The reliability of sex determination of skeletons from forensic context in the Balkans

In this study we have tested the applicability of morphological methods for sex assessment, based on seven pelvic and nine cranial traits, using contemporary Balkans Population. The material involved in the study comprises 262 pelvic bones and 180 skulls of male individuals front two mass graves in Serbia. The material was examined separately by an experienced and an inexperienced physical anthropologists. Sex was correctly estimated by the experienced anthropologist in 100% of individuals using all of the 16 pelvic and cranial criteria. In fact, sex differences in pelvic morphology were large enough to allow sexing the individuals with 100% accuracv. Among seven features observed on the pelvic bones, the least reliable single sex indicator was the width of the great sciatic notch (with accuracy of 79.15%). Looking at the skull alone. sex was correctly determined in 70.56%, cases. It was shown that the most accurate single indicators among cranial methods was the robustness of the mandible (with accuracy of 70.93%) while the sharpness of the supraorbital margins was the least reliable indicator demonstrating accuracy in only 28.15% of crania. Examination of the sample by an individual with training in physical anthropology, but no case experience, suggests that experience is likely to contribute moderately to the accuracy of the sex determination. Namely, the inexperienced anthropologist accurately assessed the sex of the sample 95.04% of the time; 4.06% less accurate than the experienced anthropologist. The two anthropologists showed the least agreement in scoring the ventral arc and composite arc on the pelvic bones. (C) 2004 Elsevier Ireland Ltd. All rights reserved.