Date of Award
Doctor of Philosophy
Richard L. Jantz
Murray K. Marks, Walter E. Klippel, Edward F. Harris
This study determined that morphometric and morphoscopic variables of the human mandible can be used to estimate the partial biological profile of an individual. Specifically, these mandibular variables were used in linear discriminant function analyses designed to estimate the biological affinity or sex of an individual, with or without biological a priori knowledge. The study data set is composed of 17 world samples including U.S. White, Black, and Hispanic individuals, prehistoric and proto-historic Native American groups, Southeast and Northeast Asian peoples, a Central American group, and a Nubian group. Eleven metric measurements were utilized: eight are standard measurements; two were designed for this study; and one was modified from its standard definition. Six morphoscopic variables were employed; most were analyzed as defined in the literature, though several involved expanded definitions and scoring categories. Specific definitions and graphics are provided for the measuring and scoring procedures.
These variables were used in three types of analyses – a morphometric analysis, a morphoscopic analysis, and both types together, forming a morphometroscopic analysis. The combination of morphometric and morphoscopic data into one statistical analysis is a relatively novel approach to the analysis of human remains. Four major combinations of samples were used in the study: all groups, groups of individuals that are forensically interesting, groups that are potentially closely associated, and single groups through time.
Results indicated that sex can be estimated using the morphometric data with high accuracy rates, typically 83% to 89%. Morphoscopic data produced sex estimation viii accuracy rates between 63% and 81%. Combining the two data sets produced accuracy rates in excess of 90%. Analyses examining the estimation of biological affinity were successful. Morphometric data produced better accuracy rates than morphoscopic data, but the morphometroscopic data sets were the most accurate. Linear discriminant functions using morphometroscopic data produced biological affinity accuracy rates that were frequently four times better than expected by random chance, and were often seven times greater than chance alone. Overall, the study concluded that mandibular morphometric and morphoscopic variables were useful for the estimation of sex and biological affinity in a variety of world populations.
Berg, Gregory Eugene, "Biological Affinity and Sex Determination Using Morphometric and Morphoscopic Variables from the Human Mandible. " PhD diss., University of Tennessee, 2008.