Increasing the Robustness of Meta-analysis Through Life History and Middle-Range Models: an Example from the Northeast Pacific

While most comparative and meta-analysis in archaeology would like to assume that taphonomic factors act randomly and do not bias the results of studies with many data points, archaeological records may suffer from systematic biases in preservation, sampling, recovery methods, analytic methods, reporting practices, or yet other factors. Using a life history and middle-range perspective, we outline an approach for assessing possible systematic biases and for explicitly evaluating factors that affect assemblages included in comparative analysis at any scale. We demonstrate the usefulness of the life history concept as a framework for holistically evaluating bias with a zooarchaeological case study from the northeast Pacific. Our comparative analysis of regional fishbone records shows that unusually high abundances of sablefish (Anoplopoma fimbria)—a nutritious and highly valued, yet rarely reported, species—at one large Native American village on the coast of Washington State, USA, cannot be explained by post-depositional destruction, screen size effects, sample size effects, or differences in fishbone identification methods. Though this study focuses on zooarchaeology, the framework we present has potential value for any large-scale meta-analysis that seeks to identify cultural and environmental patterns in “noisy” archaeological data.