Abstract
Eighteen archaeological objectives are ordered within two general classes: Class A, placement of archaeological units—traits, types, and assemblages — within a space-time framework, and Class B, integration of the units in a cultural context. Four levels of interpretation of data are distinguished: identification, unidimensional distribution, bidimensional distribution, and processual analysis. Statistical methods appropriate to the different levels of interpretation, including cluster analysis, matrix analysis, and factor analysis, are discussed. It is suggested that the matrix method provides a statistical model of findings in archaeological field work. It is concluded that archaeological theory can be advanced by interpretation in terms of series rather than in terms of classes.