Assessing the Impacts of Method and Skeletal Elements on Sex Estimation: Working Towards a Nonbinary Approach

Accurate sex estimation is a key part of constructing biological profiles in forensic anthropology and bioarchaeology, yet traditional binary approaches often oversimplify human variation. This study evaluates the accuracy, sex ratio outcomes, and utility of four established methods: Albanese’s 2003 non-group-specific pelvic and femoral and 2013 upper-limb metric methods, Albanese et al.’s 2005 sample-specific limb method, and Sutherland and Suchey’s 1991 ventral arc morphoscopic standard. Using a documented sample from the Robert J. Terry Anatomical Collection (n = 122), the results show that Albanese’s 2003 pelvic and femoral method yielded the highest overall accuracy (95.9%) and a balanced sex ratio, outperforming the upper-limb method (95.1%), which exhibited a male accuracy bias. While articular joints provided more stable estimations than diaphyseal midshafts, visual scoring proved limited, as 19.7% of the documented males exhibited ridges resembling a female ventral arc. Notably, applying a probabilistic “grey zone” (p-values of 0.2–0.8) demonstrated that sex exists on a continuum rather than a strict binary. Ultimately, sex estimation varies by the skeletal elements and criteria utilized. Implementing a probabilistic framework reduces the risk of misclassification, accounts for biological complexity, and provides a more reliable, ethically grounded approach to forensic analysis.