RELIABILITY TESTING OF METRIC METHODS FOR SEX DETERMINATION IN ANCIENT SKELETAL REMAINS IN GREECE

The skeletal study of archaeological material is the only detailed source of demographic information on ancient populations and investigating sex differences is crucial for the reconstruction of the social structure of past societies. Determination of skeletal sex can be achieved using visual (non-metric) or metric methods. Metric methods are considered more appropriate for assessing sex in archaeological skeletal remains since the accuracy of non-metric methods decreases in cases of fragmented bones. Additionally, it is well established that the expression of sexual dimorphism is population specific. Therefore, sex prediction equations should be used only when the sample is known to come from the same population from which the functions were derived. The aim of this study is to test the application of sex prediction equations, which were produced using measurements from the arm bones of a modern Greek population, in ancient Greek skeletal remains. For the purpose of this research five ancient populations were examined; Ancient Corinth, Corfu, Agia Triada Thebes, Edessa and Thebes. According to our results, ulnar sex prediction equations cannot be considered adequate for sex determination of ancient skeletal remains. On the contrary, humeral as well as radial sex prediction equations can be considered adequately reliable for sex determination of ancient skeletal remains. More specifically, sex prediction equations containing the humeral vertical head diameter, its combination with the humeral epicondylar width as well as the maximum radial distal width, achieve a classification accuracy over 72%