Paul Marshall’s research while affiliated with University of Bristol and other places
What is this page?
This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.
Social norms are standards of behaviour common in a society. However, when agents make decisions without considering how others are impacted, norms can emerge that lead to the subjugation of certain agents. We present RAWL-E, a method to create ethical norm-learning agents. RAWL-E agents operationalise maximin, a fairness principle from Rawlsian ethics, in their decision-making processes to promote ethical norms by balancing societal well-being with individual goals. We evaluate RAWL-E agents in simulated harvesting scenarios. We find that norms emerging in RAWL-E agent societies enhance social welfare, fairness, and robustness, and yield higher minimum experience compared to those that emerge in agent societies that do not implement Rawlsian ethics.
Advances in artificial intelligence (AI) have great potential to help address societal challenges that are both collective in nature and present at national or transnational scale. Pressing challenges in healthcare, finance, infrastructure and sustainability, for instance, might all be productively addressed by leveraging and amplifying AI for national-scale collective intelligence . The development and deployment of this kind of AI faces distinctive challenges, both technical and socio-technical. Here, a research strategy for mobilising inter-disciplinary research to address these challenges is detailed and some of the key issues that must be faced are outlined.
Advances in artificial intelligence (AI) have great potential to help address societal challenges that are both collective in nature and present at national or trans-national scale. Pressing challenges in healthcare, finance, infrastructure and sustainability, for instance, might all be productively addressed by leveraging and amplifying AI for national-scale collective intelligence. The development and deployment of this kind of AI faces distinctive challenges, both technical and socio-technical. Here, a research strategy for mobilising inter-disciplinary research to address these challenges is detailed and some of the key issues that must be faced are outlined.
Background
A significant percentage of the Nigerian population has type 2 diabetes (T2D), and a notable portion of these patients also live with comorbidities. Despite its increasing prevalence in Nigeria due to factors such as poor eating and exercise habits, there are insufficient reliable data on its incidence in major cities such as Port Harcourt, as well as on the influence of sociodemographic factors on current self-care and collaborative T2D care approaches using technology. This, coupled with a significant lack of context-specific digital health interventions for T2D care, is our major motivation for the study.
Objective
This study aims to (1) explore the sociodemographic profile of people with T2D and understand how it directly influences their care; (2) generate an accurate understanding of collaborative care practices, with a focus on nuances in the contextual provision of T2D care; and (3) identify opportunities for improving the adoption of digital health technologies based on the current understanding of technology use and T2D care.
Methods
We designed questionnaires aligned with the study’s objectives to obtain quantitative data, using both WhatsApp (Meta Platforms, Inc) and in-person interactions. A social media campaign aimed at reaching a hard-to-reach audience facilitated questionnaire delivery via WhatsApp, also allowing us to explore its feasibility as a data collection tool. In parallel, we distributed surveys in person. We collected 110 responses in total: 83 (75.5%) from in-person distributions and 27 (24.5%) from the WhatsApp approach. Data analysis was conducted using descriptive and inferential statistical methods on SPSS Premium (version 29; IBM Corp) and JASP (version 0.16.4; University of Amsterdam) software. This dual approach ensured comprehensive data collection and analysis for our study.
Results
Results were categorized into 3 groups to address our research objectives. We found that men with T2D were significantly older (mean 61 y), had higher household incomes, and generally held higher academic degrees compared to women (P=.03). No statistically significant relationship was found between gender and the frequency of hospital visits (P=.60) or pharmacy visits (P=.48), and cultural differences did not influence disease incidence. Regarding management approaches, 75.5% (83/110) relied on prescribed medications; 60% (66/110) on dietary modifications; and 35.5% (39/110) and 20% (22/110) on traditional medicines and spirituality, respectively. Most participants (82/110, 74.5%) were unfamiliar with diabetes care technologies, and 89.2% (98/110) of those using technology were only familiar with glucometers. Finally, participants preferred seeking health information in person (96/110, 87.3%) over digital means.
