Search

The field of Data Science has seen rapid growth over the past two decades, with a high demand for people with skills in data analytics, programming, statistics, and ability to visualize, predict from, and otherwise make sense of data. Alongside the rise of various artificial intelligence (AI) and machine learning (ML) applications, we have also witnessed egregious algorithmic biases and harms - from discriminatory outputs of models to reinforcing normative ideals about beauty, gender, race, class, etc. These harms range from high profile cases such as the racial bias embedded in the COMPAS recidivism algorithm, to more insidious cases of algorithmic harm that compound over time with re-traumatizing effects (such as the mental health impacts of recommender systems, social media content organization and the struggle for visibility, and discriminatory content moderation of marginalized individuals). There are various strategies to combat and repair algorithmic harms, ranging from algorithmic audits and fairness metrics to AI Ethics Standards put forth by major institutions and tech companies. However, there is evidence to suggest that current Data Science curricula do not adequately prepare future practitioners to effectively respond to issues of algorithmic harm, "especially" the day-to-day issues that practitioners are likely to face. Through a review of AI Ethics standards and the literature, I devise a set of 9 characterizations of effective AI ethics education: "specific", "prescriptivist", "action-centered", "relatable", "empathetic", "contextual", "expansive", "preventative", and "integrated." The empirical work of this dissertation reveals the value of embedding ethical critique into technical machine learning instruction - demonstrating how teaching AI concepts using cases of algorithmic harm can boost both technical comprehension and ethical considerations [397, 398]. I demonstrate the value of relying on real-world cases and experiences that students already have (such as with hiring/admissions decisions, social media algorithms, or generative AI tools) to boost their learning of both technical and social impact topics. I explore this relationship between personal relatability and experiential learning, demonstrating how to harness students' lived experiences to relate to cases of algorithmic harm and opportunities for repair. My preliminary work also reveals significant "in-group favoritism", suggesting students find AI errors more urgent when they personally relate to them. While this may prove beneficial for engaging underrepresented students in the classroom, it must be paired with empathy-building techniques for students who relate less to cases of algorithmic harm, as well as trauma-informed pedagogical practice. My results also revealed an over-reliance on "life-or-death reasoning" when it came to ethical decision-making, along with organizational and financial pressures that might impede AI professionals from delaying harmful software. This dissertation contributes several strategies to effectively prepare Data Scientists to consider both technical and social aspects of their work, along with empirical results suggesting the benefits of embedded ethics throughout all areas of AI education. [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.]

Polyphenism is the ability of an organism's genotype to produce multiple discrete phenotypes when exposed to environmental stress. Polyphenism is very important to organisms as it allows them to react quickly to changes in environmental stress and therefore increase their chance of survival in the changing environment. There are many different types of polyphenism, including phase polyphenism, wing polyphenism and others. Aphids are one of the key model organisms used in studying polyphenism. Under normal conditions, aphids reproduce asexually and all the offsprings are essentially clones of the mother. However, upon exposure to environmental stress such as crowding, predation, and depletion of food resources, aphids exhibit wing polyphenism whereby the offspring from the next generation will develop wings, which allows them to escape from poor environmental conditions. Even though the aphid model has been used to investigate polyphenism, the underlying mechanisms in regulating such wing polyphenism remain unclear. The main experiment in this thesis involved exposing two different aphid genotypes (N116 and N127) to crowding conditions to trigger alternative morph production and the degree and number of adult alternative morphs were recorded (mesocosm experiment). Next, we carried out two transfer experiments to investigate the possibility of trade-offs between dispersal and reproduction. At the end of this mesocosm experiment, five 4th instar nymphs of each morph were transferred to individual plants to allow them to reproduce for another 12 days (Transfer experiment 1). The total number of offspring produced by each morph was recorded. At the end of transfer experiment 1, five 4th instar nymphs produced by each morph were transferred to another individual plant and allowed to reproduce for another 12 days and the total number of the population was recorded. The adult morphs collected from the mesocosm experiment were used to investigate the underlying mechanisms of wing polyphenism in aphids and the role of epigenetics mechanisms in regulating pea aphid wing polyphenisms through different techniques: qPCR, RNA-seq, MBD-seq, pyrosequencing and mIRNA-seq. 25 The two aphid genotypes reacted very differently to starvation conditions with N116 producing winged offspring in the next generation while N127 changed their body colour from red to pale. In addition, N127 also produced a higher percentage of alternative morphs compared to N116. We found that in both transfer experiments the dispersal morphs have lower reproduction compared to the wild type. This suggests a possible trade-off between dispersal and reproduction and such a trade-off could last for more than one generation. Our RNA-seq and qPCR identify several genes that could be involved in regulating wing development, reproduction and stress response in the pea aphid. Pyrosequencing revealed that among the genes that were differentially expressed between morphs only a few were differentially methylated. Further, our MBD-seq shows that the pea aphid has the highest methylation in the gene body region. However, integrating the MBD-seq and RNA-seq suggests that there was no clear correlation between DNA methylation and gene expression but some genotypes and morphs show stronger positive trends. In addition, we found that many conserved and some novel miRNAs were differentially expressed between the aphid morphs. Target gene prediction results from these miRNAs suggest that they target many genes that were involved in wing development, reproduction and wing development. Together the results from this thesis provide insight into genes that could be regulating wing polyphenism, stress response and reproduction in pea aphids and the role of different epigenetic mechanisms in regulating this polyphenism.

