This study tested the applicability of the optimized summed scored attributes (OSSA) method for sex estimation, using five cranial traits, given the methodological similarities between classifying sex and ancestry.
The optimized summed scored attributes (OSSA) method was recently introduced and validated for nonmetric ancestry estimation between American Black and White individuals. The method proceeds by scoring, dichotomizing, and subsequently summing ordinal morphoscopic trait scores to maximize between-group differences. The current study used a large sample of documented crania from Japan and Thailand (n = 744 males, 320 females) to develop a heuristically selected OSSA sectioning point of less than or equal to 1 separating males and females. This sectioning point was validated using a holdout sample of Japanese, Thai, and Filipino (n = 178 males, 82 females) individuals. The results indicate a general correct classification rate of 82 percent using all five traits, and 81 percent when excluding the mental eminence. Designating an OSSA score of 2 as indeterminate is recommended. (publisher abstract modified)