University of Florida, Gainesville, FL
Tulane University, New Orleans, Louisiana
(Received 20 August 1998; accepted 15 January 1999)
A discriminant function analysis based on seven postcranial measurements for the metric assessment of race is presented. A sample from the Terry Collection (NMNH) was used to create independent functions for African-American males and females, and European-American males and females. The functions were tested using known forensic cases from the Maxwell Museum of Anthropology and the C.A. Pound Human Identification Laboratory.
Based on the Terry Collection sample, correct classification of race for males was 87.0%, and for females 100.0%. For the independent test population, correct classification for males was 81.8%, and for females only 57.1%. The low classification for females is most likely due to sample bias.
Paper ID: JFS12018J