Determination of sex of white femora by discriminant function analysis: forensic science applications.

Abstract

Stepwise multiple discriminant function analysis is used to establish classification functions for sex assessment of North American white femora. The functions correctly assign sex for 82% of a sample consisting of 85 femora of verified age, sex, and race, and for a similarly verified test sample of 30. The objectives are to provide criteria for sexing poorly preserved and fragmentary unknown specimens and a statement of the probable accuracy of such assessments in individual cases. The application of the method to forensic casework is illustrated by a sample case.