Confidence Intervals in the Analysis of Mortality and Survivorship Curves in Zooarchaeology

Abstract

AbstractThe analysis of age-at-death data, derived from epiphyseal fusion and dental eruption/wear patterns, is one of the most powerful tools at the disposal of zooarchaeologists studying past hunting and herd management practices. Zooarchaeologists typically analyze age-at-death data by constructing survivorship and mortality curves in order to allow insight into a variety of ecological and economic relationships between humans and animals. Since adopting such practices in the middle of the twentieth century, zooarchaeologists have proposed several methods for analyzing these curves, including visual examination and hypothesis testing. Creating confidence intervals is complementary to these two methods, allowing practitioners to graphically represent survivorship and mortality while testing hypotheses and accounting for sample sizes, which are often small in zooarchaeological assemblages. We discuss the basic concepts behind the nature of age-at-death data and the analysis of mortality and survivorship curves. We then describe how to calculate confidence intervals using bootstrapping techniques for both dental eruption/wear data and epiphyseal fusion data. To enable future users to replicate our methods, we introduce the freely available online R package “zooaRch” (http://cran.r-project.org/web/packages/zooaRch/ ), which includes a vignette to guide first-time users.