Estimation of sex from metatarsals using discriminant function and logistic regression analyses

M. A. Bidmos*, A. A. Adebesin, P. Mazengenya, O. I. Olateju, O. Adegboye

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

16 Citations (Scopus)


South Africa has one of the highest crime rates in the world and the discovery of dismembered bodies for human identification process poses a greater challenge. The South African Africans (also known as South African blacks) population group is often the victims of crimes as they are the largest group. While measurements of several bones of the human skeleton have been used for sex estimation, the potential of metatarsals have not been explored in this population group. Metatarsal bones are usually well-preserved since they are recovered in shoes protected from scavengers and they are able to withstand environmental degradation and taphonomy. This study investigated the potential of measurements of metatarsals in sex estimate amongst South African Africans using logistic and discriminant function analysis. Six measurements of metatarsals from 100 individuals of known sex and population affinity from the Raymond Dart Collection of Human skeletons were analysed. Various combinations of measurements of metatarsal bones yielded suitably high average accuracies (79% to 84%) for sex estimation and are comparable to functions derived from other skeletal elements of South African Africans. Metatarsals of South African Africans are therefore useful as alternatives to highly sexual dimorphic bones in the forensic estimation of sex.

Original languageEnglish
Pages (from-to)543-556
Number of pages14
JournalAustralian Journal of Forensic Sciences
Issue number5
Publication statusPublished - 2021


  • Estimation of sex
  • South Africa
  • discriminant function analysis
  • forensic anthropology
  • logistic regression analysis
  • metatarsals


Dive into the research topics of 'Estimation of sex from metatarsals using discriminant function and logistic regression analyses'. Together they form a unique fingerprint.

Cite this