Reconsidering the auricular surface as an indicator of age at death.

Using standards established by Lovejoy et al. to estimate age at death from auricular surface morphology, 266 individuals of documented age, sex, and ancestry from the Terry and Bass Donated Collections were scored. Analysis of covariance (ANCOVA) indicates that for the factors that could be controlled, age is the sole influence on auricular surface morphology. Ancestry and sex had no significant effect on auricular phase expression. No evidence of secular changes was detected when comparing the Terry Collection (early 20th century) to the Bass Collection (later 20th century). Pearson correlations reveal that several of the subcomponents of the auricular surface (superior and inferior demifaces, left and right sides, transverse organization, texture) correspond with age equally well, although a combined scoring of all features performs slightly better than any one indicator taken alone. Not surprisingly, only 33% of the sample was correctly aged when using the 5-year age ranges provided by Lovejoy et al., suggesting that the published ranges are much too narrow to be used in forensic contexts. To assess the variation in age per phase, standard descriptive statistics and error ranges were calculated and can be employed by forensic anthropologists when estimating the age of an unidentified decedent. Because the mean ages of some of the eight phases did not differ significantly from one another, a modified six-phase system is presented. The auricular surface performs as well as any other single skeletal indicator of adult age. This research suggests that a statistically-informed approach should be taken in order to fully understand the drawbacks and limitations of any aging method.