Forensic Anthropological Research in Indonesia: Developing a Pelvic Sex Estimation Standard for the Contemporary Population Using Clinical MSCT Scans

Estimating skeletal sex is critical in the positive identification of an unknown individual. A popular visual standard for the pelvis is that of Klales et al., (2012). Classification accuracy against the original study is reduced when the standard is applied across populations, emphasising the need for appropriate modification to achieve appropriate classification rates. Age is also known to affect the expression of pelvic skeletal dimorphism – features become ‘masculinised’ with increasing age. As Indonesia lacks forensically applicable standards, the present study aims to assess the performance of the Klales standard and the global-population model of Kenyhercz et al., (2017) in a contemporary Indonesian population. This study comprised 378 adult (213 female; 165 male) clinical MSCT scans. Indonesian-specific forensic models for the statistical estimation of skeletal sex are thereafter derived.
Results from this study indicate reduced accuracy in the global-population standard when applied to the Indonesian sample (86.0% compared to 95.9%). For the Indonesian-specific models, the three-trait function had an accuracy of 94.4% and a sex bias of –5.5%. As age was shown to significantly affect the distribution of pelvic trait scores, age-dependent models were derived. The model for individuals aged 30–49 years was the most accurate, at 93.1% with a sex bias of –4.0%. Sex bias shifted to 4.1% in the model for individuals aged ≥ 50 years, indicating greater female misclassification. Findings from this study support population-specific models to improve the capabilities of forensic practitioners in Indonesia to facilitate statistically and judicially appropriate estimations of skeletal sex.