27716 research outputs found
Sort by
Prosody and Pragmatics: How preadolescents with ASD take turns-at-talk during a cooperative goal-oriented task
Prosodic impairment in autism is widespread. It greatly affects the ability to interact, form relationships, and convey meaning. Yet, despite its prevalence, our understanding of its nature remains incomplete. This study investigates the prosody of six preadolescent children with Autism Spectrum Disorder when playing a cooperative video game with an adult. Utilising Conversation Analysis (CA), it examines twelve hours of data to explore how these children use and respond to prosodic devices for interactional purposes; specifically to manage conversational turns, create information boundaries, signal importance and express emotion. The study also examines the adult’s role in accommodating and interpreting the children’s prosody. Key findings suggest these children demonstrate competence in managing turns when signalling completion and competing for speakership using prosodic resources aligned with lexico-syntactic structure. They display understanding and competency in using prosody to create meaningful sentences as well as to highlight novel or important information. Moreover, these children produce canonical complaints-relevant designs to express frustration. However, the study reveals how these children have challenges with the articulatory and prosodic timing and coordination of their speech, leading to irregular rhythm and fragmented and dysfluent turns, particularly when using complex linguistic structures. The children appear aware of these difficulties but have problems in modifying and updating prosodic patterns. They tend to resort to less complex lexico-syntactic structures to facilitate progression in the talk. The adult accommodates these prosodic variations, intervening only when shared understanding breaks down. Overall, the children show competence and understanding of prosodic function. However, errors in the timing and coordination of phonetic and acoustic features, alongside linguistic limitations impact prosodic form, limiting communication effectiveness. These findings expand our understanding of the prosodic abilities of children with ASD. The study highlights the need for targeted interventions to address these prosodic challenges and improve communication outcomes for these children
Data Selection Methods for Semi-Supervised Learning in Automatic Speech Recognition
For an automatic speech recognition (ASR) system to perform well on new target data, it must be trained on data from the same domain as the target. Semi-supervised learning, which integrates both manually transcribed and untranscribed data, can be beneficial in this context. However, manually transcribed data are often unavailable. This challenge can be addressed by generating a training dataset from a data pool whenever new target data are introduced.
By employing the methods proposed in this thesis, training data can be effectively selected from a multi-domain untranscribed data pool, leading to improved ASR performance. The findings of this research demonstrate the potential for leveraging large amounts of untranscribed data from multiple sources to enhance ASR systems
Three essays on corporate climate activism of non-financial firms
This PhD thesis investigates the climate activism of non-financial firms. It focuses on UK and US firms to examine whether and how climate activism influences firm-level substantive environmental performance measures and financial outcomes. This thesis comprises three standalone essays. Taken together, it offers new insights into the climate activism of non-financial firms that are exposed to direct and indirect carbon emissions. Below is a summary of each essay.
The first essay examines the effect of climate risk governance and green transformation initiatives on carbon mitigation performance. Based on a unique dataset of compliant, non-financial carbon-sensitive, UK-listed firms, this essay provides evidence that climate risk governance mechanisms and green transformation initiatives enhance actual carbon performance by lowering actual carbon emissions. Further analysis shows that the combined effect of climate risk governance and green transformation initiatives helps attain sustained carbon mitigation performance. To put it plainly, the combined effect of both drivers indicates the quality of institutional transformation that is responsible for reducing carbon emissions. This study has significant implications for board members, CEOs, regulators, and other stakeholders regarding reforms in governance mechanisms and sustainable transformation.
Utilizing the non-financial components of the S&P 500 index, the second essay provides novel empirical evidence that climate risk governance is associated with environmental innovation. We find that firms with resilient climate risk governance appear to engage more in eco-driven innovative approaches to lower environmental costs and the burden on customers. This essay also demonstrates the channel through which climate risk governance supports environmental innovation. The empirical outcomes of this study contribute to the global debate on eco-friendly innovations that encourage sustainable business avenues and help limit the adverse impact of climate change.
