Estimating the length of a period by means of 14C-tests

Abstract

This paper presents a method which aims to create a set of guide-lines for estimating the fit between two irregular probability distributions. One, the observed, is made up of the sum of the probabilities of any number of ¹⁴C-tests describing the length of a period during which a certain context was formed. The other, the expected, is made up of the probabilities connected with a
series of points on the cali-bration curve, e.g., the 11 BP-values (assigned the average standard deviation of the observed tests) that match the calendar years 105, 115, ... 205 AD. This series is considered to be a model of a period with a linear relationship between the course of time and the production of ¹⁴C-test material. The discussion is centred around the construction and the interpretation of a conventional index. It draws upon one example each from the Mesolithic and Migration/Vendel Periods.