The 87th Annual Meeting of the American Association of Physical Anthropologists (2018)

Estimating Age: Reporting Strategies and Observer Reliability


Anthropology, University of Tennessee

April 12, 2018 , Zilker 1/2/3 Add to calendar

Age-at-death estimation is important for building a biological profile for unidentified human remains. Most anthropologists agree that using data from several aging methods is ideal but difficult since different studies utilize different statistical methodologies. A recent study found that final age ranges are often selected based on experience, by considering the results of aging methods along with individual familiarity of human variation, which does not meet the standards of the Daubert ruling. Effects of observer experience on final age results have not been formally quantified. Additionally, inter-observer reliability for different methods is not well understood.

The present study addresses effects of observer experience on age estimation by analyzing final age estimations of 13 observers with varying experience levels. Each observer conducted 6 age-at-death estimation methods and was asked to report result via two comprehensive age reporting approaches (overlap of ranges and experience-based) on two skeletons from the Bass Donated Skeletal Collection at the University of Tennessee, Knoxville. Accuracy of all age ranges was assessed as well as observer agreement between the 13 participants.

Results show that the overlap strategy was more accurate amongst the observers than the experience-based age ranges for both skeletons. However, some individual methods provided more accurate estimations. Cronbach’s alpha reliability test indicates that there is excellent agreement between observers with an alpha value of 0.98. Experienced-based age ranges are not the most accurate and different observers score consistently with regards to one other. These results have implications for producing age-at-death estimations in a forensic context.