1School of Social Sciences, Humanities, and Arts; School of Natural Sciences, University of California, Merced, 2Anthropology, University of Tennessee, Knoxville, 3Anthropology, California State University, Sacramento
Thursday All day, Plaza Level
Bioarchaeology has witnessed limited application of transition analysis for estimating age-at-death in past populations. Working with ancient populations appears to restrict the relevance of the method due to difficulties in locating informative priors. Building upon our previous work, which demonstrated that both an informative and a uniform prior are appropriate with the Suchey-Brooks method in bioarchaeological contexts, this paper extends the same methodology to the Lovejoy et al. (1985) auricular surface aging technique.
A documented historic Italian sample from the island of Sardinia was scored for auricular surface aging (n=404). An informative sample (documented historic sample from Torino, Italy) and a uniform sample (Terry Collection) provided the priors. Transition analysis performed on the Sardinian sample caused us to collapse phases VII and VIII into one stage to generate appropriate ages-of-transition. Gompertz hazard parameters were calculated from each prior. The highest posterior density regions for each prior were computed and subjected to cumulative binomial tests to determine: 1) which prior performed better, and 2) how the priors performed in relation to the traditional auricular surface age ranges. Cumulative binomial tests showed the priors worked similarly in the females, while in the males the informative prior functioned slightly better. In both males and females, transition analysis using either prior out-performed the traditional auricular surface age ranges. The results here support earlier work by demonstrating both informative and uniform priors are appropriate for Bayesian analysis. Furthermore, applying transition analysis to bioarchaeological populations, utilizing either prior, will increase the accuracy of aging.
The data collection was funded by a grant for Samantha M. Hens from the Office of Research and Sponsored Projects, California State University, Sacramento.