ABSTRACT
This chapter presents multivariate techniques for data reduction and pattern recognition for large datasets. It discusses regression-based techniques, such as principal components analysis (PCA) and other varieties of factor analysis. This chapter outlines the assumptions of these multivariate models and some ways of using them appropriately with archaeological data. In addition, this chapter examines some common situations in which one’s data may violate the assumptions of parametric multivariate statistical models and it presents correspondence analysis (CA) as a non-parametric alternative.