Conclusions
By identifying the influence of sociodemographic factors on diabetes care and health or information seeking behaviors, we were able to identify context-specific opportunities for enhancing the adoption of digital health technologies.
BACKGROUND
A significant percentage of the Nigerian population suffers from Type 2 Diabetes, and a notable portion of these patients also live with other co-morbidities such as hypertension and cardiovascular disease. Despite its increasing prevalence in Nigeria due to factors such as poor eating and exercise habits, there is insufficient reliable data on its incidence in major cities like Port Harcourt, as well as the influence of socio-demographic factors on current self- and collaborative T2D care approaches using technology. This, and the recurring top-down approach, i.e., from the global north to the global south of creating solutions, has resulted in a significant lack of context-specific digital health interventions for T2D care.
OBJECTIVE
The study aims to identify opportunities for designing context-specific digital health interventions for Type 2 Diabetes management, by understanding the influence of socio-demographic factors on self and collaborative care involving patients, caregivers, and community pharmacists, as well as current digital health approaches for Type 2 Diabetes care.
METHODS
We designed questionnaires aligned with the study’s objectives to obtain quantitative data, using the WhatsApp environment and in-person interactions. Following a social media campaign aimed at expanding our reach to a hard-to-reach audience, we delivered the questionnaires through WhatsApp. We also used this opportunity to explore the feasibility of WhatsApp as a data collection tool in our research context, mirroring studies that were conducted in a different context. In total, we obtained 110 responses, with 83 responses from in-person distributions and 27 responses from the WhatsApp approach. Finally, data was analyzed using descriptive and inferential statistical approaches on SPSS and JASP software.
RESULTS
Results were categorized into three, comprising socio-demographic characteristics of diabetes patients in our research context, collaborative care involving different stakeholders, and current digital health approaches for the management of Type 2 Diabetes. We discovered that men with Type 2 Diabetes were significantly older (mean = 61 years), had a higher average household income, and generally held higher academic degrees compared to women (P = .03). We also discovered that there was no statistically significant relationship between gender and the frequency of hospital (P = .597) or pharmacy visits (P = .480), and cultural differences did not influence the incidence of the disease. Regarding current management approaches, 75.45% and 60% of the population relied on prescribed medications and dietary modifications respectively, while 35.45% and 20% explored traditional medicines and spirituality respectively. Most participants were unfamiliar with technologies for diabetes care (74.54%), and 89.21% of participants who indicated their use of technology were only familiar with glucometers. Finally, participants were more inclined to seek for health information in person (87.27%) than through digital means
CONCLUSIONS
By identifying the influence of socio-demographic factors on diabetes care and health/information seeking behaviors, we were able to identify opportunities for designing context-specific digital health interventions for diabetes care.
This paper presents a systematic review of HCI literature focusing on children with ADHD, the prevailing mental health diagnosis in children. Its aim is to (i) chart the state-of-the-art in this domain (e.g. methods used), (ii) identify the ways the HCI community has addressed the needs of children with ADHD (e.g. technologies deployed), and (iii) describe the involvement of the various stakeholders playing a role in their everyday experiences (i.e. their care ecosystem). Our findings show limited engagement of the care ecosystem in the design, development and user studies of current technologies, and shortcomings in designing for multiple ecosystem stakeholders, despite their crucial role. We also find that most HCI contributions are systems aiming to address ADHD-related symptoms. Based on our findings, we provide suggestions for further research and design considerations for future systems that empower and promote the well-being of children with ADHD, while considering their care ecosystem.
... Recently, healthcare management has been undergoing a transformative evolution with the integration of self-monitoring, monitoring, and patient journey tracking into a unified system [30]. By utilizing real-time data, personalized care plans, and continuous feedback, healthcare providers can significantly enhance patient outcomes, improve operational efficiency, and deliver high-quality, cost-effective care. ...
