Premium
Estimating sex using isolated appendicular skeletal elements from Chachapoyas, Peru
Author(s) -
Anzellini Armando,
Toyne J. Marla
Publication year - 2019
Publication title -
international journal of osteoarchaeology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.738
H-Index - 60
eISSN - 1099-1212
pISSN - 1047-482X
DOI - 10.1002/oa.2810
Subject(s) - appendicular skeleton , postcrania , context (archaeology) , frequentist inference , forensic anthropology , logistic regression , bayesian probability , estimation , statistics , sample (material) , demography , archaeology , geography , biology , mathematics , bayesian inference , ecology , paleontology , sociology , chemistry , management , chromatography , taxon , economics
Abstract The estimation of sex from human skeletal remains is one of the primary steps in the creation of a bioarchaeological demography. Conventionally, sex has been estimated from the crania or os coxae, but commingled contexts pose challenges when associating these key elements with the rest of the remains becomes impossible or if the sample is mostly represented by postcranial appendicular elements. Following successful applications of logistic regression to appendicular metrics in forensic anthropology for the estimation of sex using appropriate reference samples, this study aims to create a similar method to be applied in archaeological contexts in the Andes. This is achieved through the collection of osteometric data from complete burials of previously estimated sex recovered at the archaeological site of Kuelap, in Chachapoyas, Peru. Using these data, we created logistic regression formulae, using both frequentists and Bayesian approaches, that utilize standard metrics of the humerus, femur, and tibia to estimate the sex of individuals in this sample, testing their apparent accuracy and, through their application to a commingled context, their applicability in situations when only appendicular remains may be present or complete. Both frequentist and Bayesian‐derived formulae provided a correct classification of between 82% and 93% for the previously estimated sample, on par with correct classification rates of cranial nonmetrics. No significant difference was observed between the Bayesian approach and the frequentist approach. The application of this method for sex estimation to a commingled context also provided a demography consistent with that created using only os coxae and cranial remains. These results underscore the validity and reliability of this methodology and provide further support for its application in other bioarchaeological contexts across the Andean region.