Development and Application of the Trapezoidal Model for Archaeological Chronologies

Abstract

Single- and multiphase models have been used extensively in construction of chronologies. We model more gradual transition between phases with a trapezoid model since it better reflects the nature of the information that goes into the model prior. We find that a simple trapezoid model has a bias that does not reflect prior knowledge, and thus propose an addition of a noninformative element to the prior. We also present an alternative parameterization, which transforms the current abrupt transition model into a model that allows for gradual changes. The addition of a noninformative prior ensures model flexibility. We evaluate these Bayesian models using 2 case studies.