The use of statistical models to determine the age of the deceased and the issue of children’s graves identified by the models as adult burials. Based on the example of burials of Tarnobrzeg Lusatian Culture

Abstract Childhood is the moment in every man’s life when they learn the rules governing the society to which they belong. This process of learning certain principles usually ends when the individual reaches the appropriate age. In primitive communities it was necessary to undergo complex and sometimes painful initiation processes. These moments could be different from one another depending on the community and gender, and what is more, it was often multi-stage. It usually occurred at the age of 14/15 for boys and even a little earlier for girls. It is assumed that up to 60 % of people in the prehistoric populations were under 18 years of age, so it should be assumed that the very moment of initiation could have occurred particularly early. As part of this study, which is the result of NSC grant implemented by the author concerning age and gender determinants in the graves of Tarnobrzeg Lusatian Culture, the author has developed a logit mathematical model that allows to determine on the basis of two variables (vessel size and amount of remains) whether an adult or a child was buried in a given burial. In the case of adults, this model shows very high effectiveness reaching almost 90 %. However, for children, we record ‘only’ 70 % of the correct determinations. There are some indications to believe that among these 30 % false indications there are burials of children who have undergone the initiation process. Confirmation of this thesis and a thorough investigation of this phenomenon are very important because, in a broader source perspective, they can contribute to the discussion on the cultural and biological age in prehistoric populations. It would also create additional opportunities for cooperation between anthropologists and archaeologists, as it is impossible to identify child burials wrongly qualified by the model without the help of the former. In order to verify this regularity, three groups obtained thanks to the model have been assessed in this paper: adults, children correctly indicated by the model and the ones who had been erroneously identified. The tools in the analytical process were various methods of statistical inference, such as: correspondence table, odds ratio chart and chi^2 test. On the basis of the analyses carried out, it was possible to establish that there had been numerous similarities (including similar dimensions and form of urns) between burials of adults and those of children indicated by the model as adults.