Birmingham City University Open Access Repository
Not a member yet
8823 research outputs found
Sort by
Juror Certainty about Firearms Evidence: Examination Effects
No full textFirearms examiners routinely compare tool-marks on suspect ammunition with those on ammunition test-fired by a suspect weapon to evaluate if they can be associated. The discipline has been subject to criticism, including by the National Academy of Sciences, but the testimony of firearms examiners is routinely admitted as expert evidence in the United States (US). Jurors must determine the weight of expert evidence, which opposing and proffering lawyers can aim to, respectively, discredit on cross-examination and rehabilitate on redirect-examination. The authors investigated the effect of both cross and redirect- examination on potential US jurors’ certainty about expert firearms evidence using a series of
online vignettes. Participants (n=114) were asked to rate their certainty (on a scale of 0-100) about three expert statements – Very Certain (an exact match), Certain (a match to a reasonable degree of certainty), and Uncertain (evidence is unsuitable for comparison) – when assigned to one of three conditions. These conditions were a judicial instruction about weighing the evidence (control condition); a cross-examination referencing criticism of firearms evidence; and a redirect-examination (following the cross-examination) referencing the routine admission of firearms evidence. Analysis was undertaken both between groups and between the statements given to each group. Results suggest that experts conveying high certainty create higher certainty in jurors, cross-examination has a detrimental effect on this certainty, but redirect-examination does not reduce this detrimental effect
Outlier-oriented poisoning attack: a grey-box approach to disturb decision boundaries by perturbing outliers in multiclass learning
No full textPoisoning attacks are a primary threat to machine learning (ML) models, aiming to compromise their performance and reliability by manipulating training datasets. This paper introduces a novel attack—outlier-oriented poisoning (OOP) attack, which manipulates labels of most distanced samples from the decision boundaries. To ascertain the severity of the OOP attack for different degrees (5–25%) of poisoning to conduct a detailed analysis, we analyzed variance, accuracy, precision, recall, f1-score, and false positive rate for chosen ML models. Benchmarking the OOP attack, we have analyzed key characteristics of multiclass machine learning algorithms and their sensitivity to poisoning attacks. Our analysis helps understand behaviour of multiclass models against data poisoning attacks and contributes to effective mitigation against such attacks. Utilizing three publicly available datasets: IRIS, MNIST, and ISIC, our analysis shows that KNN and GNB are the most affected algorithms with a decrease in accuracy of 22.81% and 56.07% for IRIS dataset with 15% poisoning. Whereas, for same poisoning level and dataset, Decision Trees and Random Forest are the most resilient algorithms with the least accuracy disruption (12.28% and 17.52%). We have also analyzed the correlation between number of dataset classes and the performance degradation of models. Our analysis highlighted that number of classes are inversely proportional to the performance degradation, specifically the decrease in accuracy of the models, which is normalized with increasing number of classes. Further, our analysis identified that imbalanced dataset distribution can aggravate the impact of poisoning for machine learning models
Vision‐Based UAV Detection and Tracking Using Deep Learning and Kalman Filter
No full textABSTRACT
The rapid increase in unmanned aerial vehicles (UAVs) usage across various sectors has heightened the need for robust detection and tracking systems due to safety and security concerns. Traditional methods like radar and acoustic sensors face limitations in noisy environments, underscoring the necessity for advanced solutions such as deep learning‐based detection and tracking. Hence, this article proposes a two‐stage platform designed to address these challenges by detecting, classifying, and tracking various consumer‐grade UAVs. The tracking efficacy of the proposed system is assessed using a combination of deep learning and Kalman filter techniques. Specifically, we evaluate models such as YOLOv3, YOLOv4, YOLOv5, and YOLOx to identify the most efficient detector for the initial detection stage. Moreover, we employ both the Kalman filter and the Extended Kalman filter for the tracking stage, enhancing the system's robustness and enabling real‐time tracking capabilities. To train our detector, we construct a dataset comprising approximately 10,000 records that capture the diverse environmental and behavioural conditions experienced by UAVs during their flight. We then present both visual and analytical results to assess and compare the performance of our detector and tracker. Our proposed system effectively mitigates cumulative detection errors across consecutive video frames and enhances the accuracy of the target's bounding boxes
Passive heating in sport: context-specific benefits, detriments, and considerations
