Accuracy of the Phenice (1969) and Klales et al. (2012) sex estimation methods in a Chilean population

Abstract

ABSTRACT Chile went through violent deaths and disappearances during the dictatorship (1973–1990). Many of those victims are still unrecovered or unidentified. The standards currently used to estimate skeletal sex in unidentified remains in Chile are not population-specific. Phenice’s (1969) method for sex estimation is a widely accepted standard, while Klales et al. (2012) modified Phenice’s method by developing a logistic regression equation for estimating sex in a US population. Both methods are accurate, with reported 96% and 86.2% overall correct classifications, respectively. However, due to differences in levels of sexual dimorphism between populations, the accuracy of these methods can vary when applied to groups outside the reference sample (e.g. non-US individuals). Thus, this study aims to assess the accuracy of these standards in a Chilean population. A sample of 265 coxae of adult individuals from the documented General Cemetery of Santiago collection were examined. Visual evaluations were performed following both methodologies. Phenice’s (1969) method was 97.0% accurate, with a sex bias of 7.7%, and Klales et al. (2012) method was 87.2% accurate, with a sex bias of −15.4%. Both methods can be applied to unknown Chilean remains with acceptable accuracy (>85%), but the sex bias (>5%) could lead to estimation errors.