This report lays out the research design, methodology, and analytical and data analysis techniques used to develop a medico-legal algorithm that can help classify cases of strangulation, to assess the findings from post-strangulation event examinations and to use data-based techniques that will increase the scientific validity of expert witness testimonies.
This final report describes a research study that had two major goals and one research question: the first goal was to create a large, de-identified database of forensic findings from existing medico-legal strangulation and non-strangulation exams of adult women; the second goal was to use probabilistic modeling to identify injury findings or clusters of injury findings accompanying reported strangulation of women; and the research question asked whether there were documented features, or clusters of features, that were associated with cases where strangulation was reported compared to where it was not reported. The document provides details on the research design, methodology, and data analysis techniques. The authors emphasize the critical importance of further testing the algorithm they developed in order to ensure its accuracy and reliability in a court or clinical setting, however they conclude that their algorithm shows promise regarding its ability to assist with classifying cases of strangulation, given specific forensic examination characteristics. They also note that their methodology enables forensic examiners and expert witnesses to use data-based techniques to assess the findings from examinations after strangulation events, which would allow an expert witness to quantify their degree of confidence and enhance the scientific validity of their testimony.
Downloads
Similar Publications
- The Off-season of Dental Cementum Investigations. A Critical Appraisal of Season-of-death Prediction in Medico-legal Investigations
- Grooming Traffickers: Investigating the Techniques and Mechanisms for Seducing and Coercing New Traffickers
- Evaluation of Cannabis Product Mislabeling: The Development of a Unified Cannabinoid LC-MS/MS Method to Analyze E-liquids and Edible Products