The goal of this project was to develop and evaluate free, online Open-Educational Resource (OER) content for a forensic chemistry course, aligned with the Course-based-Undergraduate Research Experience (CURE) model of instruction. Through this project, we created supplementary materials including case studies, forensic chemistry lab techniques, and demo videos. The course also provided a platform for the students to interact with forensic experts. Nine high impact practices were incorporated and aligned with CURE. The materials covered in the course are hosted on the LibGuides platform for equitable access. While student perceptions of OER material quality varied, they were strongly correlated with student experience. Additional attitudinal results suggest student self-efficacy significantly increased in response to participation in this OER forensic chemistry CURE. Instructors who wish to adopt or adapt these course materials should provide structured support for digital tools so all students find them accessible, maximizing the effectiveness of this OER CURE.
Wrongful convictions represent a persistent yet unevenly documented problem across Asian criminal justice systems. While international research has identified common factors contributing to known wrongful convictions, systematic information about wrongful conviction cases in Asia remains fragmented and difficult to compare across jurisdictions. This fragmentation limits cross-jurisdictional learning and comparison of documented cases. This paper synthesizes existing research on contributing factors in documented wrongful conviction cases in Asia and reviews current initiatives addressing such cases. It then showcases an operational map for the Asian Registry of Wrongful Convictions (ARWC), an academic documentation platform designed to record final criminal convictions that were later officially corrected or acknowledged as erroneous. By setting out the registry's scope and finality requirement, its inclusion criteria and how they were developed, its partial harmonization strategy, documentation workflow, and feasibility, the paper offers a concrete next step toward improving transparency and comparative understanding of wrongful convictions in Asia.
Cyberbullying that escalates into a victim's suicide poses significant challenges at the intersection of criminal law, causation analysis, and digital forensics. This article aims to examine how four jurisdictions - Kazakhstan, the United States, Germany, and India - conceptualise and assign criminal liability for suspected abetment to suicide conducted through online harassment, and to identify the evidentiary thresholds that shape prosecutorial decision-making. Methods comprise comparative doctrinal analysis of legislation and published judgments, supplemented by structured document analysis of official materials and digital-evidence standards. The document was the unit of analysis. The source corpus included statutes and amendments, reported judicial decisions, official prosecutorial and institutional materials, and standards or guidance. Sources were compared across legal basis, prohibited conduct, mens rea, causation, victim autonomy, procedural status, and the evidentiary functions of digital records. Results show substantial divergence: Kazakhstan expressly criminalises online forms of driving, inducing, or assisting suicide; India retains a statutory abetment offence but applies a demanding test of intent and proximity; the United States relies on state-specific offences or manslaughter; Germany generally precludes homicide attribution where suicide reflects autonomous decision-making. The reviewed authorities show that attributed communications, call records, and complete timelines can bear on authorship, knowledge, sequence, autonomy, and causation, but their legal effect depends on the applicable offence. Conclusions emphasise that rigorous digital-evidence procedures and harmonised investigative standards are essential for fair, reproducible adjudication of cyber-induced suicides, while preserving due-process and free-expression safeguards. Practically, legislators should define prohibited conduct, culpability, causal proximity, vulnerability, and autonomy safeguards; investigators should document source, lawful authority, acquisition, timestamps, integrity checks, analysis, and every transfer so that digital evidence can be assessed across borders.
Forensic testimony is where science meets the courtroom, and it is one of the critical forensic skills taught largely by apprenticeship. An examiner who overstates a match, mishandles a probability, or cannot support a method under direct or cross-examination can weaken the evidence in precisely the cases that rely on it. Yet in most forensic laboratories, testimony preparation remains largely informal, mentor-dependent, and unscalable. Consequently, many practitioners first face rigorous cross-examination at trial. To address this gap, HelixCross, an alpha-stage, publicly accessible, prompt-engineered large language model (LLM) agent, was developed. It uses retrieval-augmented generation (RAG) to simulate defense counsel cross-examination and provide structured coaching on demand. Deployed on two LLM platforms to widen access, it pairs a phase-based curriculum with a synthetic DNA case, a scripted cross-plan with embedded fallacy traps, and a ten-dimension coaching rubric. The initial curriculum targets recurrent problems in DNA testimony, including likelihood-ratio interpretation, proposition discipline, activity-level overreach, probabilistic-genotyping assumptions, and unwarranted certainty. This paper describes the architecture, training assets, and governance considerations. Evaluation to date consists of informal functional testing by the developers and a few early-stage users, without predefined acceptance criteria, independent ratings, or learner outcome data. Testing also exposed platform-dependency risks, including session-state failure after a vendor-side model update. HelixCross is therefore presented as a transparent development-stage tool and curriculum package intended to enhance the resilience, clarity, accuracy, and scientific integrity of forensic expert testimony - a contribution that a follow up phase of controlled evaluation is required to demonstrate.
