Abstract
Metric and discriminant function analyses of the skull have been used successfully to determine ancestry and sex from human skeletal remains in both forensic and archaeological contexts. However, skulls are frequently discovered in damaged condition. One structure that is commonly fragmented, even when the rest of the skull is preserved, is the zygomatic arch. The bizygomatic width is an important measurement in craniometry and in forensic facial reconstruction for determining facial width; therefore we propose a simple linear regression model to predict the bizygomatic width of skulls with damaged zygomatic arches. Thirty-one adult skulls originating from the Indian sub-continent were used to measure the bizygomatic width. Then, on the same skulls, a straight steel wire was placed at the superior surface of the temporal and zygomatic origins of the zygomatic arch to simulate the zygomatic arch reconstruction on damaged skulls. These wire measurements were used to fit a simple linear regression model between the bizygomatic widths and the wire measurements, and the estimated regression model; Bizygomatic Width (bone) = 0.61 + 1.02(wire measurement), has a very high R2 value of 0.91. Hence, this model could effectively be used to predict bizygomatic widths based on wire measurements, In addition, the bizygomatic widths and wire measurements were collected from 14 New Zealand European skulls to test the ability of the regression model to determine bizygomatic widths from different ethnic groups. This model accurately predicted the bizygomatic widths in New Zealands of European origin skulls and therefore suggests that this regression model could be used for other ethnic groups. The importance of the bizygomatic width for craniometric analysis makes this regression model particularly useful for analysing archaeological samples. Furthermore, this regression line can be used in the field of forensic facial reconstruction to reconstruct damaged zygomatic arches prior to facial reconstructions. Copyright (C) 2009 John Wiley & Sons, Ltd.