110830 research outputs found
Sort by
Theory of Mind and Language Development
No full textThis Elements presents a series of studies investigating the relationship between language, Theory of Mind, and other cognitive skills, across different languages. The first set of studies focuses on longitudinal relationships between English-speaking children’s understanding of complement-clause constructions, mental verbs, modal verbs, and Theory of Mind. The second set of studies investigates links between complement-clause constructions, mental verbs, and Theory of Mind in Mandarin Chinese and English. The last study looks at English- and Turkish-speaking children’s knowledge of evidentiality, source-monitoring, and Theory of Mind. The studies show that there are different linguistic tools across languages that enable children to represent and acquire Theory of Mind
Dark Energy Survey Year 3 results : w CDM cosmology from simulation-based inference with persistent homology on the sphere
No full textWe present cosmological constraints from Dark Energy Survey Year 3 (DES Y3) weak lensing data using persistent homology, a topological data analysis technique that tracks how features like clusters and voids evolve across density thresholds. For the first time, we apply spherical persistent homology to galaxy survey data through the algorithm TopoS2, which is optimized for curved-sky analyses and healpix compatibility. Employing a simulation-based inference framework with the Gower Street simulation suite – specifically designed to mimic DES Y3 data properties – we extract topological summary statistics from convergence maps across multiple smoothing scales and redshift bins. After neural network compression of these statistics, we estimate the likelihood function and validate our analysis against baryonic feedback effects, finding minimal biases (under ) in the plane. Assuming the wCold Dark Matter model, our combined Betti numbers and second moments analysis yields and – constraints 70 per cent tighter than those from cosmic shear two-point statistics in the same parameter plane. Our results demonstrate that topological methods provide a powerful and robust framework for extracting cosmological information, with our spherical methodology readily applicable to upcoming Stage IV wide-field galaxy surveys
Development of DEM models for analysis of multi-type single lap joints : A microscale approach
Get PDFCurrently, the concept of lightweighting is a prominent area of research to decrease the energy consumption and emissions. One of the most effective strategies for achieving lightweighting is to utilize advanced lightweight materials in place of traditional materials. Meanwhile, there are no fabrication processes capable of producing the entire structure as a monolithic unit without joints. The majority of structures are constructed by assembling different types of structural elements. Some traditional joining methods, such as bolting and riveting, can introduce high stress concentrations around the joint area and may cause cracks during the drilling process. In contrast, adhesive bonding offers notable advantages for joining multi-material structures, including reduced life-cycle maintenance costs and weight, more uniform stress distribution, and enhanced design flexibility. Microstructures such as microstructural surface roughness and internal defects of constituents are crucial factors in determining the performance and fracture mechanism of adhesive joints. However, the research dedicated to the examination of the failure mechanisms for adhesive joints influenced by microstructures at microscale is limited. This work conducts systematic experimental and numerical investigations into the effect of microstructural roughness and defects on the performance and fracture mechanism of multi-type adhesive SLJs. The adherend materials used in this study are Al and PPA, bonded with an epoxy adhesive (Loctite EA 9497). Firstly, the mechanical properties of the Al adherend, PPA adherend, epoxy adhesive, and multi-type SLJs (Al-Al SLJ, PPA-PPA SLJ, and hybrid SLJ) with three roughness grades are obtained through experimental studies. The mechanical properties of adherends and adhesive are used to determine the microparameters of the contact model of adherends and adhesive particles in the DEM model. The calibrated microparameters are validated by several experimental data, especially the interlaminar-like properties of the thin adhesive layer in joints. Then, the microstructural roughness and microstructural defect in SLJs are experimentally investigated through SEM and microCT scanning. The measured microstructural roughness and microstructural defect are realistically introduced into the DEM models for further calibration, including DEM Al-Al SLJ models, PPA-PPA SLJ models, and hybrid SLJ models. Compared to the experimental results, the developed DEM SLJ models can predict the performance and capture the microstructural fracture mechanisms of multi-type SLJs. Finally, the effects of microstructural roughness and microstructural defects on the performance of multi-type SLJs are investigated, including the failure load and stiffness. The effects of microstructural roughness and microstructural defects on the microscale fracture mechanisms of multi-type SLJs are also explored and discussed, including the crack initiation, coalescence, and propagation within the adhesive and interface