The Indiana Fingerprint Comparison Database is an open-source resource consisting of 513 image triplets of fingerprint images. Each triplet was assembled by a subject matter expert and consists of a latent impression, a mated exemplar livescan, and a non-mated exemplar livescan. The images have known ground truth and reflect many of the attributes of casework. A subset of 90 image triplets was tested in a black box study using 104 latent print examiners, and we provide response frequencies for each image pair using a 5-point response scale. We use an ordered probit model [1] to summarize the response distributions as a way of characterizing the strength of support for the same source proposition for each image pair and create likelihood ratios. The black box performance data and the freely-available images described here will support future research studies, laboratory training, validation, internal proficiency testing and calibration exercises.
Incense trees (Aquilaria sinensis) are illegally harvested across Asia due to the high market value of the agarwood they produce. This problem is particularly prevalent in Hong Kong, where government agencies are actively focused on reducing illegal harvesting by finding new means of deterrence and actively prosecuting criminal suspects. We conducted the first case study on incense trees in Hong Kong, where plant materials from criminal suspects' clothing were matched to an illegally harvested source tree using Sanger and whole genome resequencing approaches. We identified several plant particles collected from clothing of suspects to be from incense trees and traced them back to a specific illegally harvested tree. This evidence led to the prosecution of nine suspects who were sentenced to 30-50 months in jail under the Organized and Serious Crimes Ordinance (OSCO) in Hong Kong. This was one of the first times OSCO was used in illegal wildlife trade prosecution in Hong Kong, leading to 25% longer sentences. The result demonstrates that even small, aged and poorly preserved wood particulates can yield enough genomic DNA for whole genome analyses to be used as forensic evidence to support the prosecution of criminal suspects.
To illustrate the prevalence of wrongful convictions and their causes, we report critical analyses of international archival databases documenting known exonerations in the USA, England and Wales, and the Netherlands. We suggest that unjust convictions should be viewed as system failures in both adversarial and inquisitorial systems. A common theme that weaves throughout the article is that there are (often procedural) differences across the jurisdictions, yet there are many similarities in recognised wrongful conviction cases. Perhaps most striking is that not only are wrongful convictions a global problem, but that many wrongful convictions share common features regardless of the underlying jurisdiction. We identify these features, which include, for example, false confession, mistaken eyewitness identification and official misconduct, and which frequently co-exist. We also examine the relationship between other variables of interest (e.g., race, crime type, DNA) and wrongful convictions in each jurisdiction. The comparative approach allows for varying jurisdictions to learn from each other in order to ultimately enhance the efforts of researchers, policy makers, and criminal justice practitioners in transforming their systems. The transformations should in large measure target formulating future policies and reforms. Aside from the differences, the similarities across databases further indicate that these findings are robust and provide an important first step in comparing databases across varying jurisdictions and developing common reforms.
To maximise DNA yield from evidence and ensure the integrity of the chain of custody, all biological evidence must be properly managed and preserved during initial crime scene processing. This scoping review aimed to collate documented protocols and issues associated with processing blood, semen, saliva, and touch DNA at the crime scene to optimise downstream DNA processing. A synthesis review of 57 peer-reviewed studies (2015-2025) was conducted using a curated table of forensic DNA evidence. Data were presented in a table organised by study features, protocols, problems, solutions, results, and suggestions. Findings from the studies referenced favour optimised double-swab (wet + dry) or adhesive tape-lifting techniques with buffered solutions (e.g., PBS, SDS) and visualisation aids (e.g., Diamond Nucleic Acid Dye); immediate air-drying or active-drying systems; chemical preservatives (Longmire's buffer, ethanol, solid NaCl); cold storage; and carrier additives. Strict anti-contamination procedures (PPE, elimination databases, negative controls) are essential, as are separate tamper-evident packaging. These protocols improve DNA recovery, reduce degradation and stochastic effects, and enhance success in STR and mtDNA profiling. Key challenges include low-template stochasticity, shedder variability, substrate interference, environmental degradation (especially aquatic and thermal), operator variability, and secondary transfer. Most studies have been conducted in laboratories, which may not be generalisable to real-world settings. Rapid, substrate-tailored collection with minimal handling, immediate preservation, and a strict chain of custody constitute the current best practice for evidence collection at the crime scene. Future studies should focus on multi-site field validations, automation, and harmonised guidelines to address current translational gaps and enhance equity in forensic contexts.
