Iterative Temporal Hygiene and Bayesian Analyses of Radiocarbon Datasets: The Impact of Kernel Density Estimation on Clarifying Temporal Relationships among Woodland Period Phases, Middle Scioto Valley, Ohio

Summary

This is an abstract from the "Bayesian Archaeology" session, at the 86th annual meeting of the Society for American Archaeology.

The accumulation of radiocarbon dates for Scioto Valley Woodland period sites has created a palimpsest, which inhibits chronological understanding of cultural change. The project iteratively integrates temporal hygiene and Bayesian analyses of large radiocarbon datasets from multiple sites, in an attempt to clean up problematic features of such datasets and provide for accurate archaeological inference. Temporal hygiene is applied and compared using three levels of stringency, in order to eliminate problematic assays, which result from early low-accuracy radiocarbon dates and non-optimal selection and reporting practices. Kernel density estimation (KDE) smoothing reduces statistical over-dispersion, which results from summing measurement imprecision and issues with the calibration curve. Bayesian analysis is built into the KDE routine, resulting in a more accurate analysis than otherwise possible. This degree of subtlety of the analysis would be impossible if null hypothesis statistical testing (NHST) were used alone, due to inherent conceptual limitations in NHST. Analysis of 22 radiocarbon dates from five sites at the Columbus Southerly Wastewater Treatment Plant, and comparisons of 64 dates from several relevant regional sites, provide for a refined chronology. This chronology better temporally defines previously overlapping Hopewell hamlet occupations, terminal Middle Woodland–early Late Woodland nucleated settlements, and Newtown phase village occupations.

This Resource is Part of the Following Collections

• Society for American Archaeology 86th Annual Meeting, Online (2021)   •
• Bayesian Archaeology

Keywords

Record Identifiers

Abstract Id(s): 32346