Pugacheva Ekaterina Vladimirovna

Introduction. There are several standardized methods for estimating the age of a skull. Most of these methods are based on the analysis of suture obliteration and the tooth wear scoring. However, many anthropologists prefer a more subjective approach, relying on general impressions without using a set of standardized criteria. This study aimed to assess the effectiveness of a visual method for age estimation and reconstruction of age-at-death structure in a skeletal sample.

Materials and methods. The study was based on a series of 116 skulls from the early 20th century collected by the Peter the Great Museum of Anthropology and Ethnography (Kunstkamera). These specimens had documented sex and age information. Two researchers independently assessed the age of the skull specimens and recorded the degree of suture fusion on the cranial vault as well as the level of tooth wear on the occlusal surfaces. The correlation between age and estimated scores was calculated using Spearman’s rank correlation coefficient. The discrepancy between estimated and actual ages was measured by calculating the mean absolute error (MAE) and systematic error (SE) as the average difference between documented and estimated ages for the entire sample as well as for each age group. Intraclass correlation coefficients were used to assess the consistency of the authors’ estimates.

Results. The authors' estimates showed moderately high agreement among themselves and a moderate positive correlation with actual age. The accuracy of the visual assessments was found to be comparable with that of more formalized methods for assessing the degree of suture obliteration. The estimates also exhibited the phenomenon of regression to the mean, with individuals in younger cohorts being systematically overestimated in terms of age and those in older cohorts being underestimated. The accuracy of determining the age-at-death distribution depends to some extent on the actual characteristics of the sample structure. Averaging estimates from different authors or several estimates from the same author, repeated over a large time interval, makes it possible to bring estimates closer to real data.

Conclusion. Increasing interobserver agreement of age estimates can be achieved by fixing traits on the same point scales, as well as by increasing age intervals. The accuracy of estimates can be improved by repeated age estimation as well as the “wisdom of the crowd” effect. © 2024. This work is licensed under a CC BY 4.0 license