Identifier

etd-03182013-190045

Degree

Master of Arts (MA)

Department

Geography and Anthropology

Document Type

Thesis

Abstract

The assessment of ancestry from skeletal remains is a vital aspect of forensic anthropology. As such, a myriad of techniques exists for estimating this particular component of the biological profile. The most traditional of these methods utilizes the naked eye and the observer’s experience. As replicability has become more important, objective, metric techniques have been developed. This study attempts to merge these two subfields: by taking a traditionally non-metric feature, palate shape, and using a computer, evaluating it quantitatively. Using 3D digitizer technology in conjunction with shape matching and machine learning methods common in computer science, palate shape curves were collected from 376 individuals of varying backgrounds from mixed historic and modern contexts. Additionally, measurements were taken to capture palate depth, which is a novel measurement in this study. Results of the computer analysis indicated palate shape was an accurate indicator of ancestry 58% of the time. This number improved slightly when the historic sample was examined on its own (61%), but not to such a degree as to indicate a significant difference. This result may indicate that secular change in the human skeleton is not affecting this region, or at least that secular change does not affect the shape of the palate as it relates to ancestry. Cluster analysis of the curves revealed that the parabolic, hyperbolic, and elliptical shapes are relatively discrete from one another, with the only major overlap in shape being between white and Hispanic individuals. The results regarding depth are rudimentary at this stage; however, results indicate that the depth of the palate in Hispanic individuals is significantly deeper than in other ancestry groups.

Date

2013

Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Manhein, Mary H.

DOI

10.31390/gradschool_theses.1540

Download

Share

COinS