Sampling to Redundancy in Zooarchaeology: Lessons from the Portland Basin, Northwestern Oregon and Southwestern WashingtonJournal of Ethnobiology (2004)
A method for determining whether a group of samples is adequate to address a research question is presented. As each sample is analyzed the average value of a variable is recalculated and a cumulative graph is produced. When the value of the average stabilizes, one has empirical evidence that analysis of additional samples is not necessary-one has sampled to redundancy-and the collection is adequate for its intended analytical purpose. Analysis of two zooarchaeological collections of mammalian remains recovered from the Portland Basin of northwestern Oregon and southwestern Washington illustrates this point. Despite the spatial and temporal propinquity of the two sites, one assemblage is adequate for estimating taxonomic richness and diversity whereas the other, larger collection, is adequate for estimating richness but not diversity. Combined, the two collections are adequate for estimating taxonomic richness but do not provide an accurate measure of taxonomic diversity. Graphing procedures for monitoring sample adequacy, if implemented in the field, could help preserve finite archaeological resources.
Publication DateWinter 2004
Citation InformationSampling to Redundancy in Zoorchaeology: Lessons from the Portland Basin, Northwestern Oregon and Southwestern Washington. (second author with R. Lee Lyman). Journal of Ethnobiology 24(2): 329-346.