Energy sufficiency, space temperature and public policy
Get PDFReductions in absolute levels of energy demand in the Global North are increasingly understood to be important for timely net zero transitions. Accordingly, there is growing interest in sufficiency as a basis for policy. This paper explores how indoor air temperatures, as an aspect of heating and cooling demand, might be addressed. Drawing on prior reviews, it argues that sufficiency as policymaking can be distinguished from more common interpretations of sufficiency as voluntary individual-level self-moderation or post-growth socio-economic transformation or even the direct imposition of legal consumption limits. Policies could instead be oriented towards the ‘framework conditions’ that shape social practices. Moreover, common ways of articulating temperature objectives, such as limits or averages, do not reflect the distributional concern that is distinctive of a sufficiency approach. By integrating insights from social practices literature, the paper outlines how a staged thermal energy sufficiency strategy might proceed through a combination of broad guideline temperature ranges and ‘shift and improve’ objectives for lower-energy practice configurations. Working towards integrated policy packages across health, housing, welfare, energy and climate policies, this strategy would aim to create the conditions for wider debate and, over time, change within thermal norms and standard
Joint time-to-event partial order continual reassessment method and Joint time-to-event Bayesian logistic regression model : Statistical designs for dual agent phase I/II dose finding studies with late-onset toxicity and activity outcomes
No full textDual agent dose-finding trials study the effect of a combination of more than one agent, where the objective is to find the Maximum Tolerated Dose Combination, the combination of doses of the two agents that is associated with a pre-specified risk of being unsafe. In a Phase I/II setting, the objective is to find a dose combination that is both safe and active, the Optimal Biological Dose, that optimises a criterion based on both safety and activity. Since Oncology treatments are typically given over multiple cycles, both the safety and activity outcome can be considered as late-onset, potentially occurring in the later cycles of treatment. This work proposes two model-based designs for dual-agent dose finding studies with late-onset activity and late-onset toxicity outcomes, the Joint time-to-event (TITE) partial order continual reassessment method and the Joint TITE Bayesian logistic regression model. Their performance is compared alongside a model-assisted comparator in a comprehensive simulation study motivated by a real trial example, with an extension to consider alternative sized dosing grids. It is found that both model-based methods outperform the model-assisted design. Whilst on average the two model-based designs are comparable, this comparability is not consistent across scenarios
Towards Ethical and Responsible AI : Perspectives from Postcolonial Contexts
Get PDFThis thesis critically explores the ethical challenges and opportunities that arise from artificial intelligence (AI) development in postcolonial contexts, explicitly focusing on West Africa. Although the global conversation surrounding AI ethics has expanded significantly, the development of AI continues to be shaped mainly by Euro-American perspectives. These perspectives frequently neglect the intricate historical, structural, and epistemic factors that influence the adoption and integration of technology within postcolonial contexts. This oversight raises important questions about the inclusivity and comprehensiveness of ethical considerations in AI development and deployment, especially in regions where colonial legacies continue to influence technological frameworks and societal structures. Through a thesis-by-publication model, this research advances a decolonial understanding of AI ethics by combining a systematic literature review, an empirical investigation, and the development of a normative conceptual framework. The first paper presents a systematic review of 50 peer-reviewed studies on ethical AI in postcolonial contexts. Employing postcolonial theory, it identifies three interrelated dynamics: structural dependency, algorithmic colonialism, and epistemic erasure. It shows how global AI ethics frameworks marginalise local knowledge while reinforcing infrastructural and political-economic asymmetries. This review underscores the urgent need for contextually responsive and pluriversal approaches to AI governance, which can better address