Age‐at‐death estimation in archaeological samples: Differences in population means resulting from different aging methods can be predicted from the mean ages of method‐specific reference samples

Abstract

Age mimicry is a well-known phenomenon in the application of osteological ageestimation methods. Age mimicry refers to the fact that predicting age-at-death from a specific trait (age indicator) based on the relation observed in a specific reference sample implies that age estimates to some degree reflect the age structure of the reference sample. In particular, the estimated population mean in a target population in which an age-estimation method is applied is shifted towards the mean in the method-specific reference sample. Consequently, differences in population means between different age-estimation methods in the same target population may be due to differences in mean age of the reference samples used to develop the age-estimation methods. We aim at quantifying the expected magnitude for such differences. Fifteen different traditional age-estimation methods were applied to a sample of 675 adult individuals from the early medieval cemetery of Mannheim-Seckenheim. The relation of the observed estimated population age means and the mean age in the reference samples was analyzed by linear regression. We find that up to 80% of the variation in the estimated population age means can be explained by the variation of the mean age in the reference samples. Furthermore, differences in the magnitude of 3 to 4 years in the mean age between two reference samples can imply a 1-year difference in estimated target population age means. Because large differences in mean age between reference samples used to develop different ageestimation methods are common, some care is needed in interpreting differences between individual age estimates or population mean age estimates in cases where different age-estimation techniques are used.