Search

There is an increasing public concern in China on the legions of children left behind by their parents due to massive rural to urban labour migration in the past decades. It was estimated that there were 61 million left-behind children in China in 2018. The hukou system, a household registration system established in China since 1958, has intensified institutional and social barriers between rural and urban China and created far-reaching economic and social problems, which have prompted a large number of migrants to leave their children behind in their birthplace. As a result, how to keep in touch and maintain family relations has become one of the most crucial issues for Chinese separated families due to internal migration. In such a context, social media like WeChat serve as crucial tools in helping Chinese migrants to fulfil familial duties from a distance, sustain bonds with their left-behind children, and to foster emotional links with family members at home. This thesis investigates how Chinese separated families engage with social media and for what purposes; illustrates how parents, carers (mainly grandparents) and children use social media, perceive, and interpret their roles in family lives in different ways; and explores how family relationships are shaped by social media like WeChat within Chinese separated families. It employs mixed methods to comprehend the relationship between the use of WeChat and family lives of Chinese separated families, including an online survey, semi-structured interviews, child-focused interviews and creative online workshops. The participants of this research, including parents, carers, children, and schoolteachers, are mainly recruited from a primary school in a city in Liaoning province in Northeast China. Since the maintenance of long-distance familyhood is a collective work that consists of parents, carers, and children, the research findings are presented from the perspectives of the three parties. This thesis shows that social media like WeChat has primarily been seen as parenting and surveillance tools instead of communication tools among parent-participants, who attempt to exert control over children and maintain authority from a distance. It illuminates how carers play the roles as facilitators, observers, and gatekeepers in mediating parent-child communications and relationships. It also demonstrates how the child participants, exercising their agency, use WeChat to seek parental involvement and resist the practices of parental surveillance. By bringing together the perspectives of the three parties, there are three dimensions that are associated with the quality of mediated relationships in the Chinese cultural context: 1) media policy, media access, and media literacy impact online communication, which in turn shapes mediated relationships; 2) care arrangements have been identified as a key factor that has implications for long-distance relationships; 3) the value of filial piety continues to play a pivotal role in shaping parent-child and parent-carer relationships in Chinese separated families.

Opioid-induced hyperalgesia (OIH) and tolerance (OIT) counteract opioid analgesia and drive dose escalation, thereby producing adverse effects on patients. Hyperpolarization-activated, cyclic nucleotide-gated (HCN) ion channels are constitutive membrane proteins which function as nonselective cation nonselective voltage-dependence cation channels. HCN2 ion channels, that are sensitive to cAMP, play a critical role in excitability in peripheral nociceptor neurons. Prolonged opioid administration has long been considered to result in an elevation of cAMP concentration. Therefore, the activation of HCN2 via the cAMP pathway is suggested to drive the development of OIH and possible OIT. Here we showed that inhibiting or genetically deleting peripheral HCN2 in nociceptors abolishes hyperalgesia induced by chronic morphine treatment, while inhibition of central HCN channels alleviates OIT. The expression of C-FOS, a marker of neuronal activity, in second-order neurons of the dorsal spinal cord was increased in OIH mice and was reversed by blocking or genetically deleting HCN2 whereas they were unchanged by inhibiting central HCN channels in OIT mice. Long-term morphine administration caused an increase in firing frequencies of action potential in sensory neurons as a result of the increased cAMP levels, indicating the existence of peripheral neuronal hyperexcitability. We showed that this was due to a depolarizing shift in the voltage-dependence of activation of HCN2, caused by increased expression of aberrant 6TM μ opioid receptors and a consequent Gs-driven activation of cAMP signalling pathways. These experiments support the idea that OIH is caused by enhanced neuronal excitability in the periphery, driven by a potentiated activity of HCN2 ion channels. This hypothesis may provide a new therapeutic target for the treatment of OIH.

This thesis investigates the modelling and prediction of macroeconomic and financial time series by using S-vine models and vt-S-vine models. The first contribution of this thesis is to demonstrate that non-linear and non-Gaussian models in the class of S-vine processes can fit better than autoregressive moving average (ARMA) models for certain types of macroeconomic time series, such as inflation rates. The S-vine processes generalize the class of Gaussian ARMA models by modelling dependencies using pair copulas and by modelling marginal behaviour using arbitrary continuous distributions. The second part of this thesis concerns the modelling and forecasting the volatile financial returns series by vt-S-vine models. Returns are traditionally estimated by using the generalized autoregressive conditionally heteroscedastic (GARCH) type processes or variants. We show that the vt-S-vine models can compete with GARCH processes in many cases. In order to reveal the statistical properties and structures of GARCH type processes, the vt-S-vine models are applied to replicate GARCH type processes. The best combinations of margins and pair copulas in vt-S-vines for replicating GARCH type processes are determined. The higher-order vt-S-vine processes with a sequence of mixed pair copulas can mimic GARCH type processes more precisely than first-order or second-order S-vine with t copulas as proposed by previous studies. The final part of this thesis is devoted to the application of vt-S-vines in the trading book of banks. The value-at-risk (VaR) and the VaR exceedance probabilities are estimated in vt-S-vine processes, where the best quantile estimation methods are presented in each case. Surprisingly, the quantile estimator that is closest on average to the true value of the quantile of a distribution may not be the one that yields the most accurate value for the exceedance probability. Vt-S-vines are very flexible and promising models for stationary macroeconomic, financial and banking time series, and deserve to be widely used and rapidly developed in future.

Human-animal interactions since the late Pleistocene have exerted long-lasting and significant influences on animals. With the advance of ancient DNA techniques, the availability of genomic data from ancient samples has profoundly increased, and now with great possibilities we are able to investigate the past and gain knowledge of the phased genetic changes in history. Applying predominantly ancient genomic methods, I have conducted case studies on four different species under the states of domestication, semi-domestication, and commensalism respectively to explore how the anthropogenic exercises have affected the evolutionary trajectories of animals. The first case involves Chinese domestic cats (Felis silvestris catus) by examining 4 nuclear genomes and 22 mitochondrial genomes from 22 ancient Chinese samples spanning 5,500 years. Chinese domestic cats are descended from African wild cats (Felis silvestris lybica) and were likely brought to China via the Silk Road during the 7th and 9th centuries AD. Leopard cats (Prionailurus bengalensis) might have been enjoying a commensal relationship alongside humans in the Late Neolithic before domestic cats arrived in China (6,000-5,500 BP). However, this commensal link did not survive the Han Dynasty (202 BC - 220 AD). Small felids and humans did not cohabit again until the Tang Dynasty (618-907 AD) when domestic cats were imported. Since then, domestic cats have not mixed with any Chinese native wild cat species, including the Chinese mountain cat (Felis silvestries bieti) and Asiatic wild cat (Felis silvestries ornata). The second case involves the Chinese leopard cat (Prionailurus bengalensis), who were in a commensal relationship with humans since the Chinese Neolithic period (5,500 BP). This study collected seven ancient leopard cats from the Middle Reach of the Yellow River Basin. Ancient genome study shows that the leopard cats from the Middle Reach of the Yellow River Basin and Far East Russia were closely related, albeit the former was receiving introgression with Indochina leopard cats. According to archaeological, chronological, and literary analysis, leopard cats and humans ceased their commensal connection when the Han Dynasty (202 BC - 220 AD) fell, and never recovered. Domestic cats replaced leopard cats in the human niche, when they were introduced around the Tang Dynasty (618-907 AD). The third case involves the Persian and European fallow deer (Dama dama mesopotamica and Dama dama dama). The post-glacial dispersal of the fallow deer was nearly entirely mediated by people. We generated and analysed 29 partial or mitochondrial genomes of ancient fallow deer to gain insights into their evolution and range changing. Ancient DNA analysis indicates that the two fallow deer subspecies diverged 0.57 million years ago and might have interbred in Southwestern Anatolia. An ancient maritime commercial route brought Persian fallow deer from the Levant to Cyprus and westward to the Mediterranean, while European fallow deer from Anatolia were dispersed to continental Europe. The fourth case involves the Chinese domestic pigs (Sus domesticus). The Middle Reach of the Yellow River Basin was one of the domestication hubs of pigs. We have conducted ancient genomic investigation of five nuclear and nine mitochondrial genomes of ancient domestic pigs from two archaeological sites in the Middle Reach of the Yellow River Basin spanning at least 3,000 years, and zooarchaeological analysis on 1,400 pieces faunal remains excavated from the Zheng-Han City (2,500 BP). I suggest that the populations represented by the ancient domestic pigs from the Middle Reach of the Yellow River Basin had contributed to the formation of modern Chinese breeds and issued new hypothesis of the origin of the Chinese domestic pigs. Zooarchaeological approaches showed a reduction in pigs' prominence in archaeological sites and potential intensified pig husbandry in the MYR from the late Neolithic to the Iron Age. Combining the results of the four case studies, we have addressed the importance of anthropogenic forces in altering the evolutionary trajectories of animals: 1) The human-mediated dispersal can dramatically change the range of the species, even for the animals who had lost most of their genetic diversities, accompanying the migration of human, they were still able to be spread globally. 2) By establishing affinities or cultural connections with human, the population dynamics of animals would closely correlate with the development of human societies. 3) Importing new species would lead to the admixtures between exotic animals and its related local counterparts. Such admixtures could in one hand introduce new genetic diversities, while in another hand, offset the adaptability of the native species to local environment.

Protein aggregation and amyloidosis are linked to many neurodegenerative diseases as well as rare inherited disorders. Parkinson's disease in particular, is caused by Lewy bodies - amyloid fibril aggregates of α-synuclein that induce the apoptosis of neuron cells. Although the general amyloid pathway of α-synuclein aggregation has been explored, in-depth structural and kinetic data during each step of aggregation, from monomeric protein to fibrils and propagation, is yet to be fully studied. Due to the disordered nature of α-synuclein monomers and the transient nature of oligomers, the ability to structurally model them has been limited in detail with past techniques such as NMR, CD and light scattering. Despite the current focus on oligomeric aspects of α-synuclein in therapeutics, the seeded aggregation pathway and its prion-like spreading continue to be a crucial area of study. It is evident that the aggregation pathway as a whole should be considered for protein amyloid therapeutics, as each stage is an important conduit to each other, and drug targets can be found at any stage of the aggregation process. However, the constant shift in polydispersity, structure and size during protein aggregation makes it a challenging topic to explore with existing techniques. In the case of α-synuclein, the protein molecule will go through oligomeric, fibril and seeding stages, and even different conformations in the presence or absence of lipids. There has been a gap in the area of amyloids to offer real-time structural data and kinetics in a combined fashion, as well as in-depth structural analysis for disordered monomers and transient oligomers. In this thesis, we aimed to fill the gap in research by using cutting-edge small-angle scattering techniques combined with structural modelling. Chapter 4 and 5 of this thesis will explore the monomeric and oligomeric forms of α-synuclein using SEC-SAXS from Diamond Light Source. Chapter 6 will focus on developing SAXS techniques to study the seeded aggregation pathway of α-synuclein. And finally, Chapter 7 will explore the challenging topic of vesicle induced aggregation using the contrast matching technique of small angle neutron scattering. We hope the thesis will provide the reader novel insights into the aggregation of α-synuclein and illustrate the potential usage of these scattering techniques for other protein aggregation studies.

