Age‐at‐death patterns and transition analysis trends for three Asian populations: Implications for [paleo]demography

Abstract

Abstract Objectives Age‐at‐death estimation provides biological anthropologists with fundamental biographic information to help resolve various theoretical and practical questions about both individual expression and populational trends in present and past contexts. Particularly, researchers interested in reconstructing past populations' mortality profiles have favored using Boldsen and colleagues' Transition Analysis (TA2), as the method provides both methodological and statistical solutions for optimally treating traditional ordinal, categorical score data from multiple age indicators. In this article, we present an assessment of TA2 to better understand its age estimation patterns and trends when applied to a large‐scale, documented Asian skeletal sample. Materials and methods Specifically, this study examines the patterns of error and bias of TA2 when applied to three Asian populations. We assess cumulative effects of the age estimates in mortality reconstruction, and TA2's ability to accurately characterize the effects of covariates (e.g., sex, stature) that are linked to inequal social status and differential access to resources and known to further cause mortality disparities. Results The results show that age estimates of three age indicators (the pubic symphysis, auricular surface, and cranial sutures) combined with an informed prior produced the most accurate age estimates across all three Asian samples. Thus, the age composition of a target skeletal sample dictates the accuracy of TA2 estimates, and therefore demographic parameters. Discussion Although the exact parameter values may differ from the true values and a consistent underestimation of female age‐at‐death and survivorship is evident, TA2 recovers a realistic approximation of a mortality pattern by revealing (previously hidden) elderly individuals.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 13 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.