The goal of this research was to test the discrimination ability of standard and non-standard craniometric measurements and data derived from geometric morphmetric (GM) analysis.
Standard cranial measurements are commonly used for ancestry estimation; however, 3D digitizers have made cranial landmark data collection and geometric morphometric (GM) analyses more popular within forensic anthropology. Yet there has been little focus on which data type works best. The current study used 31 cranial landmarks to generate 465 interlandmark distances, including a subset of 20 commonly used measurements, and to generate principal component scores from procrustes coordinates. All were subjected to discriminant function analysis to ascertain which type of data performed best for ancestry estimation of American Black, White, and Hispanic males and females. The non-standard interlandmark distances generated the highest classification rates for females (90.5 percent) and males (88.2 percent). Using non-standard interlandmark distances over more commonly used measurements leads to better ancestry estimates for the current U.S. population structure. (Publisher abstract modified)