This research explored women's leadership in China's sport national governing bodies (NGBs). The purpose of the research is to provide insight into the ways in which women's leadership can be enhanced and increased. This is the first study in recent decades to focus on women's leadership in the context of the Chinese sport industry, covering the widest range of sport NGBs and female practitioners in leadership positions within these organisations. Extensive researches have been undertaken internationally, which have identified and acknowledged the underrepresentation of women in leadership positions across multiple industries and global contexts. According to these studies, leadership specifically within the sports industry was both historically and is still currently male dominated. Within China, the underrepresentation of women in leadership positions is also evident, yet there is limited research which has examined why this is the case regarding women's leadership in China, especially within the sport industry. In response, this research gap aimed to explore the current situation of women's leadership in sport NGBs in China. By using both questionnaires and interviews, data from NGBs in China were collected. This research achieved the following three goals: First, the research determined the low representation level of women in leadership position in China's sport NGBs. Second, it identified six major barriers hindering women from achieving leadership positions in China sport NGBs. Third, it provided options for supporting women practitioners, and increasing their representation, in China's sport NGBs. This research yielded four sets of contributions. The methodological contribution of this research included the application of the theoretical framework of the barrier matrix, quantitative and qualitative research methods. The empirical contribution of this research was to answer the three key research questions through quantitative and qualitative research results. It further summarised thirteen key findings to explain the low representation of women's leadership in China's sport NGBs, and revealed how barriers were formed and perpetuated by comparing the results of primary data analysis with other literature. The theoretical contribution of this research was to develop a new "barrier matrix" model based on Burton's (2017) multi-level framework in light of the primary data analysis and discussion. It located six major barriers among three levels. Macro level included barriers of "male dominated culture" and "gender serotyping". Meso level included barriers of "organisational policies and procedures" and "gendered organisational culture". Micro level included barriers of "work-family conflicts" and "self-limiting behaviours". The barrier matrix model also provided insights concerning the vertical correlations between barriers. These insights are helpful to understand the formation, evolution and influencing factors of barriers, so as to understand why barriers are difficult to weaken. The practical contribution provided targeted suggestions based on positioning and analysis of barriers through three aspects of guidance, namely social guidance, policy guidance and personal guidance.

Energy management of distributed energy resources (DERs) is challenging due to the distributed and uncertain nature of DERs. To optimally operate DERs and trade their energy as well as energy flexibility for financial benefits, energy management for virtual power plants (VPPs) and electric vehicle (EV) charging stations are investigated in this thesis. The research in this thesis can be summarized into three parts. Part I provides a VPP operation strategy in the electricity market environment. Part II proposes an EV charging station operation strategy considering EV user incentives. Part III develops a coordinated VPP and EV charging station operation framework based on the methods proposed in parts I and II. (1) Economic VPP operation In this part, an optimal VPP operation regime is proposed considering multiple electricity markets and multiple uncertainties. The proposed operation regime handles both the VPP market bidding and unit dispatching problems. To deal with uncertainties, a hybrid stochastic minimax regret optimization model is proposed. To reduce the conservativeness of the formulated optimization models, a self-adaptive algorithm is proposed. (2) Economic EV charging station operation In this part, an EV charging station operation strategy with an EV user incentive program is proposed to improve the EV charging station economic benefit. To maximize the long-term profit of the EV charging station, an optimal incentive price selection model is developed. In the solution methodology, a problem linearization method is first proposed. Then, a distributed solution methodology is developed based on the proposed adaptive alternating-direction-methodof-multipliers algorithm. (3) Economic VPP operation considering EV charging stations i In this part, a multi-stakeholder VPP-charging station system is investigated. Firstly, a coordinated operation framework is proposed for the VPP-charging station system to maximize the total benefit of the system. Then, an improved EV user incentive program is proposed for acquiring EV energy flexibility. At the cost allocation stage, a τ -value cost allocation method is developed. To alleviate the computation burden in calculating the τ -values, a τ -values estimation approach is proposed. The effectiveness of the energy management methods proposed in this thesis is verified through theoretical analysis and numerical simulations. Significant results suggest high potential for practical application in certain scenarios.

For years, overseas Taiwanese and Tibetan exiles have faced constant identity difficulties as a result of their contested citizenship. Citizenship is a complex status containing socio-legal meanings. It may denote a membership between people and the state and "the right to have other rights (Arendt, 1961)." Even though citizenship is crucial for people to access rights, it can also be fluid and unstable, especially for those related to contested sovereignty states. This socio-legal research analyses the contested citizenship and the predicaments faced by those without full citizenship, by focusing on the cases of the Tibetan exiles in Taiwan and the overseas Taiwanese. Taiwanese people face denial of citizenship randomly when travelling across national borders, while Tibetan exiles confront confusion about citizenship during the process of migration to Taiwan. The unstable and liminal character of citizenship may cause statelessness due to the geopolitical power between China and Taiwan and the conflicts of self-identification among these ethnic groups. To show how individuals experience and react to the debates on citizenship, I conducted semi-structured interviews with people who struggled with the uncertainty of citizenship in the process of travel, migration and settlement; the stories depict people's agency and the legal mobilisation through which they exercise citizenship acts countering the challenges of liminal citizenship in social movements and daily lives. Current research fails to reflect on the issue of statelessness in the context of the long-standing instability and tensions connected to Taiwan's international standing, the controversies and conflicts within Taiwan itself, and the relevant concrete true-life experiences of actual stateless persons. My research fills this gap by highlighting the productive character of statelessness for people who are therefore able to eschew an undesired legal identity and to maintain the possibility of regaining the desired one. Thus, with the fluid citizenship status, statelessness may be used not just as an equivalent to lack of citizenship, but as an expression of people that is cultivated from the lack of full citizenship for demanding different identities. For a better understanding to the complexity of citizenship, I apply Nancy Fraser's theories of justice to examine recognition, redistribution and representation of citizenship, and then discuss how these three dimensions are constructed in today's democratic regime to dynamically formulate citizenship not only by the legal system or bureaucracy which presents state power, but also by the people.

This thesis investigates the advancement of Acoustic emission (AE) monitoring techniques and analysis methods that could be applied in the field of Structural Health Monitoring (SHM), with an emphasis on the use of advanced AE techniques to detect and locate damage and their application on the monitoring of corrosion and Stress corrosion cracking (SCC) in complex metallic structures. The work was divided into three main areas of research, which are an investigation into using FE (Finite Element) generated delta-T mapping to locate experimental AE signals on complex structures, source location of AE signals generated by corrosion and an experimental investigation into SCC damage locations on a metallic specimen.

In this thesis, I develop a threefold response to the evidential argument from evil. I first raise a new version of skeptical theism, data-based skeptical theism, to defend theism from the evidential argument from evil, but it works only under certain conditions. Data-based skeptical theism needs to be supplemented by other arguments. I then raise the evidential argument from evil against naturalism to reduce the explanatory power of naturalism over evil. Finally, I develop the divine justice theodicy to enhance theism's explanatory power over evil. In this way, theism can defend the attack of the evidential argument from evil. My thesis can be further detailed as the following: First, l develop data-based skeptical theism to protect theism from being directly disproved by the evidential argument from evil while avoiding skeptical theism's two general unwelcomed consequences: global skepticism and moral skepticism. However, data-based skeptical theism works for people who hold no other competing theories to theism. When naturalism is regarded as a competing theory for the existence of evil, the evidential argument from evil can still disconfirm theism indirectly. Because data-based skeptical theism does not offer explanations for evil and naturalism can explain more evil than theism, evil's existence favours naturalism. Therefore, I supplement data-based skeptical theism with the other two arguments: the evidential argument from evil against naturalism and the divine justice theodicy. The former can reduce the explanatory power of naturalism; the latter can increase that of theism. The two arguments render theism's explanatory power over evil stronger than that of naturalism. Hence, theism will not be indirectly disconfirmed by the evidential argument from evil. Instead, theism becomes a more plausible theory than naturalism with respect to evil. In summary, data-based skeptical theism can prevent the evidential argument from evil from directly disconfirming theism; the evidential argument from evil against naturalism and the divine justice theodicy can help theism avoid being indirectly disproved by the evidential argument from evil. The three strategies constitute a solid threefold theistic response to the evidential argument from evil.

China's food retailing sector has experienced wide-ranging market restructuring and digital transformation during the past decade (circa. 2012-2022). As transnational retailers were declining and increasingly divesting from the Chinese market, increasingly powerful digital platform firms entered the traditional store-based food retail market and promoted an omni-channel transformation of the sector. Such market dynamism makes China's food retail market worthy of study and represents a contemporary context in many ways overlooked by economic geographers and cognate business studies scholars in recent years. The overarching aim of this thesis is to conceptualise the development of China's retailing economy during the 2010s, advancing the understanding of its structural change and digital transformations. More specifically, the research explores three particular trends that emerged during the past decade, which lead to three distinct, but related, empirical studies. These trends encompass the decline of food retail transnational corporations (TNCs) and their divestment from the Chinese market; the rapid growth of the import cross-border e-commerce (CBEC) industry and its fast-changing institutional environment; and the rise of digital platform firms and their role in the nationwide omni-channel retail transformation across food retailing. The empirical data that supports the thesis comes from forty semi-structured interviews with key informants, including senior executives from traditional store-based retailing chains and digital platforms, government officials, experienced industry consultants, and academic researchers. Importantly, the primary data were also complemented and triangulated by a wide range of secondary quantitative and qualitative materials, which encompassed information gathered from industry databases and publications, media reports, statistical yearbooks, company annual reports and official websites. The first empirical study contributes to the economic geography literature relating to retail globalisation by reconceptualising the evolving processes of territorial embeddedness of food retail TNCs, as the key competencies that had initially propelled them to success have atrophied in the face of a rapidly developing host market. The second empirical study conceptualises the industrial and institutional co-evolution of the Chinese import cross-border e-commerce (CBEC) industry over the past decade. The third empirical study conceptualises contrasting business models established by leading Platform Business Groups ('PBGs') to inform our understanding of the omni-channel transformation of the Chinese food retail market.

This thesis reports a toughening concept for carbon fibre reinforced polymer composites using continuous thermoplastic nanocomposite fibres in which dissolvable phenoxy was used as a toughener for the epoxy matrix and at the same time as a carrier to deliver graphite nanoplatelets (GNP) into desired positions of composites. Nanocomposites containing different GNP contents were prepared via masterbatch dilution method, and these were melt spun and drawn into fibres at varying take-up speeds. Structural characterisation revealed that the dispersion of GNP worsened with both increasing GNP content and take-up speed, especially in the fibre core region. Moreover, due to the strong shear stress exerted by the spinneret wall, GNP in the outer layer of the fibre showed a tangential orientation, which was absent in the fibre centre. As a consequence, the Young's modulus, which is a low-strain property, increased with both GNP content and take-up speed, while the tensile strength remained relatively unchanged because it was more sensitive to the GNP agglomeration. Theoretical analysis with simple rule of mixtures and shear-lag theory suggested the enhanced Young's modulus originated from the increased orientation degree (η₀) of GNP at high take-up speeds, which was not attainable from SEM/TEM characterisation. Moreover, the η₀ was found to be in good agreement with the normalised areal ratio (ratio of the area of the imaginary fibre sheath layer to the whole cross-sectional area) of fibres at each take-up speed, which could indicate the variation in the reinforcing efficiency of GNP in the fibre core and sheath layer due to the different dispersion qualities of GNP in these two regions. The electrical percolation threshold was 8.0 wt% for compression moulded plates, which was relatively lower than the rheological one (8.8 wt%) due to the fact that less GNP was required to form a conductive network that can be detected by the static electrical measurements. Conversely, fibres with the same GNP content were not conductive because of the orientation, separation and aggregation of GNP upon drawing. The dissolution behaviour of pure phenoxy and GNP/phenoxy fibres was studied in model systems using optical microscopy. Results showed that GNP/phenoxy nanocomposite fibres took longer than pure phenoxy fibres to completely dissolve, probably because the tangentially oriented GNP in the outer layer of the fibre hindered the interdiffusion of phenoxy chains and RTM6-2 precursors. The dissolution study also suggested that too high a dissolution temperature could lead to a premature gelation of the epoxy resin and that thinner fibres were preferred for use in composites. As a proof of concept, GNP/phenoxy fibres were 3D braided with carbon fibre (CF) to produce hybrid strands, followed by resin infusion to prepare composite laminates. Both phenoxy and GNP improved the Mode I interlaminar fracture toughness (GIc) of composites although phenoxy exhibited a more pronounced toughening effect. This was attributed mainly to the significantly intensified CF breakage and pullout in regions where pure phenoxy or GNP/phenoxy fibres initially located, as evidenced by the SEM observation on the fracture surfaces of composites. Crack deflection induced by the phase-inverted RTM6-2 nodules in the presence of relatively high local concentration of phenoxy, crack pinning caused by GNP and GNP pullout/separation were also proposed as the toughening mechanisms.