AI development's unique challenges and opportunities in postcolonial contexts. The second paper draws on 45 semi-structured interviews and one focus group with AI developers in Nigeria and Ghana to examine how ethical challenges are experienced in practice. The findings reveal patterns of structural domination, digital and data colonisation, technological mimicry, and limited representation. These dynamics highlight how AI development reproduces historical forms of dependency, labour exploitation, and epistemic marginalisation, while also surfacing forms of resistance and aspirations for decolonial alternatives. The third paper introduces the EquiAI Framework, a decolonial and contextually responsive model for ethical AI development. Being grounded in postcolonial and decolonial theory, and informed by empirical insights, the framework advances four core constructs: coloniality of power and knowledge, algorithmic colonialism, epistemic disobedience, and relational ethics. It articulates the principles of inclusivity, intersectionality, and epistemic justice, supported by structural pillars such as participatory governance, data sovereignty, and adaptive implementation. The EquiAI Framework offers theoretical and practical guidance for equitable AI development in postcolonial societies. These three studies contribute substantially to information systems scholarship by reframing AI ethics beyond universalist paradigms and centring the political economy, lived realities, and epistemic diversity of postcolonial contexts. The thesis contends that developing equitable AI is not merely an objective but a moral obligation. It calls for dismantling structural dependencies, a resistance to algorithmic colonialism, and the adoption of decolonial, participatory, and contextually relevant governance models that resonate with the values and knowledge of local communities
Treatment-control comparisons in platform trials including non-concurrent controls
No full textShared controls in platform trials comprise concurrent and non-concurrent controls. For a given experimental arm, non-concurrent controls refer to data from patients allocated to the control arm before the arm enters the trial. The use of non-concurrent controls in the analysis is attractive because it may increase the trial’s power of testing treatment differences while decreasing the sample size. However, since arms are added sequentially in the trial, randomization occurs at different times, which can introduce bias in the estimates due to time trends. In this article, we present methods to incorporate non-concurrent control data in treatment-control comparisons allowing for time trends. We focus on frequentist approaches that model the time trend and Bayesian strategies that limit the borrowing level depending on the heterogeneity between concurrent and non-concurrent controls. We examine the impact of time trends, overlap between experimental treatment arms, and entry times of arms in the trial on the operating characteristics of treatment effect estimators for each method under different patterns for the time trends. We argue under which conditions the methods lead to Type 1 error control and discuss the gain in power compared to trials only using concurrent controls by means of a simulation study in which methods are compared. Supplementary materials for this article are available online
Physical health care and staff training : An exploration of the experiences of mental health nurses, managers of mental health nurses and trainers
Get PDFBackground: It is known that people with serious mental illness have higher mortality and morbidity rates compared with the general population and the reasons for this are multifaceted. Mental health nurses are in a pivotal position to address physical health inequalities experienced by this population group, but research indicates that they may lack the skills, knowledge, and confidence to provide patient centred and integrated care. It is important to understand the views and experiences of mental health nurses and other key stakeholders regarding physical health care provided to this population group and training to improve physical health care. An initial systematic review with narrative synthesis explored the views and perceptions of key stakeholders (mental health nurses, managers of mental health nurses and trainers) in relation to physical health care training needs of mental health nurses and the effectiveness of training. The results from the review informed the research questions and qualitative design of this empirical study Aim: The aim of the study is to explore the views and experiences of mental health nurses and other key stakeholders about (a) the care given to people with serious mental illness who have physical healthcare needs in acute in-patient mental health settings(b) the training provided to mental health nurses to enable them to provide care to people with serious mental illness who have physical healthcare needs. Methods: The design of the empirical research is a