Ancestry Variation in the Accuracy of Rogers's Method of Sex Estimation

Rogers’s (1999) method of human skeletal sex estimation evaluates morphological variation in four traits of the distal humerus. Although this method has the potential for widespread application in forensic and biological anthropological contexts, previous tests have been unable to replicate Rogers’s initial accuracy rate of 92%. Additionally, the role of ancestry in the accuracy of the method has not been sufficiently explored. This study expands on previous blind tests of Rogers’s (1999) original method, though it differs methodologically from prior studies (Ammer et al. 2019; Falys et al. 2005; Harrison 2017; Horbaly et al. 2019; Rogers 2009; Tallman & Blanton 2019; Vance et al. 2011; Wanek 2002; Watkinson 2012) by explicitly controlling for ancestry (85 American Black and 114 American White individuals, as defined in the Hamann-Todd Osteological Collection), by seriating humeri according to trait expression, and by using logistic regression in addition to chi-square and Fisher’s exact tests for analyzing the results. The findings determined that the method was 67% accurate overall and that correct classifications were 2.03 more likely for American Whites than American Blacks, posing an important consideration for practitioners of this method.