question sick patients about their symptoms, you will undoubtedly have a very long list of complaints. You may think that you will find as many combinations of symptoms as you have patients. However, if you study the types of symptoms that frequently occur together, you will probably be able to put the patients into groups ” those who have respiratory disturbances, those who have gastric problems, those who have cardiac difficulties. Classifying the patients into groups of similar individuals may be helpful both for determining treatment strategies and for understanding how the body malfunctions.

In statistics, the search for similar groups of objects or people is called cluster analysis. By forming clusters of objects and then studying the characteristics the objects share, as well as those in which they differ , you can gain useful insights. For example, cluster analysis has been used to cluster skulls from various archeo-logical digs into the civilizations from which they originated. Cluster analysis is also frequently used in market research to identify groups of people for whom various marketing pitches may be particularly attractive.