Students in Higher Education (HE) are increasingly engaging with Generative Artificial Intelligence (GenAI), raising concerns regarding learning, academic integrity, and equitable access to technology. In forensic science, these concerns are particularly significant, as professional practice relies on critical judgement, evidentiary reliability and ethical accountability. This study examined how undergraduate forensic science students engaged with a university-approved GenAI tool (Copilot) within crime scene simulations, with a focus on decision-making, ethical and responsible use, and trust. Copilot was embedded into a second-year subject (Major Scene Investigation) of a three-year undergraduate forensic science degree, as a structured decision-support tool. Students critically evaluated AI-generated outputs during simulated crime scene investigations and documented their use through reflective entries, including interaction records and reflections that could be presented in various formats (e.g. written, audio, video). These data were analysed thematically to explore student perceptions, patterns of engagement, and the development of critical AI literacy. Findings indicate that students did not adopt GenAI uncritically or as a substitute for professional judgement. Instead, they demonstrated a cautious selective approach, positioning GenAI as a conditional support tool. Engagement evolved over time, with initial scepticism giving way to more informed strategic use. Importantly, students differentiated between low- and high-stakes decisions, demonstrating reduced reliance on GenAI where accountability and evidentiary consequences were greater. Embedding GenAI within assessments supported the development of critical and ethical GenAI literacy, enabling students to articulate appropriate, cautious, and limited uses of the technology. While GenAI offered efficiencies in tasks such as hazard identification and documentation, its role remained bounded by the need for human oversight and justification. This study demonstrates the value of integrating GenAI into discipline specific learning environments to support critical engagement and understanding rather than passive use. It highlights the importance of explicitly teaching the ethical and professional boundaries of GenAI in forensic science decision-making to prepare students for responsible use of this technology in future practice.
Explosive detection canine (EDC) teams are widely used to find explosives and protect public safety, but training practices and standards vary across agencies/organizations. This study investigated how EDC handlers train their dogs, their familiarity with national standards, and what challenges they face in preparing for effective operation. Data from a nationwide survey of handlers (102 unique responses, 62 complete), across 29 states, and in-depth interviews were conducted during two field evaluations of the ANSI/ASB Standard 092 Standard for Training and Certification of Canine Detection of Explosives. Results revealed most handlers train regularly and value certification; however, training structures differ widely. Many handlers reported limited familiarity with national standards and scientific research, and some lacked access to formal training guides, realistic scenarios, or certain explosive materials needed for effective practice. Handlers emphasized the importance of practical training that reflects real-world conditions, including complex environments, low target prevalence, and presence of distractors. They also reported challenges related to limited access to true training materials, uncertainty about how odors change or how canines generalize between like odors, time and resource constraints, and differences between testing formats and field work. Concurrently, many viewed structured evaluations as helpful for identifying weaknesses and improving performance. Overall, the findings suggest successful detection depends on multiple factors working together: consistent standards, access to resources, realistic training, scientific guidance, and a strong handler-canine partnership. Improving communication between researchers, standards organizations, and practitioners, as well as increasing outreach and access to evidence-based information, may help strengthen training, reduce errors, and improve operational readiness.
Forensic expert certification relies on evaluation systems designed to assess professional performance under high technical, analytical, and probative demands. However, certification outcomes are often interpreted through aggregate indicators, such as approval rates or overall mean scores, which may obscure where performance is stable, sensitive, or insufficient within the internal structure of expert work. This study proposes a Performance Criticality Index to examine certification evidence at the level of competency units in a national institutional forensic certification system in Colombia. The index combines two dimensions that are central to certification decisions: the distance between observed performance and a predefined sufficiency threshold, and the variability of performance across evaluation cycles. Using aggregated historical results from forensic certification processes conducted between 2021 and 2025, the analysis identifies differentiated configurations of performance rather than global levels of achievement. The findings show that criticality is not uniformly distributed across disciplines or competency units. Robust competencies coexist with localized sensitivity, distributed sensitivity, concentrated criticality, and structural insufficiency. Criticality tends to emerge in units involving integrative forensic reasoning, analytical reconstruction, or the formulation of expert conclusions, whereas highly standardized technical-procedural components tend to show greater stability. The index does not measure expertise directly; rather, it identifies competency units in which certified performance may be structurally vulnerable because mean scores approach the sufficiency threshold while dispersion remains high. The proposed approach contributes to interpreting certification evidence beyond averages and to understanding how evidence, technique, and expert judgment are articulated in forensic practice.
