Anthropology Insider - Spring 2007
Recent Conferences - American Association of Physical Anthropology
Erica Chambers, Editor | Dr. Jeff McKee, Faculty Advisor
Selection of variables for discriminant analysis of human crania for determining ancestry.
An almost endless number of measurements may be made of the human cranium. Most anthropologists use only a small standard set of measurements. These only require spreading and sliding calipers, relatively inexpensive instruments found in most biological anthropology laboratories. Unconventional cranial measurements require costly instruments such as the radiometer and coordinate caliper that are not common in laboratories. These measurements also are more difficult to complete. Which cranial measurements are best for classifying when constructing a discriminant function? Will the use of variables requiring more time, training, and equipment improve discrimination? 65 measurements were taken from 155 crania of European American, African American, and Coyotero Apache ancestry. Measurements were separated into four subsets for statistical analysis: (1) FD2 (1996), (2) Howells (1973), (3) Gill (1984), and (4) All Measurements. Forward stepwise methods were used to develop a predictive discriminant analysis using SPSS 14.0. Classificatory power of discriminant functions was determined using the Leave- One-Out hit-rate estimates for each subset: (1) 85.5%, (2) 90.3%, (3) 61.9%, (4) 93.5%. Results lead me to suggest that biological anthropology laboratories should purchase radiometers and coordinate calipers. They record data missed with spreading and sliding calipers. In addition, standard measurements may be combined with nonstandard measurements to produce more powerful discriminant functions for determining ancestry. This study was funded in part by The Connective Tissue Graduate Student Research Award (2004-007).
A. Kolatorowicz. Department of Anthropology, The Ohio State University.