High stress in the teaching sector may be a contributing factor to teacher turnover and low retention. Across different populations, research has found that resilience can buffer against stress, anxiety, and depression. However, there is a lack of understanding about the psychological mechanisms that maintain and predict resilience based on theoretical frameworks. The first aim of this thesis was to examine the underlying mechanisms of trainee teachers' resilience at a given time, and whether affective-cognitive mechanisms predict resilience later in teacher training and when working as newly qualified teachers (NQTs). A limitation of existing resilience interventions is that they are often not theoretically driven and do not target on the key processes of resilience. Therefore, resilience interventions have been inconsistent in yielding positive outcomes. The second aim of this thesis was to evaluate whether a mechanism-based resilience intervention is feasible and acceptable to trainee teachers during teacher training. A series of studies were conducted to examine what mechanisms are associated with teachers' resilience. First, a pilot study confirmed the assumption that interpretation bias is associated with resilience. The purpose was to evaluate whether interpretation bias could be a modifiable mechanism associated with short-term stress response, stressors, and resilience among teachers, which informs the decision to adapt a positive interpretation bias training for resilience in Study 3. Then, using a cross-sectional and longitudinal approach (Study 1 & 2), we examined whether perceived stress and cognitive-affective mechanisms (i.e., interpretation bias, emotion regulation, cognitive control, cognitive flexibility, positive affect, anticipatory pleasure, responses to positive affect, meaning in life, and positive meaning finding) are related to resilience, and whether these mechanisms predict trainee teachers' future resilience. Overall, the results suggest that perceived stress and cognitive-affective mechanisms were associated with trainee teachers' resilience cross-sectionally (while controlling for the effects of age and gender). However, they did not consistently predict trainee teachers' future resilience. Notably, interpretation bias was significantly associated with resilience among qualified and trainee teachers, and consistently predicted trainee teachers' future resilience. Based on this finding, a resilience intervention that targets interpretation bias is warranted. Therefore, in Study 3, we conducted a pilot and feasibility study to promote resilience in trainee teachers via a mechanism-focused intervention (cognitive bias modification for interpretation; CBMI). Overall, we found that the adapted CBMI was feasible and acceptable to trainee teachers. However, revising the intervention length is necessary to improve adherence rates. This thesis contributes to the literature by identifying modifiable cognitive-affective mechanisms of resilience among teachers. This thesis sets the stage for the development of effective mechanism-based resilience interventions and informs future randomised controlled trials.

The importance of self-complexity on leadership has been discussed by several theoretical studies. The present research draws upon Linville's (1985, 1987) social cognitive perspective of self-complexity, which is defined as a function of two things: the number of aspects that one uses to cognitively organize knowledge about the self, and the degree of relatedness of these aspects" (Linville, 1985, p. 97), to serve as one of the limited empirical studies that examine its associations with leadership behaviors. To assess self-complexity, I develop a measurement that focuses on capturing the structure of the self at the role level and adopt the pairwise comparison method to enrich our understanding of self-complexity. In addition, I also draw upon self-control and goal shielding theories to examine the psychological factors such as time and centrality in self-complexity. Regarding leadership behaviors, I studied two often-seen leadership behaviors in the current research. For positive leadership behavior, I study servant leadership, which refers to "an (1) other-oriented approach to leadership (2) manifested through one-on-one prioritizing of follower individual needs and interests, (3) and outward reorienting of their concern for self towards concern for others within the organization and larger community" (Eva, Robin, Sendjaya, Van Dierendonck, & Liden, 2019, p.114). For the negative leadership behavior, I study abusive supervision, which is defined as "subordinates' perceptions of the extent to which supervisors engage in the sustained display of hostile verbal and nonverbal behaviors, excluding physical contact" (Tepper, 2000, p.178). To study the associations between self-complexity and leadership behaviors, I employ two different perspectives: Self-control and cognitive theories. From the self-control perspective, I adopt self-control and conservation of resource theories to examine the indirect associations between self-complexity and leadership behaviors via ego depletion, which is defined as "a state in which the self does not have all the resources it has normally" (Baumeister & Vohs, 2007, p.116). From the cognitive perspective, I employ cognitive theories to examine the indirect associations between self-complexity and leadership behaviors via cognitive flexibility, which is defined as "the ability of individuals to restructure knowledge in multiple different ways depending on changing situational demands" (Gino & Ariely, 2012, p.446). Four studies were conducted in the current research. The results in Study 1 (U.K. sample, N = 179) showed that the number of self-aspects and the degree of overlap among self-aspects were negatively associated with ego depletion and were positively associated with cognitive flexibility. Moreover, ego-depleted leaders were found to be more likely to behave abusively toward their followers, whereas cognitively flexible leaders were found to be less likely to abuse their followers. Regarding the psychological factors, results in Study 1 showed that the relationship between the interaction of a high number of self-aspects and a high degree of overlap among self-aspects and ego depletion was weaker when the time was short rather than high. The results in Study 2 (Taiwan sample, N = 332) also showed that the degree of overlap among self-aspects was negatively associated with ego depletion and positively associated with cognitive flexibility. Nevertheless, the number of self-aspects was not found to be negatively associated with ego depletion nor positively associated with cognitive flexibility. Study 2 also showed similar findings regarding leadership behaviors; ego depletion was positively associated with abusive supervision, and cognitive flexibility was negatively associated with abusive supervision. Study 3 specifically focused on testing the differences between the traditional pen and paper trait sorting task (N = 98) and the newly introduced method of measuring self-complexity (N = 110) and on examining the positive leadership behavior (i.e., servant leadership) in the current research model. The results in Study 3 showed that both methods had similar findings to Studies 1 and 2. However, the newly introduced method that measures the structure of the self at the role level seems to be a better approach than Linville's free-response format in predicting ego depletion and cognitive flexibility. In addition, ego-depleted leaders were not found to be less likely to display servant leadership in the workplace, and cognitively flexible leaders were found to be more likely to display servant leadership behaviors in the workplace. The results in Study 4 (N = 144 leaders and N = 450 followers) showed that the number of self-aspects and the degree of overlap among self-aspects were negatively associated with ego depletion and were positively associated with cognitive flexibility. Moreover, ego-depleted leaders were more likely to be perceived as abusive supervisors, and cognitively flexible leaders were less likely to be perceived as abusive supervisors and more likely to be perceived as servant leadership by their followers.

Cryptocurrencies, a product of financial technology (FinTech) innovations, are a recent phenomenon drawing extensive attention. This essay dissects cryptocurrencies in two distinct dimensions: asset pricing and portfolio management. Cryptocurrency returns are highly non-normal, casting doubt on the standard performance metrics. Therefore, we apply almost stochastic dominance (ASD), which does not require any assumption about the return distribution or degree of risk aversion. From 29 long-short cryptocurrency factor portfolios, we find eight that dominate our four benchmarks. Their returns cannot be fully explained by the existing three-factor coin model, so we develop a new three-factor model where momentum is replaced by a mispricing factor, based on size and risk-adjusted momentum, which significantly improves pricing performance. The new three-factor model is robust to various out-of-sample tests. Additionally, we investigate diversification benefits on a cryptocurrency-factor level within a portfolio management framework. The uncertainties of input parameters cause the traditional mean-variance framework to be misleading. To enhance the out-of-sample performance of optimised portfolios, we combine machine learning and various asset-allocation strategies to tackle the estimation errors of these input parameters. Through estimating out-of-sample performance metrics, we find that cryptocurrency factors formed on size and momentum groups can add substantial diversification benefits (e.g., statistically significant risk-adjusted returns) to a traditional stock-bond portfolio across both conventional and novel asset-allocation strategies. Furthermore, our results are robust to transaction costs, an alternative benchmark, and a rolling-window estimation. Therefore, this essay suggests that investors could i) estimate the expected cryptocurrencies by our new three-factor model, and ii) include cryptocurrencies in a stock-bond portfolio to gain considerable diversification benefits.

In this dissertation, I K-8 mathematics teacher agency and learning within a research-practice partnership (RPP) context. I asked three primary research questions: (a) How is teacher agency enacted within the RPP?; (b) what do teachers perceive they are learning from their participation in the RPP?; and (c) how is teacher agency connected to teacher learning in the RPP context? I utilize both an ecological framework for teacher agency (Priestley et al., 2015a) and situated learning theory (Lave & Wenger, 1991) about teacher learning to frame the questions and analyses. As an independent investigation that drew from a larger, multi-site project (i.e., as part of the broader RPP), case study methodology (Yin, 2018) was employed to closely examine K-8 mathematics teachers' participation and experience in a co-design team in one school district. The team's primary goal was to improve the quality of mathematics instruction across their school district by employing a design-based implementation research (DBIR) process. The district and school administrators and teachers were involved in the improvement team's work. Multiple types of data were collected including semi-structured interviews, meeting observations, and related documents for two years. Additionally, those data sources were supplemented with a targeted set of semi-structured interviews about teacher agency and teacher learning to better address the study's research questions. I analyzed the data using primarily qualitative methods and, specifically, with coding approaches associated with grounded theory (Bernard et al., 2016; Charmaz, 2006). The study's findings suggested teachers enacted their agency in various ways in and outside of their team meetings and the quality of engagement changed over time. From the ecological perspective of teacher agency, multiple mediating factors from iterative, projective, and practical-evaluative dimensions were found to influence teachers' participation by functioning as supporting and inhibiting factors. Amid the teachers' involvement in the RPP, teacher learning occurred not just about mathematics teaching but also about other areas, such as leadership, collaborative abilities, and understanding of the DBIR process, which are all crucial components of capacity building for creating sustainable learning opportunities even after the partnership with the university. Also, the interviews revealed potential connections between teacher agency and learning. In the final chapter, I delve into the implications of these findings. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]

