Is There a Research Design Role for Sensitivity Analysis (SA) in Archaeological Modeling?

Abstract

The absence of sensitivity analysis (SA) in archaeological predictive modeling, and indeed in contemporary archaeological research design generally, places the discipline in stark contrast to other disciplines routinely employing such approaches. The lack of routinized use of SA diminishes credibility of model outcomes, and without it, loci of model uncertainty remain undetermined—whether attributable to parameterization or other model elements. This essay explores the nature of archaeological modeling goals, the constraints of archaeological data and the impact on uncertainty, and, finally, the potential position and role of SA in archaeological research design. In particular, issues of the tandem application of environmental and sociobehavioral modeling are addressed. It is argued that simplification of both models and expectations may well result in enhanced ability to effectively employ SA in archaeological modeling enterprises, and bolster outcome confidence.