Compared to other violent crimes, sexual assaults suffer from very low prosecution and conviction rates, especially in the absence of DNA evidence. As a result, the forensic community needs to use other forms of trace contact evidence, such as lubricant evidence, in order to provide a link between the victim and the assailant. In the current study, the instrumental data were analyzed by multivariate statistics, including hierarchal cluster analysis, principal component analysis, and linear discriminant analysis. By interpreting the mass spectral data with multivariate statistics, 12 discrete groupings of lubricants were identified, indicating inherent chemical diversity not only between but within the three main marketing groups. A number of unique chemical markers, both major and minor, were identified, other than the three main chemical components (i.e. PEG, PDMS and nonoxyno 9) currently used for lubricant classification. The data were validated by a stratified 20 percent withheld cross validation, which demonstrated that there was minimal overlap between the groupings. Based on the groupings identified and unique features of each group, a highly discriminating statistical model was then developed that aims to provide the foundation for the development of a forensic lubricant database that may eventually be applied to casework. (Publisher abstract modified)
Downloads
Similar Publications
- The Role of Simulated Data in Making the Best Predictions (from the 87th Annual Meeting of the American Association of Physical Anthropologists - 2018)
- Human Decomposition Ecology at the University of Tennessee Anthropology Research Facility
- Genetic Architecture of Skin and Eye Color in an African-European Admixed Population