The third essay looks into the effects of firm-level actual environmental investment, a fundamental aspect of tangible environmental engagement and risk management, on firm financing decisions and the cost of debt in US firms. Reusing the sample employed in the second essay, this study shows that firms that spend on environmental initiatives tend to boost their use of financial leverage. We also find evidence that a higher level of environmental investment is negatively related to the cost of debt. However, this negative association between environmental investment and the cost of debt financing is more pronounced for small firms as opposed to their larger counterparts. The empirical outcomes of this essay provide noteworthy insights for socially responsible investors, executive management, and regulators regarding firm-level substantive environmental engagements and financing decisions
Real-Time Detection and Classification of Motile Microbes Using Machine Learning and Holographic Microscopy
The tracks of motile pathogens can be used as a phenotypic `footprint' that can be used to predict the species of the pathogen. This has potential uses across many industries including healthcare, environment and manufacturing. Digital holographic microscopy (DHM) is a unique, high-throughput tool for obtaining the 3D time-series tracks of microbes for study and analysis. Here I show that the process of analysis DHM can be sped up to real-time speeds so that tracks can be obtained rapidly, and that these tracks can be used to predict the species using machine learning techniques. I also discuss how this work could be applied to the industries mentioned above
QUANTIFICATION OF COLLAGEN STRUCTURE IN OSTEOGENESIS IMPERFECTA USING SECOND HARMONIC GENERATION ON POLYCAPROLACTONE SCAFFOLDS
Othered from the offset: decolonial narrative analysis of female Indian university student experiences in the North of England
Attending Western universities remains popular, with international students making up a fifth of UK university income. India is a top three country of origin for international students in the UK, with 126,535 Indian students studying in 2021-2022. This educational connection highlights the ongoing relationship between Britain and India post Indian independence. An important consideration is how the British Empire, and colonial legacies have informed Indian international student experiences in UK universities. Despite turbulent political and cultural times, universities remain sites of belonging, and due to increased political and historical awareness, international students' sense of belonging is an important consideration for UK higher education. I ground this project in the decolonial principles of having an ethos of power sharing and the challenging of colonial and Western positions of power; these have helped me to operationalise my decolonial orientation. This thesis follows the narratives of three Indian women, exploring their experiences as international students and their study-related and free time communities. Miguel Zavala's (2016) decolonial strategies of counter/storytelling, healing and reclaiming informed an in-depth qualitative approach.
Two data generation periods of reciprocal interviews were carried out during and after the Covid-19 pandemic. I interpreted, from the women’s narratives, their in- betweenness. In-betweenness conceptually exists as experiencing a lack of belonging, leading to indecision. I contribute to the development of an additional decolonial strategy to help participants work through their in-betweenness which connects past colonial implications and present experiences to prepare for future self-decolonisation. The key messages illustrated conflicting feelings about being labelled as international students and the value of a strong national identity. This shaped the women’s understanding and experiences of their study-related and free time communities. Recommendations include the acknowledgement of the colonial legacies within British universities and better collaboration between students, staff, and management to decolonise the international student experience
UNDERSTANDING POSTCOLONIAL PENAL REFORM IN BANGLADESH SPECIAL LEGAL REGIME IN CONTEXT
This research involves a socio-legal analysis of postcolonial special penal reform in Bangladesh. In spite of robust presence of constitutional commitment to the rule of law ideals, the shadow of colonial hangover continues to haunt the justice institutions in this post-colonial state. In addition to its colonial legacies, some postcolonial statutes offer harsh sentences and stringent procedures in investigation, trial and punishment of many vaguely defined new forms of crime. In practice, such sporadic postcolonial reforms have turned out to be either inappropriate or inadequate because the persons at helm of the state have the tendency to arbitrarily manipulate the course of law to buttress their power with an eye on controlling their adversaries.
In order to fathom the dynamics of an offshoot of criminal justice system, this thesis is premised upon some philosophical and practical debates surrounding some highly invoked expressions including rule of law, criminalisation, and postcolonial legal politics.
The key query investigated in this research is how and why a particular postcolonial penal trend is systematically problematic. The data sources used to answer this question include interviews with 15 key informants and court observations supplemented by an auto-ethnography.