No full textExercise and passive heating share some acute physiological responses. These include increases in body temperature, sweat rate, blood flow, heart rate, and redistribution of plasma and blood volume. These responses can vary depending on the heating modality or dose (e.g., temperature, duration, body coverage) and are beneficial to athletes in specific scenarios. These scenarios include being applied to increase muscle or force production, induce rapid weight loss, stimulate thermoregulatory or cardiovascular adaptation, or to accelerate recovery. The rationale being to tailor the specific passive heating protocol to target the desired physiological response. However, some acute responses to passive heating may also be detrimental to sporting outcomes, such as exercising in the heat, having unintended residual negative effects on performance or perceptions of fatigue, or even resulting in hospitalisation if implemented inappropriately. Accordingly, the effects of passive heating should be carefully considered prior to implementation by athletes, coaches, and support staff. Therefore, the purpose of this review is to evaluate the physiological responses to different modes and doses of passive heating and explore the various sport contexts where these effects may either benefit or hinder athletes. Understanding these responses can aid the implementation of passive heating in sport and identify potential recommended heating protocols in each given scenario
Perceived Critical Success Factors for Implementing Building Information Modelling in Construction Small- and Medium-Sized Enterprises
No full textBuilding information modelling (BIM) is an emerging technology in the building sector. As with any emerging technology, the identification of critical success factors (CSFs) for BIM is essential. On the other hand, small- and medium-sized enterprises (SMEs) consistently play a vital role in the construction industry. Therefore, it is essential to determine the critical success elements for the effective implementation of BIM in these companies. Hence, this study aims to determine the CSFs for implementing BIM in SMEs in the developing country of Iran. To accomplish this, three rounds of the Delphi technique were carried out with the participation of fifteen BIM professionals from SMEs based in Iran. According to the Delphi survey findings, a total of 27 CSFs were identified for the effective utilisation of BIM in SMEs. Subsequently, to assess the CSFs, a questionnaire utilising a five-point Likert scale measurement was designed. Then, it was distributed among specialists in construction SMEs in Iran. The questionnaire included twenty-seven factors categorised into four primary groups: technical, managerial, financial, and legal. A total of 56 questionnaires were gathered and examined. The findings indicate that the CSFs highlighted for implementing BIM in SMEs are above the average level. Furthermore, the CSFs with a high impact on successful BIM implementation in construction SMEs in Iran were determined. Four high-impact CSFs are (1) the employer’s demand; (2) understanding the advantages and practicality of implementing BIM; (3) awareness of and ensuring a return on investment; and (4) efficient and suitable legislation. The findings of this study can serve as a valuable resource for stakeholders, providing them with a useful tool to enhance decision-making about the implementation of BIM in SMEs, especially in developing countries
Annual Vitamin D Status of World-Class British Swimmers Following a Standardised Supplementation Protocol for Three Years
No full textBackground/objectives: British swimmers are at a heightened risk of vitamin D deficiency (serum 25-hydroxyvitamin D (25(OH)D): 75 nmol∙L−1: no supplementation). Results: Mean serum 25(OH)D concentrations increased each year (2018: 76.4 ± 28.4 nmol∙L−1, 2019: 91.5 ± 24.8 nmol∙L−1, 2020: 115.0 ± 36.6 nmol∙L−1, p 125 nmol∙L−1, although it is currently debated whether this is a concern or a benefit for athletic populations. Conclusions: Supplementing with 4000 IU∙day−1 of vitamin D3 throughout the winter can increase the vitamin D status of swimmers. However, more frequent testing may be required to ensure that serum 25(OH)D remains within the sufficient range across the season (75–125 nmol∙L−1)
Twenty-Four-Hour Compositional Data Analysis in Healthcare: Clinical Potential and Future Directions
No full textCompositional Data Analysis (CoDA) is a powerful statistical approach for analyzing 24 h time-use data, effectively addressing the interdependence of sleep, sedentary behavior, and physical activity. Unlike traditional methods that struggle with perfect multicollinearity, CoDA handles time use as proportions of a whole, providing biologically meaningful insights into how daily activity patterns relate to health. Applications in epidemiology have linked variations in time allocation between behaviors to key health outcomes, including adiposity, cardiometabolic health, cognitive function, fitness, quality of life, glycomics, clinical psychometrics, and mental well-being. Research consistently shows that reallocating time from sedentary behavior to sleep or moderate-to-vigorous physical activity (MVPA) improves health outcomes. Importantly, CoDA reveals that optimal activity patterns vary across populations, supporting the need for personalized, context-specific recommendations rather than one-size-fits-all guidelines. By overcoming challenges in implementation and interpretation, CoDA has the potential to transform healthcare analytics and deepen our understanding of lifestyle behaviors’ impact on health
Authentic assessment in event management education: Integrating technology and industry for authentic learning
No full textPurpose
This paper explores authentic assessment in event management education, with emphasis on emerging technologies and industry collaboration. It addresses sector-wide concerns around course credibility and employability, examining how authentic assessment can bridge the gap between academic learning and professional practice.