qualitative interview-based study using reflexive thematic analysis (Braun & Clarke 2019) underpinned by hermeneutical-phenomenological epistemology. In total, eighteen in depth interviews were conducted with inpatient mental health staff (nine mental health nurses, four managers of mental health nurses and five trainers in physical health care) from a single NHS Foundation Trust in England. The analysis of qualitative data was informed by Braun & Clarke (2019) reflexive thematic analysis. Findings: Four themes were developed under the following headings: (1) Culture of care and sense of competence: “Oh yeah, they asked me to do blood sugar. I told them I'm not there for the physical health, I’m there for the mental health” (2) Promoting professional autonomy: “Training went to waste” (3) The influence of mental state: “They tell you that, “I don’t need it; I don’t think anything is wrong” (4) Training success: what does it mean to me? Conclusion: As an original contribution to knowledge, this thesis concludes with discussion which highlights that adequate pre-registration education and post registration training, existence of supportive conditions and culture of care that equally prioritises mental and physical health care shape the delivery of optimal physical health care
Disease burden attributable to intimate partner violence against females and sexual violence against children in 204 countries and territories, 1990–2023 : a systematic analysis for the Global Burden of Disease Study 2023
No full textBackground Violence against women and against children are human rights violations with lasting harms to survivors and societies at large. Intimate partner violence (IPV) and sexual violence against children (SVAC) are two major forms of such abuse. Despite their wide-reaching effects on individual and community health, these risk factors have not been adequately prioritised as key drivers of global health burden. Comprehensive x§and reliable estimates of the comparative health burden of IPV and SVAC are urgently needed to inform investments in prevention and support for survivors at both national and global levels. Methods We estimated the prevalence and attributable burden of IPV among females and SVAC among males and females for 204 countries and territories, by age and sex, from 1990 to 2023, as part of the Global Burden of Diseases, Injuries, and Risk Factors Study 2023. We searched several global databases for data on self-reported exposure to IPV and SVAC and undertook a systematic review to identify the health outcomes associated with each of these risk factors. We modelled IPV and SVAC prevalence using spatiotemporal Gaussian process regression, applying data adjustments to account for measurement heterogeneity. We employed burden-of-proof methodology to estimate relative risks for outcomes associated with IPV and SVAC. These estimates informed the calculation of population attributable fractions, which were then used to quantify disability-adjusted life-years (DALYs) attributable to each risk factor. Findings Globally, in 2023, we estimated that 608 million (95% uncertainty interval 518–724) females aged 15 years and older had ever been exposed to IPV, and 1·01 billion (0·764–1·48) individuals aged 15 years and older had experienced sexual violence during childhood. 18·5 million (8·74–30·0) DALYs were attributed to IPV among females and 32·2 million (16·4–52·5) DALYs were attributed to SVAC among males and females in 2023. IPV and SVAC were among the top contributors to the global disease burden in 2023, particularly among females aged 15–49 years, ranking as the fourth and fifth leading risk factors, respectively, for DALYs in this group. Among the eight health outcomes found to be associated with IPV, anxiety disorders and major depressive disorder were the leading causes of IPV-attributed DALYs, accounting for 5·43 million (–1·25 to 14·6) and 3·96 million (1·71 to 6·92) DALYs in 2023, respectively. SVAC was associated with 14 health outcomes, including mental health disorder, substance use disorder, and chronic and infectious disease outcomes. Self-harm and schizophrenia were the leading causes of SVAC-attributed burden, with SVAC accounting for 6·71 million (2·00 to 12·7) DALYs due to self-harm and 4·15 million (–1·92 to 13·1) DALYs due to schizophrenia in 2023. Interpretation IPV and SVAC are substantial contributors to global health burden, and their health consequences span a variety of individual health outcomes. Importantly, mental health disorders account for the greatest share of disease burden among survivors. Investing in prevention of these avoidable risk factors has the potential to avert millions of DALYs and considerable premature mortality each year. Our findings represent strong evidence for global and national leaders to elevate IPV and SVAC among public health priorities. Sustained investments are needed to prevent IPV and SVAC and to implement interventions focused on supporting the complex social and health needs of survivors. Funding Gates Foundation