Testing the validity of the DSP2 sex assessment method on two European populations

Sex estimation from human skeletal remains can be achieved using a diverse range of skeletal elements, methodologies, and statistical approaches. While population specificity presents a challenge for most metric sex assessment methods, the os coxae seems to bypass this issue due to its universal pattern of sexual dimorphism. The DSP2 (Diagnose Sexuelle Probabiliste, version 2) is a non-population-specific metric sex estimation method that employs a combination of at least four of ten measurements of the os coxae to estimate sex probabilistically. Despite the method being based on a worldwide meta-database, many populations remain unrepresented. The present study aims to test the DSP2 method on two previously untested European populations. A sample of 51 modern Italian os coxae and a sample of 53 British historical os coxae were used. The validity of the DSP2 method was evaluated through the sexing rate (88.24% for the Italian sample and 86.79% for the British sample) and sexing accuracy (ranging from 95.83% to 100%). The sex estimates were obtained by DSP2 with a 0.95 posterior probability threshold. Intra- and inter-observer error analyses revealed close to excellent reliability for most measurements (with the depth of the great sciatic notch being the only exception). The results of this study support the use of DSP2 on populations not represented in the original datasets, including modern Italian and historical British populations.