Design/methodology/approach
An interpretivist, qualitative approach was used, drawing on focus groups and interviews with UK-based event management academics. Thematic analysis identified key themes in assessment design, technology use, and industry engagement.
Findings
Academics widely support authentic assessment for its role in developing industry-relevant, employability-focused skills. Technology was used to replicate professional contexts but was limited by institutional constraints, digital inequality, and staff capability. Industry partnerships improved assessment relevance but often relied on individual networks and goodwill. A revised version of Gulikers et al.’s (2004b) Five-Dimensional Framework is proposed, incorporating student co-creation, reflexive practice, and industry feedback as additional dimensions.
Originality
The paper offers new insights by highlighting academic perspectives and advancing existing frameworks to better reflect current practice. The proposed model integrates technology, reflexivity, and industry feedback, offering a more adaptable tool for vocational disciplines.
Research limitations/implications
The study reflects only academic perspectives within UK event management educators. Further research should include student and industry views and test the revised framework in practice.
Practical implications
To support authentic assessment, institutions must provide clearer definitions, targeted staff development, and resources to embed industry collaboration in curriculum design
Disrupting norms of time and talk: Email interviews as a neurodivergent-affirming method for ethical and rich qualitative research
No full textThis article challenges the marginal status of email interviews in qualitative research. It argues that neuro-normative assumptions underpin prevailing ideals of ‘good’ research, privileging qualities aligned with neurotypical modes of interaction. Such qualities include verbal, real-time communication and bodily co-presence. Drawing on an interpretative phenomenological analysis study with women diagnosed in adulthood with autism and attention deficit hyperactivity disorder (ADHD), this article challenges these neuro-conventions of research. It argues that email interviews enabled embodied engagement and disrupted normative temporalities by allowing participants to regulate sensory environments and engage on their own terms, in their own time. The asynchronous and written format of email interviews facilitated participant agency, supported their well-being, and enabled reflexivity for both participant and researcher. This produced rich, authentic data and enabled ethical research practices grounded in autonomy, beneficence, and justice. Reframing email interviews as a neurodivergent-affirming method, this article highlights their potential as a robust, inclusive and theoretically generative approach within qualitative research
Bacterial agents of otitis media and their antibiotic resistance profile: a retrospective clinical presentation at St. Dominic Hospital, Akwatia, Ghana
No full textBackground
Otitis media is an important public health condition, particularly in children. Globally, it is estimated that 391.3 million people were affected by otitis media in 2021. It is ranked the fifth leading global burden of disease and the second cause of hearing loss. This study was conducted to identify the bacterial agents causing otitis media and their antibiotic resistance patterns at the St. Dominic Hospital in Ghana from 2020 to 2022.
Methods
A retrospective study involving 130 archived records was conducted at St. Dominic Hospital in the Eastern Region of Ghana. Data from the laboratory records in the department from 2020 to 2022 were retrieved. Data collected were entered into a Microsoft Excel 365 file, cleaned and exported into STATA version 15 statistical software for analysis. P-value < 0.05 was considered statistically significant.
Results
Within the study period, 130 suspected cases of otitis media were investigated in the laboratory. Of these, 46.2% (60/130) recorded bacterial growth. Isolates were predominantly Gram-negatives, with Pseudomonas aeruginosa [(28.3%) 17/60] and Proteus species [(26.7%) 16/60] accounted for majority of the cases. Staphylococcus aureus [(16.7%) 10/60] and Coagulase Negative Staphylococci [(8.3%) 5/60] were the only Gram-positive organisms isolated. All the Klebsiella species and Proteus species tested against ceftriaxone and meropenem were resistant, respectively. None of the isolates tested against amikacin was resistant.
Conclusion
The study highlights varying antimicrobial resistance patterns among the S. aureus and Gram-negative isolates, emphasising the importance of prudent antibiotic use, robust antibiotic stewardship practices, and continued surveillance to combat antimicrobial resistance effectively