A method for constructing life table bands for paleodemographic mortality analysis

Marc A. Luy, University of Rostock
Ursula Wittwer-Backofen, Institute for Human Genetics and Anthropology, University Clinics Freiburg

We suggest a procedure for the construction of life table bands for mortality analysis of archaeological skeletal populations, based on modern demographic and statistic methods. It takes into consideration the uncertainty given by wide individual age ranges and uses Monte Carlo approach. The method is based on random extractions of exact ages at death out of the given age range for each skeleton, reproduced with a sufficient number of extractions. This leads to a relatively narrow bundle of age-at-death distributions, characterising the specific mortality pattern of the skeletal population and, in contrary to the usually used standard methods, is including the individual age estimation errors. The upper and lower limits of the age-at-death band can be defined by the innermost 95 per cent of the age-at-death distributions as base for life table construction. Several simulation tests show that this method yields robust and informative results for paleodemographic mortality analysis.