On inter-observer error in age estimates based on the adult skeletons: A case study of the Okunev culture sample

Abstract

The study is dedicated to analyzing inter-observer disparities in age estimations based on skeletal remains of varying completeness. Five anthropologists estimated the sex and age of a series of skeletons from two burial grounds of the Okunev culture (Khakassia, Early Bronze Age). Age estimation was conducted in two variants: 1. according to traditional methods used by physical anthropologists; 2. using the "Transition Analysis 3" (TA3) free software, which is based on a random generalized linear model trained on documented information about age-related variability of dozens of features. The intraclass correlation coefficient (ICC) was utilized as a measure of agreement between the assessments. The results indicate a high consistency of assessments obtained within both approaches, with an advantage for TA3. The new method also reveals an additional advantage in estimating the age of individuals older than 50 years. The higher consistency of TA3 assessments is likely explained by the standardized procedure of recording numerous skeletal features with a limited number of options. The consistency of age-at-death structure assessments for the sample as a whole is also noticeably higher when using TA3 compared to the traditional approach. However, in this case, the larger age intervals proposed by TA3 compared to subjectively assessed intervals in the standard approach likely play a key role. Sex assessments coincided among participants in 19 out of 23 (83%) cases. A comparison of morphological and genetic sex estimations obtained for 9 skeletons revealed an average agreement in 84% of cases.