This report describes a research study to establish a large dataset of dental data in order to improve estimations of sex and ancestry or population affinity in forensic anthropology.
The authors of this biological anthropology report describe a research study that had five main objectives: collect data on standard and novel morphological traits and metric dental data in modern populations across the globe, to identify their utility in sex and ancestry/population affinity estimation; create a large databank of dental data that will be publicly available for research; make available a free web-based computer application to provide statistical estimates for use in forensic anthropological casework; explore methods to estimate sex using dental data from modern collections; and provide a free database for the collection of dental morphological and metric data. Results demonstrated that every metric variable in the researchers’ dataset for every tooth under investigation was found to be significantly different between males; the percent dimorphism was positive in every case, indicating that males were larger; cervical measurements of the teeth exhibited much higher levels of sexual dimorphism than those of the crown; cervical measurements of the upper and lower canine showed the highest levels of dimorphism, however the mesiodistal crown measurement of the lower canine was in the same range of dimorphism. The authors report that in a simple discriminant function analysis to estimate sex using only those five variables, the cross-validated model was able to correctly estimate sex in 81.0% of cases. They conclude that the preliminary analyses indicate discriminatory power using the dentition to estimate sex, particularly when the canine teeth can be used.
Downloads
No download available
Related Datasets
Similar Publications
- Mental Health and Rape History in Relation to Non-Medical Use of Prescription Drugs in a National Sample of Women
- Genetic mapping of 15 human X chromosomal forensic short tandem repeat (STR) loci through multi-core parallelization
- An efficient clustering-based retrieval framework for real crime scene footwear marks