The use of decision tree analysis for improving age estimation standards from the acetabulum

Area of Science:

• Forensic Anthropology
• Bioarchaeology
• Biostatistical Analysis

Background:

• Accurate skeletal age estimation is crucial in forensic and archaeological contexts.
• Traditional statistical methods for multifactorial age estimation face challenges with non-linear relationships.
• Existing methods like transition analysis have limitations in precision and ease of use.

Purpose of the Study:

• To investigate the utility of decision trees for estimating adult age from skeletal remains.
• To develop and test a decision tree model using acetabular morphological data.
• To compare the accuracy of decision trees against established methods like transition analysis.

Main Methods:

• A dataset of 100 South African acetabula was utilized to construct a decision tree model.
• Morphological features of the acetabulum were employed as variables for age classification.
• A separate test sample of 25 individuals was used to validate the decision tree's performance.

Main Results:

• The developed decision tree correctly classified 80% (20 out of 25) of individuals into broad age groups (young, middle, older adults).
• The remaining 20% of individuals were classified into adjacent age groups, indicating good accuracy.
• The decision tree method demonstrated superior accuracy compared to previous analyses using transition analysis on the same dataset.

Conclusions:

• Decision trees show potential as an accurate and user-friendly tool for adult age estimation from skeletal material.
• This method may better accommodate the complex, non-linear relationship between chronological and biological age.
• Further research is warranted to refine decision tree models and expand their application in forensic anthropology.