This study aimed to enhance student engagement in language classrooms by developing a professional learning community (PLC) in the Modern Language Department in a high school. This exploratory study uses a mixed-method design collecting both quantitative and qualitative data through various instruments. Incorporating improvement science, the teachers engaged in an 8-week plan-do-study-act cycle to learn and share instructional practices focusing on student-centered strategies. This study aimed to understand: (1) How do teachers engage in a PLC? (2) How do teachers describe their learning of new instructional practices and strategies and apply what they have learned in their classes? (3) How do students and teachers identify and describe any changes in student engagement after teachers implemented what they learned from the PLC? The methodology included a pre-PLC survey, teacher interviews, exit tickets for PLC meetings, meeting notes, and pre- and post-student surveys on the song project intervention. The findings revealed that veteran teachers could learn and develop knowledge and skills by sharing and discussing with each other within the structured time and space provided by the PLC. Student engagement was enhanced as the teachers used student-centered instructional practice they learned from the PLC. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]

Language is a communication tool used by everyone in their daily life as a means to convey information and arguments to others. Even though the Ministry of Education (MoE) in China has been switching from exam-oriented education to cross-cultural communication education, the classroom practices, teaching methods, evaluation system, school structures and materials, and aspects of the administrators and teachers are still traditional and contrary to the principles of cross-cultural communication and exchange. This study offered a biography-driven instruction (BDI) approach as a new method in teaching English as a Foreign Language (EFL) in China and examined two EFL teachers as they navigated the conceptual, pedagogical, structural, and political dilemmas of implementing a BDI approach in a Grammar Translation Method (GTM) context. Professional development about BDI, including coaching sessions, was provided to two EFL teachers working in two public junior high schools in the province of Shaanxi and Gansu, and the understanding of the BDI approaches on teachers' practice and implementation was analyzed. Different types of data sources, such as in-depth interviews of participants, pre- and post-surveys, and lesson plans and artifacts were collected and analyzed to better understand how these teachers made meaningful use of this new approach and negotiated its use in a GTM-dominate context. The findings, in general, showed that the BDI approaches helped these teachers build a more constructivist classroom while combining GTM and BDI methods, views, and classroom applications. This study provided findings regarding exploration, integration, and localization of a BDI approach in the context of GTM. Well-planned and well-organized teacher development with on-going support and counseling is necessary. Further, professional development and coaching conversations have important roles in teacher development. This study also revealed that the teachers were the most important people in implementing a new educational approach and shift in teaching methods as required by the MoE to ensure they can reach a more communicative and constructivist classroom, regardless of any insufficiencies. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]

Composite materials are attractive in aerospace and automotive applications due to their high stiffness-to-weight ratios. Non-crimp-fabric (NCF) reinforced composites have been receiving attention in the composite market due to their cost-effectiveness and excellent mechanical performance. However, compared to traditional unidirectional laminates (UD), NCF composites have different mechanical behaviours due to their heterogeneous internal structures. The ability to predict the mechanical performance of NCF composites is necessary for a robust and reliable design. The objective of this PhD research was to develop numerical methods to predict the in-plane mechanical behaviour and damage tolerance of the carbon fibre reinforced biaxial NCF composites. The in-plane mechanical behaviours of NCF composites were investigated at different scales by conducting multiscale analyses in the LS-DYNA finite element (FE) software. The macroscopical FE modelling results were validated by a series of in-plane characterisation tests of biaxial NCF composites. The compression-after-impact (CAI) test was adopted to assess the damage tolerance of the composite laminates. The complex failure mechanisms of NCF composites involved in a CAI failure process were comprehensively studied by experimental methods. The experimental results contributed to the validation of FE models to predict the low-velocity impact (LVI) and CAI behaviours of NCF composites in LS-DYNA. Furthermore, different laminate designs were employed to change the CAI behaviour of NCF composites by altering layup sequence and ply-level hybridisation. An optimised scheme was proposed to enhance the CAI behaviour of NCF composites, providing a practical guide to damage tolerance design.

This thesis presents an empirical analysis of the influence of regional institutions on the social and economic inequalities through crises and policies changes. It is organised into four independent chapters. Chapter 1 investigates how an indigenous government institution affect contemporary economic development. Government institutions have long been known as critical drivers of growth. However, it is still unknown whether the institution is resilient after experiencing large political shocks. Using the textual data from government records, we examine how differences in local government quality in Imperial times determined current differences in economic development. Chapter 2 focuses on government institutions' role in explaining performance differences in pandemic prevention. Current research so far has primarily overlooked linking institutions to social outcomes. How institutions react to public health crises has attracted minimal attention. We collected and combined the Black Death's mortality rate and duration data with the location of autonomous cities to study whether a greater degree of autonomy could have allowed cities to react effectively after the pandemics. Chapter 3 moves beyond public-order institutions to study the persistent economic effect of the private-order institutions. We used historical administrative data, the Statistics of Agriculture and Commerce, to digitalise the conditions of the chamber system of China in the 1910s. To measure contemporary business performance, we rely on the Firm Registration Database, which systemically documents start-ups' emergence since 1975. Using a difference-in-differences strategy, we identify cities heavily permeated with the historical chamber system as enjoying more entrepreneurs and flourishing with innovative performances when the market economy reopens. Our instrumental variable strategy has also verified these results. Chapter 4 investigates the influence of culture. Within the background of son preference, we discuss how the family planning policy distorts a couple's decision-making about childbirth. Under natural conditions, parents satisfy potential son preferences by giving birth until a son is born. However, with restrictions on family size, the cultural preference for sons can no longer be smoothed out by adding additional children. Parents may practise prenatal and postnatal discrimination against daughters, lowering their probability of survival relative to sons. Drawing on the natural experiment of the Two-Child Policy, we examine whether inconsiderate policies could further trigger the dark side of a backward culture.

Among all the ways in which Taiwan's transition to democracy could be examined, this thesis focuses on the role of the media and their relation to generations. Complementing Karl Mannheim's generation theory with narrative theory and mediation theory, it develops a 'mediated generation' approach to examining how people born in different times make sense of and engage with democracy. The thesis compares the generation of Democratic Consolidation, born in the 1980s and coming of age around the first party change in the ruling government in 2000, with their parents, dubbed the generation of Soft Authoritarianism, experiencing authoritarian rule first-hand as they lived through the Chiang Kai-shek or Chiang Ching-kuo regimes. By using qualitative methods including focus groups, paired interviews and individual interviews, I found that both generations commonly associate democracy with polarisation, corruption and populism. However, their responses to these problems of democracy differ as regards their media engagement and orientation to politics. Blaming legacy media for driving partisan polarisation, the Democratic Consolidation generation normally relies on search engines for fact-checking, whereas the Soft Authoritarianism generation is more likely to trust television but will compare news across multiple television channels. As a reaction to corruption and populism, both generations are attracted to authentic, unconventional politicians, but their preferences and reasoning are different. In ascribing blame for populism, the generation of Democratic Consolidation criticises irrational netizens and Internet anonymity, whereas the generation of Soft Authoritarianism criticises the youth for failing to think deeply and for their susceptibility to popular culture influences, considering that their use of the Internet and social media for political expression has made Taiwan democracy populist. In addition, this thesis argues that, on the one hand, the two generations develop their knowledge of democracy and strategies for dealing with political and media problems in separate mediated spaces. On the other, in everyday life they come together as family, and their family practices with the media, especially in Family Groups on LINE, dovetail with those in the real world; that is, media predominantly perform functions of reinforcing the principle of 'jia chang' (literally meaning 'home ordinary') and a harmonious but hierarchical parent-child relation which shapes practices of (dis)engagement in political talk within the family. This thesis demonstrates that generational experiences and identities can be empirically examined as narratives. Evidence of how the media contribute to certain generational narratives, identities and actions confirms the usefulness of adopting mediation theory to study generations. This mediated generation approach, used to explore intra-generational differences, successfully provides some new insights into 4 Mannheim's concept of generation units. The thesis concludes by exploring the implications of these findings beyond the Taiwanese case for the relationship between media, generations and democracy in the 21st century and the importance of critical reflection on transitional societies.

With the development of omics technologies, there is a large amount of biological data available. The biological data is usually complex, ill-sampled, and high dimensional. As a result, gaining insightful knowledge from the biological data is still a challenging problem. Many of those biological data can be represented using graphs. Over the years, many graph-based machine learning methods have been developed to analyse graphs in many tasks, such as module detection, feature engineering, and link prediction. This thesis provides the applications of graph-based machine learning methods on three types of biological data with the following contributions: • A novel method that allows overlapping between the peak modules in the metabolite annotation problem on liquid chromatography-mass spectrometry untargeted data. • A novel method that allows high freedom of the shapes of the topological modules in the detection of network biomarkers for gene expression data. • A novel method that incorporates the implicit networks constructed from the drug-target interaction network in the drug-target interaction prediction problem. With these contributions, we can gain more understanding of the biological data.

Advancements in information and communication technology (ICT) have rev-olutionized communication between service providers and consumers. Both academics and industry have acknowledged that social media platforms play a vital role in tourist decision-making and destination branding, but a comprehensive assessment of how social media affects tourism behaviour and its capacity to boost destination value remains elusive. The purpose of this study is to fill that gap by examining the impacts of user-generated content (UGC) in social media on tourist behaviours, as well as its role in contributing to destination value co-creation. Gulangyu, a well-known World Heritage Site and a heritage tourism destination in southeast China, served as a case study in this study. An on-site questionnaire survey and online narrative analysis were conducted to investigate how UGC influences visitors during their entire vacation travel process and how travellers share their travel experiences through social media, further enhancing the value of the destination. Both qualitative and quantitative methods were used in this investigation. During the first-stage on-site questionnaire collection, 500 Gulangyu tourists were approached in 2018. The structural equation modelling (SEM) analysis was conducted based on a sample of 439 travellers who reported they planned their trips using user-generated travel information on social media. Study results indicate that UGC impacts tourist behaviour across the entire holiday-related decision process, from destination selection, purchasing be-haviour to loyalty behaviour. On the second round of data collection, 97 travel blogs and their interactive comments with readers were selected as a suitable qualitative data set. The online narrative analysis of travel blogs identified three unique themes that contribute to the co-creation of destination value through the sharing of holiday travel experiences on social media (Xu et al., 2022). A discussion is presented on the theoretical and managerial implications of each study regarding how UGC influences tourists' decision-making as well as how tourists' sharing behaviour affects destination value co-creation. In a conclusion, the limitations and future research directions are indicated.

This thesis explores how brands have come to (re)organise the fresh food sector in China, investigating how brands have increasingly intervened in the way in which fresh food is produced, distributed, exchanged and consumed in China in recent years. While there has been some literature examining the role of brands in the fresh food sector, little attention has been paid to it in the East and emerging countries. Until recently, most fresh food products were treated as ordinary commodity items in China, sold in wet markets or other traditional channels without any brand information, but in recent years brands have begun to enter the fresh food sector. This thesis explores the brand as a significant object of mediation that assembles and intermediates a series of actors, materials, practices, associations, values and norms between food production and consumption, operating alongside wet markets. This innovative understanding of brands and fresh food branding uses a socio-cultural framework for assessing the transformation of the food sector in China, as well as expanding the field of brand studies. Through an analysis of a selection of government policies and documents on regulating agriculture and food, the thesis describes the recent history of the state-led transformation of the food sector, developing an analysis of the political and economic context in which brands emerged in China's fresh food sector. The thesis also presents a multi-case study of four brands, analysing the specific ways in which they (re)organise the fresh food sector. The conclusions propose that brands, as significant social mediators, are adaptable to various political, economic, social and cultural contexts.

