Statistics

The science of making valid inferences about the characteristics of a group of persons or objects on the basis of numerical information obtained from a randomly selected sample of the group. Artifacts, ecofacts, features, sites, etc. can be reduced to a series of measurements, analytically determined values, or systematic observations which can be represented as numbers. The distributions of items with respect to these variables can then be studied. A wide variety of quantities summarizing the distributions may be calculated and compared, possibly determining the degree of similarity between distributions. The use of statistics procedures in archaeology most often involves assumption because archaeologists are dealing with remains which are only representative of the original population of artifacts, ecofacts, features, sites, etc.

The study of numerical facts. Artefacts, sites, skulls, items etc are reduced to a series of measurements, analytically determined values or systematic observations which can be represented as numbers. The distributions of items with respect to these variables can then be studied. A wide variety of statistics, quantities summarizing the distributions, may be calculated (e.g. mean and standard deviation) and are themselves compared, as representatives of the distributions. This may determine the degree of similarity between distributions. Various statistical ‘tests’ can also be carried out to determine the ‘reality’ of any differences or similarities. Best-known of these are the Chi Square test and Student’s ‘t’ test. The degree of relationship (correlation) between variables may also be calculated. Most conventional statistics of this kind are carried out on one or two variables. Large numbers of variables can be treated by multivariate analysis. Any kind of statistics must be applied carefully to archaeological information. Most procedures involve some kind of assumption and clear thought is required as to how valid it may be. Of prime concern is any assumption that the sample which remains, after the ravages of time and archaeologists, is representative of the original population of artefacts, sites, animals or plants. Archaeology is by its very nature unable to test this and any archaeological conclusions, even if non-statistical, are subject to uncertainty as a result.