... The management of the condition involves pharmacotherapy, which includes using hypoglycemic drugs (eg, sulfonylureas and meglitinides) and lifestyle modifications focusing on exercise and diet [36,37]. Digital interventions also play a role across different stages of T2D care, including diagnosis or prognosis with rapid test kits or machine-learning algorithms [38] and providing information, remote monitoring, and lifestyle modifications through mobile apps and websites [2,39,40]. Continuous glucose monitoring systems for blood sugar sensing and personalized medicine [41][42][43] are also critical components. Because T2D is a long-term condition that requires constant monitoring, its care involves self-management by people with the condition and collaborative care with both caregivers and health care practitioners in both formal and informal settings. ...
... This number is expected to grow [33]. Scholars have reviewed these apps [4,12,15,20], and found that these services typically include features which support users logging/charting their blood sugar readings, insulin dosages, and foods eaten. Other common features include providing users with educational tips and advice on how to manage T1D [22,36,45]. ...
... This work leverages cutting-edge privacy preserving and federated machine learning methodologies to enable machine learning on data across all the sources interactively and in real-time while guaranteeing that the identity of the individual patients and the data they provide will not be leaked through for example training-data leakage attacks (Chen & Campbell, 2022). Strand (ii) works with hard-to-reach patients (those suffering from secondary health conditions, mental health conditions, or living circumstances that prevent them from accessing health care unaided and limit their use of technology), plus their carers and clinicians, to address issues of trust, usability and efficacy in ethical AI for informing healthcare decision making across patient populations (Stawarz et al., 2023). The over-arching challenge for both strands is to leverage population-wide data collection for informing robust individualised decision-making without compromising anonymity and under realistic data and user assumptions. ...
... This gap is even more pronounced in pediatric research, where managing hypoglycemia is compounded by the unpredictability of children's routines and hormonal changes leading to insulin resistance 24 . Earlier studies present diverse methods for BG prediction, e.g., autoregressive moving-average models and various machine learning algorithms [25][26][27][28] . While recent models employing neural networks and ensemble machine learning show promise [29][30][31][32][33] , the complexity of pediatric diabetes is proving harder to model; some research aims at improvement by constraining time periods, e.g., nocturnal hypoglycemia prediction models and variable importance plot feature selection 34,35 . ...
... In fact, Merino et al. suggest that "understanding the role of MR/AR in working places and in communication and collaboration" [26] represents an important opportunity for humancentered evaluations. This can be made by running evaluation out of the laboratory in concrete scenarios of application, or in a remote manner where researchers may or may-not supervise participants [26,33,36,37]. The importance of the last has even been emphasized by the circumstances associated with the COVID-19 Pandemic, which forced research teams worldwide to adopt new methods [30,38]. ...
... As the field study was conducted within a research "in-the-wild" approach [54,55], we had to administer short questionnaires to fit into the tight training schedule of the police professionals. Thus, this questionnaire was designed for our study, drawing from existing scales on presence and technology acceptance and including uniquely crafted items to target our research objectives. ...
... Though such attempts were made, they implemented similar, one-dimensional assessment questions. And, as has been critically pointed out before, consideration of a general audience, which will inevitably involve people with different ability levels, is still in need of improvement in visualization research [3,4,40]. ...
... Human-Computer Interaction (HCI) and Child-Computer Interaction (CCI) researchers have increasingly explored technologies to support ADHD children and their care ecosystem [114]. Recent studies have highlighted various limitations in existing approaches for creating and evaluating technologies for neurodivergent individuals [109,110], particularly those with ADHD [111,114]. ...
... Steve and Yuen observe that AR applications, now portable and operable on mobile devices, have permeated various aspects of daily life. The feasibility of AR technology allows users instant access to location-specific information compiled from multiple sources [5]. This suggests that AR could effectively project current media information onto AR applications, potentially becoming one of the first widely adopted applications on any AR device. ...