Focusing on Chinese-English bilinguals, this research project aims to explore the cognitive mechanisms underlying bilingual language processing and the effects of bilingual language experience on domain-general cognitive control in language comprehension and production processes through four behavioural studies. Specifically, it examines (1) how cognitive control adapts to dynamically-changing cognitive demands in language processing; (2) how bilinguals' habitual language use patterns and contexts affect their development of domain-general cognitive control, through the convergence of linguistics and behavioural experiments approach. Study 1 explored how bilinguals' habitual language experience affects their cued code-switching production and performance in cognitive shifting and inhibition tasks. Similarly, Study 2 also focused on understanding the effects of bilingual language use habits on cognitive shifting and inhibition; however, it investigated the relationship between bilinguals' spontaneous language production and cognitive control. Studies 3 and 4 examine the adaptations of cognitive control in comprehending bilingual utterances in different interactional contexts. Specifically, Study 3 investigates how different language interactional contexts modulate domain-general inhibitory control in bilingual language comprehension. Study 4, consisting of two experiments, compares the effects of bilinguals' code-switching habits and L2 proficiency on their inhibitory control during language comprehension in habitual and induced language use conditions. The results show that (1) cognitive control processes are adaptively deployed to produce and comprehend bilingual utterances in different interactional contexts; (2) bilinguals' habitual language use patterns and language proficiency are significant factors modulating their domain-general cognitive inhibition and shifting efficiency; (3) the magnitude of bilingual language use experience effects on cognitive control vary significantly across processing languages in naturalistic and mandatory (i.e., induced) language use conditions. In general, the project addresses bilingualism being a dynamic process that includes multifarious individual differences that may affect cognitive control. Therefore, more comprehensive measures to characterise bilingual habitual language use experience are needed in future studies.

Videos are an appealing source of data to train computer vision models. There exist almost infinite supplies of videos online, but exhaustive manual annotation is infeasible. The goal of this thesis is to learn strong video representations efficiently via self-supervised learning: a method that learns from the data rather than human annotations. The thesis is structured around three themes: (1) self-supervised learning for short-term videos, (2) efficient video representation learning, and (3) self- supervised learning for long-term videos. For short-term videos lasting only a few seconds, we show that predicting the video in the future is a strong learning signal at a large scale. We further show that strong video representations can be learned by taking two complementary modalities, namely RGB and optical flow, and using them to teach each other. For efficient video representation learning, we show that large-scale pre-trained vision-language models can be effectively adapted via a prompt tuning technique. We also show that dropping image patches can accelerate the finetuning of classification tasks and pre-training of video-language models. For long-term videos that last longer than a few minutes, we show that temporal alignment networks can be trained from the weak visual-textual correspondence within instructional videos. The resulting networks can automatically clean up the natural videos for effective vision-language training. In addition, we show that movie description models can be trained by leveraging the pre-trained vision- language models.

Creativity is critical to the growth of service organisations. Yet, the question of how service organisations promote creativity remains unanswered. Prior research has highlighted the importance of team empowerment in frontline service and found that being empowered leads to greater performance; this raises the question of whether empowering leadership can bring up team creativity in a frontline service context. Thus, the purpose of this research is to investigate the relationship between empowering leadership and team creativity in a frontline service context. To do so, a theoretical-driven model was developed wherein empowering leadership could influence team creativity through team identification - knowledge sharing; relationship conflict - knowledge sharing, respectively. Based on a sample of 51 frontline service teams from China's banking sector, this study found that empowering leadership can facilitate knowledge sharing by enhancing team identification and reducing relationship conflict. The presence of a high level of team efficacy magnified the positive effect of empowering leadership on knowledge sharing through team identification but was not significant through relationship conflict. More importantly, the results support the proposed sequential mediation model in which empowering leadership influences team creativity through the two parallel mechanisms: team identification-knowledge sharing and intragroup relationship conflict-knowledge sharing. Based on these findings, this research revealed the mechanisms by which empowering team leaders may elevate team creativity. By doing so, this research provides some theoretical suggestions for future research and practical implications for service organisations and managers.

Human activity recognition (HAR) has been studied for decades in computer vision and has shown great success. However, as people are caring more about privacy issues, researchers are investigating non-intrusive HAR systems using radar-based techniques, among which millimetre-wave (mmWave) radars have received great popularity due to their capability of capturing high-resolution spatial information about the scene. This thesis presents a systematic study of HAR using mmWave radars. It explains the fundamentals of mmWave sensing techniques, discusses its use in HAR applications, and highlights the challenge of the sparse and noisy data through a purpose-built simulation system that can import arbitrary 3D models to form a scene and simulate the radar signal with configuration antenna settings. A software framework for managing multiple radars is presented that allows real-time data transmission, data processing, and result visualization. Based on the software framework, three HAR systems are presented. First, a human detection and tracking system is presented as the fundamental of HAR. The system operates two radars simultaneously that verify each other's detection and significantly reduce the probability of false alarms. The system achieves 90.4% sensitivity and 98.6% precision when detecting up to four people in the room. Then, a human posture estimation system is presented that uses two radars as a vertical array and a neural network model to estimate the joint positions of the person. The system achieved over 71.3% accuracy when detecting postures that are commonly seen in an office environment with arbitrary limb motions. Finally, a human vital sign detection system is presented that uses one mmWave radar to detect a person's heart rate when exercising on a treadmill. It overcomes the challenge that the heartbeat signal can be difficult to extract when there is body movement, and achieved a low error rate of 5.4%.

Neurodegenerative diseases cause severe health and social problems. They create a health and economic burden on individuals and their families. Currently, there is no efficient treatment, apart from some attenuating medicines, for the majority of neurodegenerative diseases. Neurone loss is a consistent characteristic of these diseases. The hippocampal dentate gyrus is one of the regions in the adult mammalian brain capable of generating new neurones throughout life. The neurogenic niche appears to play an important role in the neuronal dysfunction associated with neurodegeneration. However, the detailed mechanisms underpinning this process are still unclear. To explore the neurogenic niche in Frontal Temporal Dementia (FTD) and Parkinson's disease (PD) we investigated disease-relevant rodent models and donated post-mortem human brain samples. Immunohistochemistry and immunofluorescent assays were developed to allow the quantification of dividing cells and immature neurones in the adult hippocampus of selected tissues. The TDP43-Q331K knock-in mice, a model of FTD, revealed a significant reduction in the number of dividing cells (Ki67+) and immature neurones (Dcx+). Distinct morphological stages of immature neurone development were observed in this disease model. Further rodent neurotoxin-based models were used to characterise neurogenesis in PD. However, contrary to previous reports, these models did not consistently induce significant deficits in adult hippocampal neurogenesis (AHN). Human post-mortem brain samples were used to explore the neurogenic niche, but we were unable to consistently observe immature neurones in non-diseased brain samples. In summary, this project identified significant AHN deficits in the FTD mouse model. These require further analysis to determine their function and possible role in human disease.

With the rapid development of electronics towards miniaturization and increased functionality, the heterogeneous integration of electronic components necessities a paradigm shift in the design of physical dispositions and selection of materials, as the conventional joining approach is becoming increasingly restrictive. Low temperature bonding is considered as one of the most favourable methods to enable advanced packaging with the uses of a wide range of intermediate materials, such as pastes, preforms and conductive adhesives. However, various drawbacks are commonly observed in the current practices with existing processes and materials. Therefore, the development of advanced interconnect technologies and novel materials is imperative to meet the ever-increasing demands. In this study, interconnection manufacturing routes have been developed with an attempt of exploiting the unique Cu nanowire/micropillar arrays produced by template-assisted electrodeposition (TAED). In the present work, the Cu nanowire array (Cu NWA), Cu-Sn nanocomposite interlayer (Cu-Sn NI) and Cu micropillar array (Cu MPA) have been successfully prepared through TAED. The growth characteristics of these arrays or interlayers are examined to make them more tailorable in the subsequent bonding processes, by considering the effects of deposition parameters on the microstructures, morphologies, and process optimisation. In line with demands of three-dimensional integrated circuit (3D IC), the uses of purposely designed Cu-Sn NI are made in the experiments to accelerate the transient liquid phase bonding (TLPB) process, which is extremely slow based on conventional bonding process. The Cu-Sn NI consists of Cu NWA infiltrated with Sn, in which the nanowires are densely but separately distributed perpendicular to the surface of Cu substrates. Results show that this Cu-Sn NI can accelerate the TLPB process effectively by over 20 times. Moreover, the intermetallic compound (IMC) joints formed with this interlayer possesses refined equiaxed Cu6Sn5 grains which exhibits ~29.1% higher bonding strength than the joints bonded with Sn interlayer of the same thickness and bonding conditions. The underlying mechanism of such bonding process is therefore investigated, with an emphasis put on the elaboration of the formation of various defects and the potential mitigation solutions. In order to supress the formation of voids and solder bleeding that are commonly observed in self-propagating exothermic reaction (SPER) bonding, the Cu-Sn NI is also utilised to replace conventional solder interlayers (e.g., Sn-based alloys). The bonded structures achieved under different preheat temperature and pressure are examined in terms of their microstructures and mechanical integrity. The research has found that the joints formed with Cu-Sn NI exhibit significantly improved bonding quality in comparison with those with the Sn solder layers of same thickness under the same bonding conditions. It has been proven that the uses of Cu-Sn NI are effective to reduce porosity and solder bleeding, which has resulted in a higher shear strength. The fundamental mechanism of such improvements has therefore been explicitly explained in connection with the unique properties of Cu-Sn NI during SPER bonding. A novel fluxless insertion bonding is also proposed and realized by pressing the Cu MPAs into solder bumps under a low process temperature. The morphologies, microstructures and shear strengths of the joints formed under different bonding parameters with Cu MPAs have been investigated in detail to understand the formation of the interfaces and joints. It is found that effective interconnections could be achieved even at 20 °C due to the mechanical interlock between the solder bumps and Cu MPAs. Joints formed with longer bonding time at higher temperature possess higher shear strength due to the formation of more IMC at the bonding interfaces. Moreover, the morphology of the Cu MPA is found to have significant influences on the shear strength of bonded joint. Sn coating is also found to be helpful in the formation of effective metallurgical interconnection, leading to the enhancement of bonding strength. All aforementioned materials and bonding methods developed, with a wealth of research findings yielded in this work, may open a new window of innovative manufacturing routes with great potential for advanced electronic interconnections in the future industrial applications.

