Evaluating the Validity and Reliability of Textile and Paper Fracture Characteristics in Forensic Comparative Analysis
Date of Graduation
Eberly College of Arts and Sciences
Forensic and Investigative Science
In a comparative forensic analysis, an examiner can report that a physical fit exists between two torn or separated items when they realign in a manner unlikely to be replicated. Due to the common belief that it is unlikely that two unrelated fractured objects would match with distinctive characteristics, a physical fit represents the highest degree of association between two items. Nonetheless, despite the probative value that this evidence could have to a trier of fact, few studies have demonstrated such assumptions' scientific validity and reliability. Moreover, there is a lack of consensus-based standard protocols for physical fit comparisons, making it difficult to demonstrate the basis for the features that constitute a “fit.” Since these analyses rely entirely on human judgment, they are highly subjective, which could be problematic in the absence of harmonized examination and interpretation criteria protocols.
As a result, organizations like the National Institute of Justice and NIST-OSAC have identified the need for developing standardized methods and assessing potential error sources in this field. This research aims to address these gaps as applied to physical fits of textiles and paper. Here, standard criteria and prominent features for each material are defined to conduct physical fit examinations in a more reproducible manner. Additionally, a quantitative metric is used to quantify what constitutes a physical fit when conducting comparative analyses of textiles and paper, further increasing the validity and reliability of this methodology and providing a manner of assessing the weight of this evidence when presented in the courtroom.
The first aim of this research involved the development of an objective and systematic method of quantifying the similarity between fractured textile samples. This was done by identifying relevant macroscopic and microscopic characteristics in the comparative analysis of a fractured textile dataset. Additionally, factors that affect the suitability of certain types of textiles for physical fit analysis were evaluated. Finally, the systematic score metric was implemented to quantify and document the quality of a physical fit and estimate error rates.
The second objective of this study consisted of establishing the scientific foundations of individuality concerning the orientation of microfibers in fractured paper edges. In comparative analysis of paper, it is assumed that the microfibers deposited across the surface of paper are randomly oriented, a key feature for addressing the individuality of paper physical fits. However, this hypothesis has not been tested. This research evaluated the rarity and occurrence of microfiber alignments on fractured documents. It also quantified the comparative features of scissor-cut and hand-torn paper and the respective performance rates.
Finally, the comparative analysis of textile and paper physical fits was validated through ground truth datasets and inter-examiner and intra-examiner variability studies. A ground truth blind dataset of known fits and known non-fits was created for 700 textile samples with various fiber types, weave patterns, and separation methods. Also, a set of 260 paper items, including 100 stamps and 160 office paper samples, were examined. The paper specimens contained handwritten or printed entries on two paper types and were separated by scissor-cut or hand-torn methods.
This proposed research provides the criminal justice system with a valuable body of knowledge and a more objective and methodical assessment of the evidential value of physical fits of textiles, paper, and postage stamps.
Andrews, Zachary Bailey, "Evaluating the Validity and Reliability of Textile and Paper Fracture Characteristics in Forensic Comparative Analysis" (2022). Graduate Theses, Dissertations, and Problem Reports. 11373.