1Department of Anthropology, University of Toronto, 2Forensic Science Program, University of Toronto Mississauga
April 16, 2020 , Platinum Ballroom
In forensic anthropology, ancestry assessment describes population affinity using morphological and metric analyses. However, morphological analyses are particularly difficult to reproduce and standardize, since they strongly depend on an anthropologist’s subjectivity, based on their experience with human variation. The purpose of this research was to improve the rigour of morphological analyses of ancestry by using three-dimensional (3D) technology to quantify relevant features on the human skull. The sample consisted of 50 European-American, 24 Canadian Inuit and 13 African-American adult female crania, for a total sample size of 87 individuals. The samples were imaged using photogrammetry, the 3D models were constructed in 3DF Zephyr, and the shape analysis was performed in 3DS Max. The data were statistically analysed using a non-parametric multivariate analysis of variance (PERMANOVA), a linear discriminant function analysis (LDA) and a principal component analysis (PCA). Results showed that major differences between groups were clearest when 3D measurements were combined. Overall, European-Americans were statistically different from the other two groups, while Canadian Inuit and African-American individuals were harder to distinguish statistically. The current method was satisfactory in presenting a classification rate of 87.36% (jackknifed: 80.46%) and an average repeatability of 97%. Nonetheless, this project had some limitations. Future research should evaluate the technique with a larger sample size, more diverse populations, other ancestry-related cranial traits (e.g., oval window), and other 3D measurements (e.g., volume). Despite its few limitations, this new and simple method of 3D shape analysis shows promise for the future of ancestry assessment via 3D technology.
This research was supported by the Social Sciences and Humanities Research Council of Canada (SSHRC).