Molecular systems with functional domains serve as a practical motivation for understanding the factors that contribute to the interaction energy. The ability to decompose the interaction energy of a group of interacting subsystems is an important method in studying the chemical nature of the interactions. Energy decomposition analysis (EDA) is a family of schemes that allows such dissection of the interaction energy into chemically relevant components depending on the scheme used. Since different EDA schemes decompose the interaction energy differently, the interpretation of the resulting components differs among schemes. However, various EDA schemes provide complementary insights into the interactions between chemical entities. In this work, two EDA schemes are developed or extended: Hybrid Absolutely Localized Molecular Orbitals (HALMO) and Combined Localized Molecular Orbitals (CLMO). Both EDA schemes have been implemented as part of a linkable library alongside the computational chemistry package, ONETEP. Since ONETEP is a linear-scaling software package, an important application of such decomposition analysis is in the study of large, nontrivial molecular systems for more insightful understanding of chemical interactions, which in turn can lead to more accurate and focused design of chemical systems. Systems such as biomolecules usually contain several self-stabilizing domains that can fold independently and have important functions. Defining the fragments of a supermolecule is necessary in EDA, and if done appropriately given the context of a particular application, the fragmentation of a biomolecule can elucidate the intramolecular interactions that contribute to the functions of the system as a whole. Two major methods of self-consistent field for molecular interaction (SCF MI) are examined and made more mathematically transparent. SCF MI was originally designed to exclude basis set superposition error (BSSE) from molecular interactions. However, SCF MI is also used in separating charge transfer from polarization in HALMO EDA. Studying several SCF MI methods provides HALMO EDA alternatives for separating charge transfer as an EDA component. Due to ONETEP's linear scaling, large biomolecules or large samples of biomolecules could be studied for their interactions or distributions of specific properties. The primary type of biomolecule studied in this work is double-stranded DNA (dsDNA) and how its stability relates to guanine-cytosine (GC) content, which is a measure of the amount of guanines or cytosines in nucleic acids. By examining the interaction energies in terms of EDA components, contributions to the variabilities within and across GC-content groups are examined and are correlated with differing stabilities despite having same GC content. Ensemble density-functional theory (EDFT) optimizes the molecular orbitals of a system at finite electronic temperatures and allows occupancies to be fractional when the band gap is sufficiently small. EDA methods are normally developed for the pure state and pose difficulties in decomposing the interaction energy of a conductor due to the fractional occupancies that are part of the optimization process in EDFT. As such, fractional occupancies have been incorporated in the EDA optimization process of the fragmented species, thereby allowing EDA to be applied to systems of relevance to catalysis and metallic systems. The adaptations of EDA and SCF MI to metallic systems are novel and were validated using samples from catalysis and batteries, and HALMO EDA has provided reasonable decompositions of interaction energies and revealed some trends from SCF MI that correlate with charge distributions and chemical intuition.

This thesis comprises three essays on international finance and asset prices. The research uses international macroeconomic data and asset prices to answer questions in the field of macro-finance. The first chapter shows that global inflation risk is priced in the international equity market and that its risk premium is strongly time-varying. This time variation is intimately tied to the fluctuations in the cyclical nature of inflation around the world. Indeed, I document a simultaneous shift in the global inflation risk premium and the covariance between inflation and future consumption growth during the 2008 financial crisis. I rationalise the time variation and sign-switch in the predictive content of inflation with the effect of the zero lower bound (ZLB). The second chapter is co-authored with Juan Antolin-Diaz, Gino Cenedese and Lucio Sarno and focuses on the impact of the US monetary policy on currencies. We find that currencies that react more negatively to tightening monetary policy shocks or more positively to positive central bank information shocks earn higher average excess returns. Currency characteristics help explain the heterogeneity of this reactivity across currencies and over time. We build exposure indices to gauge this effect using high-frequency co-movement of interest rates and stock prices around policy announcements. Long-short trading strategies that use such exposure indices as a signal display significant alpha after controlling for the dollar and carry factors. Thus, investors require compensation to hold the currencies that are more sensitive to the US monetary policy. The third chapter is co-authored with Lucio Sarno and Maik Schmeling. It provides an empirical investigation of the importance of trade balance news in explaining the puzzling high returns to the well-known 'currency price momentum' strategy, i.e. the currency strategy that relies on past return continuation among winner and loser currencies. To do this, we introduce another currency strategy, denoted as trade news momentum, which is driven by the heterogeneous currency responses to trade balance news announcements. We find that returns to the trade news momentum command strong explanatory power toward the currency price momentum alpha. Furthermore, we show that the cross-sectional variation in the currency exposure to trade news is understandable in terms of basic economics of exchange rate regimes and the level of capital account restrictions of a country.

This dissertation is an ethnographic bricolage of portraitures of dogs and their humans in Edinburgh and on the internet. Through stories of dogs and dog people before and during the COVID-19 pandemic, it explores the formations, negotiations, and maintenance of knowledge, childhood, race, loneliness, class, and music in the nature-cultures of more-than-human kinship. In the chapters, I navigate the diverse and (sometimes) contradictory collections of sentiments, sensibilities and subjectivities as produced by and projected onto bodies and movements while situating them in the temporally and spatially specific contexts of Edinburgh and the internet from 2019 to 2020. At the heart of this work is (more-than-human/multispecies) kinship, seen as a form of domestication with intended and unintended consequences. Even though warmth, joy, and companionship of more-than-human kin relations did endure and flourish, so did conflicts, hardships, and tensions. While kinship across the species boundary was described as a source and practice of unconditional love, the histories of selective dog breeding, neoliberal late-capitalism, and animalisation of racialised humans were also entangled with and permeated more-than-human kin- making. An exercise in staying with the trouble (Haraway 2016), kinship between dogs and their humans constantly meshed together the warm and the chilling, the fuzzy and the prickly, the comfortable and the tense. Building on more-than-human anthropology, anthropology of kinship, multidisciplinary human- animal studies, and anthrozoology, this dissertation speaks to wider questions and concerns about what it means to make kin with one another and how to live a good life together amidst various troubles of our time-space as well as its various reflections, refractions, and reverberations in the mundaneness of everyday life in Edinburgh and on the internet.

Purpose: The purpose of this doctoral thesis sets out to explore and elaborate on the impact of artificial intelligence (AI) and blockchain adoption in corporate governance from ethical perspectives. Positioned within the corporate governance domain, this study adopts an explicit business perspective to study corporate governance change with emerging AI and blockchain technological tools in general and focuses on the ethical use of technologies specifically. As such, this empirical investigation aims to help organizations understand the ethical benefits and ethical dilemmas of using AI and blockchain in businesses and draw plans on how to govern these technologies ethically for the benefit of the business and society. Design/Methodology/Approach: This study adopts specific techniques and a pragmatic, step-by-step netnography approach to investigate online traces from social media sites and extends these online explorations with online semi-structured interviews. The research design of this investigation follows step-by-step procedures that are methodologically sound to ensure rigor in this investigation to enhance the trustworthiness of this study. In total, this research collects an abundance of data: 34 LinkedIn Posts with Comments; 12 Webinars; 22 YouTube Videos; 19 Videos; 10 Podcasts, and 17 semi-structured interview videos. The video, audio, and interview data have been transcribed into textual data total of 453065 words for thematic analysis using NVivo software. Enough time has been allocated to the iterative process of data collection and data analysis. The analysis moves back and forth to the point when theoretical saturation is achieved. The data structure extracts from data in this study illustrate the analytic claims that match the analysis and data together, to ensure a good fit between described method and reported analysis are consistent. Findings: This study develops a thematic framework that constitutes the corporate governance transformation with the ethical use of AI and blockchain technology. This framework provides a holistic understanding of why corporate governance needs to change, especially with the emergence of blockchain and AI technologies, what changes will corporate governance encounter, and how corporate governance can imperatively respond to the ethical use of these technologies. Specifically, it explicitly provides comprehensive understanding of the ethical benefits and ethical concerns of using AI and blockchain technologies in corporate governance, and reveals how companies can govern the use of these technologies ethically. In general terms, the findings of this study support the notion of corporate governance change to transform business models and processes to leverage the new capabilities of AI and blockchain technologies, to priories creativity, speed, and accountability, to replace the old business model, to foster agile or collaborative governance to deal with uncertainty, agility, adaptiveness, and cooperation in the digital world, to foster a network and platform strategies to drive success. This study goes beyond the extant corporate governance scholarship to assess the technological impact to capture values for companies in ethical ways to sustain future growth. Additionally, the notion of corporate governance is further specified and significantly expanded by this study to assess the adoption of AI and blockchain as new corporate governance tools or mechanisms, to enhance ethical benefits when used properly, and mitigate ethical dilemmas with proper checks and balances, safeguards in place, to help organizations stay relevant in this digital transformation and be ethical and sustainable. This study empirically corroborates that in theory, the use of blockchain and AI can enhance ethical practice by detecting fraud and anomaly activities, due to the unique capabilities of blockchain and AI technologies. Further, this research adds depth and specificity by identifying the ethical concerns of using blockchain and AI in corporate governance. The study empirically reveals the ethical concerns of privacy issues, unethical use of data, job transformation and replacement, and algorithm bias that companies will encounter when they use these technologies. In addition, the findings of this study suggest how companies can ethically govern the use of these technologies in socially responsible ways as they transform digitally. Originality/Value: The emergent thematic framework is constructed from the empirical and analytical procedures specifically and purposely designed for this study. This study makes theoretical contributions to knowledge and enriches the extant works of literature, and also provides practical contributions to the ethical use of disruptive technologies, future workforce, and regulations. However, the study was conducted within certain theoretical, methodological, empirical, and pragmatic conditions, which might constitute particular limitations and constraints. Therefore, the last section of this thesis elucidates and suggests the directions for future research.

In complex natural environments, sensory systems are constantly exposed to a large stream of inputs. In order to encode the everchanging sensory surroundings efficiently, sensory systems rapidly adapt their responses to common stimuli that are parts of predictable patterns while retain their responsiveness to the novel, unexpected stimuli that usually contain behaviourally important information. Such a differential processing capacity is commonly referred to as novelty detection, whose neural signatures have been identified in human electroencephalogram (EEG) recording as mismatch negativity (MMN) and single-neuron electrophysiology as stimulus-specific adaptation (SSA). Despite the prevalence of the rare-selective responses in several sensory modalities, the intracellular and network mechanisms underlying are not yet fully understood. In this thesis, we employ the modeling tools in the computational neuroscience community, combined with computer simulation, to investigate the neuronal network dynamics producing novelty-detecting signals in the auditory and somatosensory cortex of the rodent model. Following an earlier auditory SSA mechanism in recurrent network with synaptic depression [Yarden and Nelken, 2017], we provide further support for the SSA circuitry hypothesis by using for the first time a similar recurrent network structure yet mediated with neuronal adaptation implemented in both simulated mean-field and physical neuromorphic spiking models. Our results suggest that SSA and its pertinent properties could arise from the differential excitation of tuned neuronal populations via propagation of population activity and short-term adaptation mechanism (neuronal adaptation or synaptic depression). In addition, we adapted and expanded this cortical circuit of the auditory SSA to a multi-scale thalamocortical network model that explains and reproduces physiologically observed neuronal response patterns that demonstrate signatures of both SSA and MMN in the somatosensory cortex [Musall et al., 2017]. Specifically, our results indicate that the novelty signal arises from the complex recurrent interplay between thalamic neurons and cortical neurons in layers 4 and 6. This work therefore provides a concrete mechanism that can serve as a starting point for further investigating the neural circuit mechanisms underlying novelty detection.

