STATISTICS IN ARCHAEOLOGY

Robert D Drennan, University of Pittsburgh, Pittsburgh, PA, USA

© 2008 Elsevier Inc. All rights reserved.

Quantitative analysis Analysis of information that comes in the form of numbers, relies heavily on the tools of statistics. sampling bias Selection of a sample in such a way that some members of the population are less likely to be included than others.

Vagaries of sampling The variation that can be expected to occur by pure random chance between different samples selected from the same population.

Archaeological data are irrevocably (although not exclusively) quantitative in nature. The phenomena that archaeologists work with (including artifacts, ecofacts, features, remains of architecture, and many more) are classified, counted, and measured in various ways. The results are numbers, often quite a lot of numbers. Describing things quantitatively, then, along with finding patterns in and comparing numbers, are essential archaeological tasks. Statistical analysis is especially associated with certain schools of thought in archaeology; for example, it was strongly advocated as processual archaeology developed (see Processual Archaeology). Counting and measuring different kinds of things found in the archaeological record, however, are so fundamental to the process of using material remains to reconstruct past human activities, that statistical analysis cannot be ignored by any school of thought in which it matters to know what people did in the past. The importance of statistical analysis in archaeology is evidenced by the number of books published in recent years with the purpose of introducing and explaining the tools of statistical analysis in a specifically archaeological context. Any of these works can be consulted for further discussion of aspects of statistical analysis touched on here. Traditional statistical tools and terminology developed between about 1920 and 1950 are most often used in archaeology, although the perspectives and vocabulary of the more recent ‘exploratory data analysis’ school are becoming more widely known.