Sally Clark's 1999 conviction for the murder of her two infant children represented one of the most conspicuous miscarriages of justice in modern English criminal law. A central component of the prosecution case was statistical evidence presented by Professor Sir Roy Meadow, which suggested that the probability of two sudden infant deaths occurring naturally within the same family was approximately one in 73 million. This paper examines three methodological limitations that affect the evidential value of that estimate. First, the family characteristics used to reduce the estimated probability of natural death were not considered in relation to the probability of the alternative explanation, thereby distorting comparative inference. Second, sudden infant death syndrome is a diagnosis of exclusion; population frequency data derived from such a category cannot determine causation in an individual case and are contingent on the limits of postmortem investigation. Third, the probability estimate was presented without reference to the likelihood of competing explanations, limiting its interpretive value. The paper further argues that the significance of the Clark case lies not only in the arithmetical error in the one in 73 million figure but in the broader evidential process by which probabilistic evidence was admitted, framed and interpreted within a criminal trial. The paper also examines the empirical and unstructured approach of English criminal courts to expert evidence. It considers the extent to which subsequent developments, including the Criminal Practice Directions 2023, respond to these difficulties, and compares that position with the contrasting procedural framework associated with the Daubert standard in the United States. The analysis situates the Clark case within broader issues concerning the interpretation, contextualisation and scrutiny of expert evidence in legal proceedings.
Wrongful convictions represent a profound failure of criminal justice systems with severe individual and societal consequences such as, undermining public trust in the justice system, allowing true perpetrators to remain free, and inflicting lasting harm on exonerees and their families. Despite growing international attention to this issue, systematic empirical data on wrongful convictions in Europe remains limited. This article presents a descriptive analysis of the 144 exoneration cases documented in the European Registry of Exonerations (EUREX) as of May 2026. The dataset includes exoneration cases from across Europe, with Germany, Italy, the Netherlands, Sweden, and Spain most represented. Findings show that the majority of exonerees were male, with most initial convictions relating to homicide. On average, individuals spent around 6 years and 5 months wrongfully imprisoned, collectively representing 927 lost years, while the average time from conviction to exoneration was nearly 12 years. False confessions, and false accusation or perjury emerged as the leading contributing factors, frequently occurring in combination with other factors. DNA evidence played a role in only a minority of exonerations These findings are consistent with North American findings and underscore the need for structured case review mechanisms and continued cross-national research into the causes and prevention of wrongful convictions.
Melissopalynology can provide scientific evidence about the authenticity of honey products, which is becoming a medical, economic and legal concern globally. Investigations can evaluate whether the pollen and spores (collectively termed palynomorphs) in a questioned honey sample are consistent with its purported geographical origin and botanical source(s). Legislation can therefore use such melissopalynological information to create benchmarks for authentic honey products and, by requiring explicit labelling of origins on honey products, can counteract honey fraud, protect consumers, and create premium economic products for small-scale honey production. If the authenticity of a honey product is challenged, melissopalynology must be able to provide a robust forensic service. This review evaluates how well melissopalynology qualifies as a forensic science that can be used to litigate under the legislation with which it is intimately associated. It meets criteria of scientific testability; professional peer review and publication; codified standard operating procedures for data collection, quantification and analysis; quantified error rates; and widespread acceptance within its scientific community, and it has been used in court cases. Its public databases are growing and proprietary databases need to be probatively auditable. Like other forensic sciences, melissopalynology is growing in technical sophistication but its ability to address specific relevant questions may depend on the contingencies of each case, e.g. the inadmissibility of illegally-obtained evidence. Melissopalynology is a science that satisfies the baselines of fitness for court, and this overview highlights areas where it can be refined so that its evidence is not merely admissible and legally defensible, but inherently robust forensic science.