Upper limb prostheses offer those with limb loss a solution to restore some of their lost functionality by allowing them to participate in bilateral tasks, especially those required for daily living. Whilst there is a wide range of upper limb prostheses available, there remain high device rejection rates. Low functionality and discomfort are major factors in prosthesis rejection, which had been identified as challenges more than 60 years ago. These issues have not been effectively addressed due the lack of design tools for engineers and clinicians. Upper limb prostheses have seen greater technological advances than the methods to evaluate them effectively, which has resulted in over-engineered designs which do not meet the needs of their user. In this thesis , I aim to improve future upper limb prostheses through the development of three design tools. These design tools seek to quantify the functionality of prosthetic devices using motion capture analysis, virtual environments, and joint optimisation. By developing these tools, there is greater opportunity to optimise prostheses earlier in the design cycle which can result in improved functionality. It is anticipated that improvements in functionality will increase user satisfaction and therefore reduce device rejection rates Motion capture analysis was used to study the compensatory movements that arise from operating an upper limb prosthesis. Using a motion capture suit, the motor strategy of a participant was compared between using their biological hand and using a prosthesis through the use of an able-bodied adaptor. It was found that the shoulder and trunk had to make the most compensatory movements to complete several grasping tasks due to the lack of degrees of freedom at the distal end of the prosthesis. Without forearm supination/pronation and wrist extension/flexion, the participant had to approach the grasping tasks from a different angle, sometimes having to lean backwards and abduct their upper arm. The methodology of utilising a motion capture suit as a design tool to quantitatively assess the compensatory movements caused by a prosthetic device was successfully demonstrated. Virtual environments, in conjunction with quantitative grasp quality metrics, can be used to assess the performance of the upper limb prosthesis extremity alone, uninfluenced by user bias. A dynamic virtual environment is presented to simulate several grasping tasks with five upper limb prosthetic devices. Contact information from these grasping tasks are used to calculate the quality of the grasp and provide an overall grasping functionality score. From the simulation results, it was found that more degrees of freedom do not necessary equate to better grasping performance. The positions of force vectors during grasp formation are vital and they must be well- balanced in order to result in stable grasps. Simulated grasping and quantitative analysis in a virtual environment has been demonstrated, which can be used to better plan grasping paths and therefore improve the grasping functionality of upper limb prosthesis designs. Prosthesis users desire their devices to have a low mass, have a low cost, and have high functionality. However, these are conflicting design objectives and decisions must be made to which design considerations to prioritise. A multi-objective model was used to balance these three objectives and select the most suitable components that make up a prosthesis. A modularity scheme was used to divide an upper limb prosthesis into three categories: socket, forearm, and terminal device. In each category, several components were considered which can either be manufactured by conventional engineering or additive manufacturing. Each component would provide a unique value determined by a several quantitative utility functions. Based on satisfaction studies in the literature, the multi-objective optimisation model found that a Split Hook terminal device with an additively manufactured socket and forearm was the optimal design as it provided a low mass and excellent grasping functionality. This model has been demonstrated to work with different user requirements to intelligently select the most appropriate upper limb components within the modularity scheme. Overall, methods were developed which covered aspects of prosthesis design from clinical testing of prosthetic devices, functionality assessments of Computer Aided Design models, and intelligent selection of prosthesis components for individual requirements. It is hoped that these design tools may enable better communication between engineers and clinicians to ensure that users receive devices that are to their satisfaction.

As one of the key technologies of the fifth generation and future wireless communication systems, massive multi-input multi-output (MIMO) has been, and will be, deployed in many practical network scenarios. By increasing the number of antenna elements, the massive MIMO system could significantly improve the spectral efficiency, reliability, coverage, and energy efficiency. Due to the large-scale antenna array, the channel sparsity has been exploited in the literature to solve a number of problems, such as channel estimation, precoder design, and pilot assignment. For the centralized massive MIMO system, the sparsity property lies in users' channel in the angular/beam domain. In the distributed network, the sparsity occurs in the partial connectivity between access points and users due to the signal blockage effect with the obstacles. In practice, however, the existence of such sparsity property is still under debate for sub-6GHz frequency bands with a practical size of antenna array. This makes the developed algorithms that rely on such sparsity not applicable to some other scenarios where channel sparsity may not exist. To deal with this challenge, recently, a novel active channel sparsification (ACS) approach has been proposed for a uniform linear array (ULA) to design sparsifying precoders for frequency-division duplex (FDD) massive MIMO. The proposed ACS approach is able to actively impose the channel sparsity structure in the angular domain with the aid of a partially-connected bipartite graph to represent users' channels. Such a graph representation captures the users' channels through the linear combinations of a set of carefully chosen common basis vectors. Consequently, the precoder design problems can be transformed into graph problems and solved by existing graph algorithms. A question then arises as to whether the ACS approach is limited only to FDD downlink precoder design or it is a more general concept can be applied to a wider range of applications. To answer this question, in this thesis, the concept of ACS is generalized to both centralized and distributed massive MIMO and solve the following problems: pilot decontamination in time-division duplex (TDD) mode; downlink precoding in FDD mode; and pilot assignment for distributed massive MIMO. Pushing forward this line of research, this thesis aims to facilitate the potential deployment of ACS in massive MIMO systems, by extending the ACS from ULA to uniform planar array (UPA), dual polarized-uniform planar array (DP-UPA), and distributed massive MIMO scenarios. The contributions of this thesis are three-fold. 1) For TDD massive MIMO with UPA antenna, the ACS concept is used to solve the pilot contamination problem. A joint beam and user selection approach is proposed that considers channel estimation and sum rate maximization simultaneously. 2) For FDD massive MIMO, the ACS concept is applied to the downlink channel precoder design. The ACS method is generalized from scalar to multi-dimensional to deal with the dual polarization structure of DP-UPA. 3) In distributed massive MIMO, the ACS concept is deployed to resolve the pilot assignment problem. By exploiting the topology structure of the distributed massive MIMO network, this thesis connects the pilot assignment to the topological interference management (TIM) problem that could be solved by off-the-shelf algorithms. Through this thesis, It demonstrates that the ACS is a broad concept, which is able to take multiple factors, e.g., geography, angular, power, and time domains, into account to represent the users' channels by a graph. Thanks to such graph representation, many communication problems with a combinatorial nature can be reformulated as a combinatorial optimization problem. As such low complexity algorithms for solving combinatorial optimization problem could be used in communication problems. In this thesis, it demonstrates the applications of ACS to pilot decontamination in TDD mode, downlink precoder design in FDD mode, and pilot assignment in the distributed setting. It is worth to believe that the ACS concept can find much wider applications including but not limited to beam/user/link scheduling problems in wireless communication networks.

This thesis aimed to investigate two types of decision short-cut, decision avoidance (DA) and heuristic which can save cognitive effort by (partially) skipping or simplifying the decision making process. Firstly, I conducted two studies focusing on whether DA strategies share common antecedents and emotional consequences based on the rational-emotional model. The meta-analysis in chapter 2 indicated that DA strategies could reduce experienced regret. However, since most of the eligible studies were focused on status quo preservation, the identified relationship was primarily driven by this strategy. The results of chapter 3 showed that high personal fear of invalidity (PFI) could promote the use of most DA strategies. This finding offers evidence that DA strategies deserve to be classified as a group of strategies. Then, I explored how changes in the decision environment influence the use of heuristic. The results of chapter 4 suggested that people have a tendency to maintain their decision strategy when the information provision structure changes. However, this inertia effect was weakened when maintaining a strategy requires more cognitive effort in the new decision environment, where people tend to adopt more heuristic strategies. Finally, results of chapter 5 showed that high PFI promotes systematic strategy instead of heuristic. Combining the results of chapter 3 and chapter 5, it could be found that PFI drives people towards and away from using decision short-cuts. High PFI promotes DA in the presence of a choice precedent. However, when the choice precedent or opportunity to evade responsibility is not available, PFI drives people away from using heuristic to achieve a higher expected decision accuracy. These findings imply that the internal drivers for adopting these two types of decision short-cuts are different.

The emergence of contemporary slow world cinemas and their formal aesthetics - the use of long shots and long takes, de-dramatisation and the attentive observation of everydayness - have been read as nostalgic evocations of European cinematic canon of realism and modernism. This Eurocentric type of reading, on the one hand, facilitates the circulation of contemporary slow world cinemas within the global film festival and cinephilic circuits; on the other hand, it also runs the risk of reducing the diverse cultural specificities displayed in a world of representations of cinematic slowness to purely formulaic and teleological universality. In response, this thesis adopts Jiaoyou (Stray Dogs, Tsai Ming-liang, 2013), Japón (Carlos Reygadas, 2002), Yi Yi (A One and a Two ..., Edward Yang, 2000) and La soledad (Solitary Fragments, Jaime Rosales, 2007) as case studies to repurpose slow world cinemas' formal aesthetics as, rather than a universalist strategy, a coincidence of hesitation that leads to a consensus of resistance against the homogenising force of globalisation and the hegemony of capitalist society and neoliberal modernity. This thesis further adopts The Body Remembers When the World Broke Open (Elle-Máijá Tailfeathers and Kathleen Hepburn, 2019) in a coda as a standpoint, or even counterpoint, to interrogate an ethical dilemma that lies in this transformation of coincidental hesitation into consensual resistance. Besides recognising that the theories of Gilles Deleuze, Henri Bergson and André Bazin provide productive readings of cinematic slowness, this thesis also brings into dialogue the existing scholarship on slow cinema, the de-Westernisation of film studies, Transnational Cinema and World Cinema to not only elaborate on but also reconfigure these theorists' established frameworks. Finally, this thesis demonstrates that a world of slow cinemas' resistant tendency, which addresses the detrimental impacts of modern hegemonic social mechanisms in different sociopolitical and geopolitical contexts, can facilitate the dissemination of a localised, vernacular sensibility with the aim of achieving a transnational empathy - an empathy that recognises and interconnects, rather than distills, filters and homogenises, a world of different voices.

The primary objective of my research project was to work with five worldwide dry bulk terminals to assist them to adopt and/or extend Key Performance Indicators (KPIs) to improve their business performance. The project also explores the case for standardisation of KPIs within the wider Dry Bulk Terminals sector. The project aims to answer a number of business and academic research questions. On the business front, these include: does effectiveness dimension improve performance of KPIs; how does organisational culture affect implementation of KPIs; and what role do stakeholders play in planning and implementation of KPIs? On the academic front, these include: how do management style and organisational culture interact during the change process; what role do information systems play in the planning and implementation of KPIs; to what extent were planned PAR stages followed in practice; and challenges in becoming an insider researcher? In answering above questions, the project uses PAR as a research framework through a single research cycle with multiple stages. Representatives from each organisation actively participated during each stage of the project. Even though I was an outsider to participating organisations, I played a leading role and directly participated during each stage of the project. The project makes a novel contribution in a number of areas. First, it has focused on dry bulk terminals rather than ports in general, therefore bringing forward the important role of this industry sector as well as addressing a gap existed in this area in previous research. Secondly, it has introduced the effectiveness dimension into the planning and implementation of KPIs in addition to the efficiency in order to bring in key stakeholders' expectations and views into the picture. Thirdly, if focussed on dry bulk terminals which are significantly more complex and diverse in nature in comparison with container terminals where the majority of previous research concentrated. Fourthly, the project examined a large number of areas where performance KPIs are utilised, led by the participants rather than by a limited or predetermined set of KPIs primarily driven by efficiency parameters. Finally, the project worked with multiple categories stakeholders, driven by the participating terminals rather than limiting those to two or three traditionally recognised stakeholder categories. As a result of adding an effectiveness dimension to performance KPIs, the project made a positive impact at participating terminals in areas such as competitiveness, stakeholder relationships, business continuity and performance. Participating terminals made effective use of elements of their organisational culture and adjusted their management style during the project to ensure success. Stakeholders played a pivotal role in the success of the project by actively participating in it. Information systems played an important role in the implementation of KPIs in coordination with the organisational culture in each participating organisation. The level of standardisation of KPIs in participating terminals was high and this in turn provided a positive platform to explore options for wider standardisation within the industry sector.

