Archaeological Applications of Factor, Cluster, and Proximity Analysis

Abstract

AbstractAn effort is made to give readers a good idea of what the techniques of factor, cluster, and proximity analyses are; of what are their good features and their limitations; and where one should look for further information. Archaeological uses of these techniques which are discussed include those of Hodson on La Tene brooches and on Mousterian assemblages, L. and S. Binford on Mousterian assemblages, Freeman and Brown on the Carter Ranch Pueblo, and the author on data from Teotihuacan, Mexico. The dangers of using correlations based on inadequate samples are stressed, and one necessary (though not sufficient) condition for sample adequacy is suggested. It is argued that proximity analysis, among other applications, is probably better than the Brainerd-Robinson approach for seriation.