Establishing Metric Method’s to Identify Variation in Human Biological Sex Using the Upper and Lower Limb Bones of Non-Population Specific Samples.

Author

Roland P. Sanchez, University of MontanaFollow

Year of Award

2025

Document Type

Thesis

Degree Type

Master of Arts (MA)

Degree Name

Anthropology (Forensic and Biological Anthropology)

Department or School/College

Department of Anthropology

Committee Chair

Dr. Randy Skelton

Committee Co-chair

Dr. Meradeth Snow

Commitee Members

Dr. Randy Skelton, Dr. Meradeth Snow, Dr. Mark Heirigs

Keywords

Osteology, Metrics, Sex Estimation

Subject Categories

Biological and Physical Anthropology

Abstract

Sex estimation of human skeletal remains is a facet of the biological profile. The pelvis has been suggested as the best source to accurately predicting the biological sex, but not always available after the recovery of human remains. At times, only long bones of human skeletal remains are recovered for identification. It has been shown that metric measurements using long can accurately predict biological sex. The most common methods use discriminant functions to predict the likelihood an individual falls within a certain population. These studies rely on population specific collections of skeletal remains for increased accuracy. It’s advised to avoid using these methods on different populations and when producing discriminant functions in the absence of a populationspecific collection. Because of this limitation, it’s difficult to identify biological sex using non-destructive means. This research aims to find evidence that support the use of nonpopulation specific methods to accurately predict biological sex of human long bones. Discriminant analysis was used as a test of variance using a sample of 2920 individuals from the University of Tennessee Forensic Databank and the Goldman Dataset of Postcranial Measurements. Discriminant analysis consistently yielded 70-90% classification results. A test of comparison was performed against discriminant analysis using logistic regression yielding 80-90% classification results. Finding evidence that metric methods can predict some degree of sexual dimorphism without the need for population affinity

Recommended Citation

Sanchez, Roland P., "Establishing Metric Method’s to Identify Variation in Human Biological Sex Using the Upper and Lower Limb Bones of Non-Population Specific Samples." (2025). Graduate Student Theses, Dissertations, & Professional Papers. 12447.
https://scholarworks.umt.edu/etd/12447