There is a lack of UK audit pricing literature to review the long period of audit fees trends and investigate the impact of earnings management. This thesis focuses on UK audit pricing and the data for FTSE All-Share firms from 2001 to 2019 and integrated abnormal accruals into audit pricing models. The thesis aims to ascertain whether auditors will charge higher fees for lower abnormal accruals, which is absent from the related audit pricing literature. First, this thesis's descriptive analyses report the extremely high concentration of Big4 in the FTSE All-Share audit market and as an auditor industry specialist. However, the auditor change frequency is extremely low. The high concentration of Big 4 and rarely auditor change further raised concerns about the quality of the audits. These results have uncovered the development of the UK largest listed firms audit market in recent two decades. Secondly, the empirical analysis utilises the pooled regressions with industry-year fixed-effect and annual regressions. The primary findings of the thesis are that the Big4 premium is not significant and auditor industry specialists charge the fee premium. Also, the initial engagement discounts and audit fee recovery theory exist. Furthermore, strong evidence suggests that lower abnormal accruals lead to higher audit fees. Therefore, this thesis's empirical results indicate that abnormal accruals could be a key audit pricing determinant and a better audit quality measurement, especially for the market with a high concentration of Big 4 clients, like FTSE All-Share firms.

This research work is centred on the selective oxidation of alkanes or cyclic alkanes by using molecular oxygen, and developing supported metal nanoparticles catalysts in a solvent-free system under mild conditions. It aims to design of novel supported metal nanoparticles for the direct oxidation of hydrocarbons to synthesize the corresponding oxygen-containing products like alcohols and ketones, which are valuable precursors for the manufacture of fibres, nylon and their derivatives. To this scope, Ag supported metal particles were developed, and these catalysts were capable of activating O2 and the organic substrate. In particular, a novel supported metal catalyst Ag/Nb2O5 was developed for the oxidation of cyclooctane and cyclohexane, which to the best of our knowledge it is firstly reported, developed and applied for catalytic purposes. A systematic study was conducted to investigate both the catalytic activity of Ag and Nb2O5 in cyclooctane oxidation, as well as the possibility of existence of other active species in the parent Nb2O5 or cyclooctane. This included a systematic evaluation on the effect of traces of water or alkyl hydroperoxides in cyclooctane and characterization of metal and support by means of XPS, XRD and TEM. By using this new catalyst, it was possible to achieve conversion values of 81%, 13% for the substrates, cyclooctane, cyclohexane, respectively. With a total selectivity of ~70%, ~75% to the combination of ketone and alcohol (mainly ketone) generating from these substrates, posing a potential application of this catalyst in the production of oxygen containing products, especially for ketone. Given these promising results, the roles of supported Ag particles and Nb2O5 in the oxidation process were systematically investigated. It was found that Ag species (Ag0, Ag+) could activate molecular oxygen to give reactive superoxide species, or Ag+ was responsible for the direct dissociation of C-H bond, whereas metallic Ag0 was proved to be effective for the abstraction of α-H in the crucial intermediates alkyl hydroperoxide to generate the corresponding ketones as a higher selectivity to ketones was found in comparison with Nb2O5 in the oxidation of cyclooctane and cyclohexane. Nb2O5 was also having an active role in the decomposition of alkyl hydroperoxides. In addition, based on the studies about the different reaction performance of Ag/Nb2O5 prepared by different methods: wet impregnation(WI), deposition precipitation(DP) and sol immobilization(SI), the impregnation protocol was the one that exhibited the highest catalytic activity towards C-H activation, which was probably because the method led to a relatively larger amount of Ag+ species versus Ag0 and correlated to a higher Ag loading. Moreover, in view of these results, a bimetallic supported catalyst with the incorporation of Fe was developed, with the aim to have a metallic partner promoting the initiation of the reaction and a metallic partner promoting the selectivity of the reaction, and therefore in practical terms to have a bifunctional catalyst. In this case supported Ag-Fe/Nb2O5 prepared by wet impregnation method exhibited a superior reactivity with an enhanced selectivity to ketone (~60%) and a higher K/A (ketone to alcohol) molar ratio (3.0) in comparison with supported monometallic Ag/Nb2O5. According to the collected results, a simplified reaction scheme was proposed for the understanding of the catalytic performances of Ag/Nb2O5 to provide insight of the effects of active species in the oxidation process, laying foundation for the further exploitation of this catalyst into the oxidation of other hydrocarbons or alcohols.

Machine learning (ML) is becoming ubiquitous in real-world applications, spanning from our domestic devices such as vacuum cleaner robots, smart phones, virtual assistants, to industrial applications such as manipulators, warehouse robots, to public services such as medical imaging, surveillance cameras, energy grid, etc. With the growing ubiquity and interconnection of these devices, interactions among them will soon become commonplace. These emerging ML problems and the interactions among agents can be formulated as Multiagent ML Systems. On the one hand, many complex ML tasks require the cooperation among multiple agents carrying distributed resources and capabilities. On the other hand, multiagent solutions often lead to improved efficiency, robustness and scalability. Among the many aspects of multiagent ML systems, an important research problem is payoff allocation. This is because multiagent ML systems typically receive a global payoff for the overall performance of all agents, while a fine-grained evaluation of each agent's contribution is absent. Nevertheless, these agent-specific payoffs not only offer natural incentives for their contribution and cooperation, but also provide crucial feedback signals for interpreting the agent's contributions and improving the their future cooperative policies. In this thesis, we investigate the properties of emerging multiagent ML systems and address new challenges that arise when applying classic payoff allocation methods from cooperative game theory to them. The first part of the thesis investigates payoff allocation in submodular multiagent ML problems. Though convex games are commonly studied in cooperative game theory, many multiagent ML applications naturally exhibit submodular characteristics. Moreover, properties of the ML applications give rise to increased computational burden, as well as a new type of malicious attack (i.e., the replication manipulation). To address these challenges, we present a theoretical analysis on payoff allocation methods (in particular, semivalues) in submodular cooperative games, and characterise their robustness against the replication manipulation. We then extend our theoretical results to an emerging application of ML data markets and discuss related and more complex attack models. Moreover, we present a sampling method for estimating the payoff allocation methods to address the computational issue. Finally, we empirically validate our results on a classic facility location problem and on real-life machine learning datasets. In the second part of the thesis, we investigate the integration of the payoff allocation into the learning process for guiding multiagent reinforcement learning (MARL) updates. To this end, we present a framework of game-theoretic payoff allocation in MARL for robotic control. Though the payoffs to agents provide crucial reward signals for improving their cooperative policy, a new challenge arises when performing the payoff allocations during learning. That is, the characteristic values of coalitions are absent except for the grand coalition, i.e., the global reward. On the one hand, it is impossible to re-run the simulation for all possible coalitions at each step. On the other hand, brute-force inferring the value of the unseen coalitions would suffer from highly inaccurate estimations due to high dimensional continuous state and action spaces in the robotic application. To address this challenge, we further incorporate a model-based RL module into our payoff allocation framework, and use a learned model of the environment to simulate values of the unseen coalitions. We empirically demonstrate that our model-based payoff allocation framework greatly improves the performance and sample-efficiency of the MARL system on (multiagent) MuJoCo robotic locomotion control tasks. In conclusion, the results presented in this thesis pave the way towards the application of game-theoretic payoff allocation methods in emerging multiagent ML systems, and opens up a number of exciting directions for future research in this topic.

This thesis studies how mispricing caused by investors' biased expectations of future cash flow affects the cross-section of stock returns. Since the Capital Asset Pricing Model (CAPM), asset pricing models have long been associated with risk-based explanations, and the standard finance paradigm fails to account for investor irrationality. Studying the conditional CAPM model and using cuttingedge machine learning algorithms, this thesis provides novel evidence demonstrating that mispricing from the cash flow channel is systematic and explains characteristic-based anomaly returns. In the Second Chapter, I explore the conditional version of the CAPM on sentiment to provide a behavioural intuition behind the value premium and other anomalies. I find betas and the market risk premium to vary over time across different sentiment indices and portfolios. More importantly, the state beta derived from this sentiment-scaled model provides a behavioural explanation of the value premium and several anomalies. Different from the static beta-return relation that gives a flat security market line, I document upward-sloping security market lines when plotting portfolio returns against their state betas and portfolios with higher state betas earn higher returns. The Third Chapter extends the exploration of the relation between the timevarying sentiment effects and the cross-section of returns by analysing how beta overpricing drives returns of a large set of long-short anomaly strategies. Highbeta stocks are more prone to speculative overpricing (Hong and Sraer; 2016). I find that anomaly-shorts comprise more high-beta stocks, especially during optimistic periods, and thus earn lower returns than longs. Further supporting beta-driven anomalies, the principal component analysis shows that both long and short legs exhibit strong beta comovement. More importantly, beta overpricing gets attenuated in recent years, consistent with the post-publication decay in anomaly returns. I summarise two sources contributing to beta overpricing-investors' biased expectation and leverage constraints and show that these two channels do not contradict each other but co-exist and affect prices of high-beta stocks simultaneously. In the Fourth Chapter, I use classification-based machine-learning methods to decompose 32 anomaly payoffs into risk exposures and mispricing. The component driven by risk earns statistically insignificant returns, despite its efficacy in explaining the time-series variation in anomaly payoffs. The mispricing component is driven by biased cash flow expectations and earns significant returns that subsume anomaly payoffs. These findings indicate that the unconditional averages of anomaly returns can be fully explained by biased expectations, whereas risk exposures play an important role in explaining the time-series variation in anomaly returns. This thesis contributes to both behavioural and asset pricing studies in several ways. First, findings in this thesis demonstrate that sentiment is an important conditional variable to the CAPM and incorporating it into the CAPM improves its ability to explain the cross-sectional variations in stock returns substantially. Moreover, by studying the sentiment-scaled CAPM I provide a behavioural insight to the value premium that the value effect arises from overpricing of growth stocks. Second, I consider overpricing of high-beta stocks as a mispricing-based explanation for a large number of characteristic-based anomalies; the attenuation to beta overpricing seems to drive the post-publication attenuation in anomaly payoffs. Finally, I develop a classification-based machine learning test to disentangle the risk- and mispricing- driven components in anomaly returns. I also introduce the real-time bias in analysts' earning forecasts as proxies for investors' biased expectations to examine the mispricing-based explanation for out-of-sample anomaly returns. My findings suggest that these anomalies are related more to investors' incorrect expectations on future cash flows rather than a reflection of exposures to risk factors.

This thesis presents the exact discrete time representations of first order continuous time models with unequally spaced stocks, flows and mixed data. With unequally spaced data, given that the underlying continuous time models have constant coefficients and homeskedastic disturbances, the exact discrete time representations exhibit more complicated properties such as time-varying coefficients and heteroskedastic moving average disturbances, which arise due to the irregularity in sampling intervals. When data are purely stock variables, the exact discrete time representation follows a VAR(1) process with time-varying coefficients and serially uncorrelated heteroskedastic disturbances. When data are purely flow variables or a mixture of stocks and flows, the exact discrete time representation follows a VARMA(1, 1) process with time-varying coefficients and moving average heteroskedastic disturbances. Based on unequally spaced real life data, the empirical results show that the parameter estimates are different when accounting for the unequal sampling intervals compared to the approach that assumes data are equally spaced. In addition, the Monte Carlo evidences indicate that there are gains to be made in the estimation, such as smaller estimation bias, when the irregular sampling intervals are correctly accounted for.