Whether the goal is to estimate the age of a single individual or those in a larger skeletal assemblage, accurate and precise estimates are desired but cannot be consistently produced for adults in all parts of the lifespan using existing methods. In 2014, National Institute of Justice (NIJ) grant #2014-DN-BX-K007 was awarded for the development of an age-at-death estimation method for modern forensic cases using data from a suite of traits throughout the skeleton in a new analytical framework. The project combined the characterization of novel skeletal features, investigation of various statistical approaches for generating age estimates for unknown individuals from reference data, and development of software to simplify the generation of estimates by forensic practitioners. In 2018, a reference dataset containing TA3 trait data for more than 1,700 documented individuals from four continents—North America, Europe, Africa, and Asia—was released. Although the investigation of methods and development of software are ongoing, a trait manual, data collection form, and beta version of a software program (Beta Version 0.8.4) have been released for public testing.1,2
The current TA3 traits include 74 binary and ordinal features, many of which are scored bilaterally. All traits are visible with the naked eye and no specialized equipment or destructive sampling are required for data collection. Trait scores, along with some descriptive fields (e.g., case information, bone condition, number of bones available for analysis), can be recorded using the TA3 Trait Manual1 and recorded on the genetic data collection form. An age estimate can be generated using these data by entering the scores into the TA3 software2 using one of two different entry methods.
Despite the similarity in names, TA3 uses a different approach for estimating age than the "transition analysis" method used in TA2 (ADBOU). Preliminary testing of the current beta software program, presented at the American Academy of Forensic Science (AAFS) in 20223, has demonstrated that the statistical model selected and the data provided in the software output have ramifications for method validation and use in forensic casework. Furthermore, testing indicated that the TA3 software (Beta Version 0.8.4) produces estimates that are inappropriate forensic or archaeological contexts and that the program should be modified before stable release. Despite this, other recent work has verified that the TA3 traits collectively provide age-informative data for the entire adult age life span4 and that the traits provide a way to estimate age from adult skeletal remains when standard features, such as the pubic symphyses and iliac auricular surfaces, are unavailable for analysis.5 Ultimately, the TA3 traits and the NIJ reference dataset are valuable tools in the process of improving adult age estimates. Although the TA3 software should not be used in the current form, it is recommended that the TA3 trait data be collected, as the trait definitions will not change and new methods for generating age estimates are currently being developed.
This presentation will provide a brief history of the TA3 project from its beginnings in the late 1990s to its current status two decades later. Participants will learn about the iterative process used to identify traits and develop definitions and will be introduced to the TA3 trait manual and generic data collection form. The structure and use of the TA3 software (Beta Version. 0.8.4), including its output and limitations, and recommendations for data collection and future research will also be discussed.
Detailed Learning Objectives
- Attendees will be able to describe the process and samples used to develop the TA3 traits and the modern reference dataset.
- Attendees will understand the current capabilities and limitations of the TA3 program (Beta Version 0.8.4).
- Attendees will be able to critically discuss the potential of the TA3 traits and avenues for future work.