The thesis argues that an eclectic complex of law, justice and society has emerged in a volatile state that tends to move between the high ideals of the rule of law and oxymoronic rule by law. It finds that the special legal regime per se is not at adds with the rule of law; rather, empirically it forms a host of problematic menifestations of erosion of rule of law as special laws leave wider scope for abuse. It also presents a sui generis case of methodological barriers undertaken in socio-legal contexts that posed formidable challenges to a researcher
Cost reduction and benchmarking practice in complex infrastructure decommissioning projects
Though cost reduction is a key duty of public infrastructure project practitioners to ensure taxpayers get more for less, the literature has been drawn away from this subject. Cost performance has been seen as synonymous with reducing cost overruns that stem largely from projects being insufficiently planned. Better front-end definition of these projects would increase their estimated costs and thus make it more likely for the actual cost to match. However, though this improves cost performance on the face of it, it can fall short of delivering value for money for the taxpayer. It neglects reduction of cost during the planning of projects through optimisation of their design and delivery, something that frees up funding for other important projects and is thus of considerable value to society. This is even more crucial for the larger, more complex projects like nuclear decommissioning projects/programmes (NDPs), where their cost is higher and thus the value of cost reductions greater. However, the infrastructure projects realm generally lacks understanding of the key areas for cost reduction opportunity: the cost drivers. Moreover, the cost reduction opportunities themselves are not clearly defined, with researchers tending towards quantitative methods but doing little to present project teams with implementable cost reduction solutions. Benchmarking is a tool that can be used to identify how the best-in-class projects reduce their cost, but it is not yet known if complex project teams actually perform benchmarking effectively. This could justify underperforming NDPs as compared to similar but higher performing complex project industries like offshore oil & gas decommissioning projects (OGDPs). Therefore, this thesis aimed to investigate the cost drivers of infrastructure projects and the cost reduction and benchmarking practices of NDP and UK OGDP practitioners.
Three stages of research were conducted to meet this aim. In the first stage, an integrative literature review of the cost drivers of infrastructure projects was conducted. It presents a novel definition, a taxonomy, and conceptual framework of cost drivers. The conceptual framework highlights those from the taxonomy that are controllable, i.e., able to have their cost reduced. This formed the basis of the second stage of research, where a single-case study of a UK NDP was conducted to determine how cost is reduced in relation to the cost drivers. To uncover this in more detail, this stage utilised a practice theory lens to understand the actuality of cost reduction. It provides seven practice-based cost reduction solutions that extend previous knowledge by offering insights both into what can be done to reduce cost in the planning of complex projects and also how these solutions are actually developed and implemented. The third stage of research uses practice theory again to understand the actuality of benchmarking in complex projects. Semi-structured interviews were conducted with project practitioners from UK NDPs, Continental Europe NDPs, and UK OGDPs, to find that the actuality of benchmarking – presented in the form of a novel five-domain qualitative and quantitative benchmarking framework – looks considerably different to the guidance. It shows that benchmarking needs to be adapted to better support complex projects teams in the identification of performance improvement solutions
Texture Evolution and Its Relationship with Strain Path of Mg-Ca-Zn Lean Alloys
The project was based on magnesium-calcium-zinc (Mg-Ca-Zn) alloys, a topical study the area of Mg alloys. There has been extensive research, leading to many explanations for the mechanism of texture weakening in the Mg-Ca-Zn alloys. However, the explanations of the texture weakening remain controversial. Different materials and different processing routes (extrusion and rolling) lead to different results. Different processing routes cause different strain paths in the materials. The role of the strain path on the texture weakening mechanism is unknown and is therefore important to investigate. The aim of this project was to discover the relationship between the different strain paths and the mechanism of the weakening effect in Mg-Zn-Ca alloy. The relationship between strain/strain rate and mechanism of the weakening effect was also investigated. While the role of the Ca and Zn in weakening the texture of the Mg alloys is clearer, the relationship between the Ca/Zn ratio and the texture weakening in Mg-Ca-Zn alloy still needs further investigation and is also the subject of this work