Variable selection in artefact composition studies

In artefact compositional studies, the selection of variables to use in ana lysis is unavoidable. Given this ubiquity, surprisingly little attention ha s been paid to ways variables might be selected. After arguing the case for the importance of variable selection, two systematic approaches to making a choice, which have had little or no application in archaeometry, are disc ussed and illustrated. One, based on the use of principal components, is ap propriate of structure in the data is not known. The other, based on the us e of classification trees, is applicable when there are known or assumed gr oups in the data.