During the CSAFE All Hands Meeting, the opportunity to listen to many researchers discuss their current research projects was presented. The projects ranged from statistical foundations to several kinds of forensic evidence, such as firearms and tool marks, bloodstain pattern analysis, and fingerprint analysis. Two of these projects will be discussed based on their posters and presentations.
Poster 1:
The first study that will be discussed is entitled “Building a Shoe Sole Database Using Longitudinal Data Collection”, which was conducted at Iowa State University. The person presenting the poster was James E. Kruse, and his co-authors were Dr. Guillermo Basulto-Elias, and Dr. Alicia Carriquiry.
This particular research project was developed to determine whether or not an algorithm could be created to identify shoes. This study focused on the longitudinal data collected from two styles of shoes, and two sizes for each men and women. 160 volunteers wore the shoes that were assigned to them over a period of six months, with at least 10,000 steps per week1. These shoes were then brought into the lab every five weeks for data collection. Six categories of data were collected, including HR pressure mats, high resolution photography, 2D scanning, 3D scanning, film-dust prints, and paper-dust/crime scene prints. Each category of data was replicated, or performed twice, to ensure accuracy between the measurements. Researchers are ISU compiled this data into a database that will later be made public so statistical methods can be used to analyze this data and attempt to create an algorithm.
In the future, more styles and sizes of shoes will be studied to be included in the database. One statistical concern in the algorithm will be the variability of each shoe print after the minimum number of steps and time. Statistical methods will be used to determine how similar or different a shoe print is depending on how many steps have been walked in the shoe, as well as the characteristics of the person wearing them, such as their height and weight. Another future extension of this research will include prints that could possibly be made at crime scenes, such as prints in mud, blood, dry dirt, etc.
This project was fascinating because it included a lot of information that beginners do not normally think about when it comes to shoe print analysis. Every person walks and wears their shoes differently, so it was interesting to see how this study takes many factors into account when examining shoes.
Poster 2:
The second poster that will be discussed is entitled “Analysis of Forensic Evidence and Wrongful Convictions”, conducted at the University of California, Irvine. The poster was presented by Dr. Simon A. Cole, and his co-authors were Alyse Bertenthal and Matt Barno.
This poster focused on two main research questions: What types of forensic evidence are the most frequent contributors to wrongful convictions, and what problems with forensic evidence appear in these cases that can be addressed by statistics2. In order to answer these questions, the research team looked at the National Registry of Exonerations, which has information about cases in which someone was found to be wrongly convicted in the past. The cases that the research team wanted to find in this registry fit the criteria of a person who was wrongfully convicted, and later cleared of all charges based on new evidence that proved a person’s innocence. Of the 2182 cases that were studied, 524 of those cases were found to have false or misleading forensic evidence used in their cases. The two major types of forensic evidence that were seen to have more frequent mistakes were drug chemistry, with a frequency of 119 cases, and serology, with a frequency of 108 cases. The disciplines that CSAFE studies, which are footwear, friction ridge, questioned documents, bloodstain patterns, digital, and fire arms evidence had a frequency of 61.
This research was interesting because it was different from what the Innocence Project focuses on. The Innocence Project focuses on DNA evidence, while the National Registry of Exonerations looks at all types of forensic evidence. One interesting fact that was included in this poster was a mention of how Harris County, Texas had made many mistakes of wrongfully convicting people based on field presumptive testing for drugs. These tests were later found to be inaccurate by intensive laboratory testing and return many false positives. Dr. Cole discussed that Harris County realized their mistakes and made a public apology for their actions. Most other counties try to hide when mistakes are made, and clean things up quietly without making a public apology.
The future direction of this project will dive deeper into these cases of the wrongfully convicted, looking at what specific mistakes are being made in each area of forensic evidence, and using statistics to understand why these mistakes are being made, and how to decrease the frequency of these mistakes.
References
Using Longitudinal Data Collection. Poster session presented at the CSAFE Annual All Hands Meeting, Ames, IA.
Convictions. Poster session presented at the CSAFE Annual All Hands Meeting, Ames, IA.