My dissertation consists of three chapters in the field of labor and education economics. The first chapter studies the impact of early morning classes on students' educational trajectories. The second chapter explores how college course shutouts affects students' educational outcomes. The third chapter investigates the impact of kindness on wage returns. In the first chapter, I examine how early morning classes affect students' educational trajectories by exploiting a natural experiment which randomized class time to students. I find that enrolling in early morning classes lowers students' course grades and the likelihood of future STEM course enrollment. Early morning classes also cause a 79% reduction that a student study in the corresponding major. To understand the mechanism, I conducted a survey of undergraduate students enrolled in an introductory course, some of whom were assigned to a 7:30 AM section. I find evidence of a decrease in human capital accumulation and learning quality for early morning sections. The second chapter is written with Rich W. Patterson and Kevin J. Mumford. We explore what happens when college students do not get the courses they want. We use a natural experiment at a large public university in which freshman students are conditionally randomly assigned enrollment in over-subscribed courses. We find that first-semester freshman students who are shut out from an over-subscribed course are approximately half as likely to ever take the oversubscribed course and approximately 43% less likely to ever take a course in the same subject. Students who are shut out of a course are also 4% (0.05 percentage points) less likely to major in a corresponding subject. The effects of shutouts on major choice go beyond shifting students to closely related majors. The third chapter explores the impact of kindness on wage returns across five occupation classes: unskilled, partly skilled, skilled, managerial, and professional. Using data from the 1958 National Child Development Study (NCDS), the study reveals that kind women are rewarded while kind men face penalties in unskilled occupations. Gender norm bias and workplace cultures that undervalue interpersonal skills may explain this disparity. However, both kind men and women in managerial roles receive rewards, with men enjoying higher returns. This may be attributed to the greater need for interpersonal communication skills in managerial positions, where kindness reduces communication costs and enhances productivity. These findings highlight the significance of kindness, gender bias, and interpersonal dynamics in wage outcomes. [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.]

Radio pulsars are rotating neutron stars (NSs) identified by a periodic detection of radio emission with intervals of seconds to milliseconds. The periodicity of the emission is closely tied to the NS's spin frequency such that timing a pulsar allows its spin frequency to be accurately determined. The spin frequency is generally predictable except for the case of two timing irregularities: glitches and timing noise. Both represent phenomena that occur on different time-scales with glitches affecting short-term evolution and timing noise affecting long term evolution. Glitches are sudden increases in the spin frequency which is sometimes followed by a post-glitch recovery. Timing noise is the term used for the residuals left over after the predictable spin evolution is subtracted from observations. There are existing models that can explain glitches, their recovery as well as timing noise, but in this thesis, we focus on creating two more, one for the post-glitch recovery and another for timing noise. The key aspect of these models is their connection to gravitational waves (GWs), which is an idea not well explored. In the first part of this thesis, we create the "transient mountain" model which can explain post-glitch recoveries and make falsifiable predictions for the GWs that come off. This is based off the idea that a NS "mountain" would create an extra braking torque on the NS, hence, spinning it down during the post-glitch recovery. For the second part of this thesis, we create a model for timing noise which we propose is caused by successive (micro)glitches which cause the NS to oscillate. These glitches excite the f-modes on the NS which are known to emit GWs. In modelling these two types of timing irregularities, we hope to strengthen the bridge between theory, radio observations and GW observations.

The effectiveness and feasibility of psychological interventions for populations under ongoing threat: A systematic review: Current views of psychological therapies for trauma typically assume the traumatic event to be in the past. Yet individuals who live in ongoing organised violence or experience intimate partner violence (IPV) may continue to be (re)exposed to related traumatic events or have realistic fears of their recurrence. This review considers the effectiveness, feasibility and the requirement for adaptations of psychological interventions for individuals living with ongoing threat. Eighteen papers featuring 15 trials were included (12 on organised conflict, three on ongoing IPV). For political violence, most studies showed moderate to large effects in reducing trauma-related symptoms when compared to waitlists. For IPV, findings were varied. Most studies found providing psychological interventions were feasible. The findings, albeit preliminary with mixed methodological quality, showed psychological treatments can be beneficial and should not be withheld in context of ongoing organised conflicts and IPV. Clinical and research recommendations are outlined. Patients' experiences of clinical team meetings (ward round) at an adult inpatient eating disorders ward: a mixed-method service improvement project: Aims: Ward rounds are key to treatment-related decision making but are often stressful for patients. This project aimed to explore and improve patients' experiences of the clinical team meeting (CTM, historically known as ward round) in an adult inpatient eating disorders unit. A mixed-method approach was adopted using in-vivo observations, two focus groups, an interview, and a staff survey. Six patients participated. Two former patients contributed to data analysis, co-production of service improvement initiatives and write-up. Results: The mean CTM duration was 14.3 minutes. Patients spoke half of the time, followed by psychiatry colleagues. 'Request' was the most-spoken category. Thematic analysis identified (1) CTMs are important but impersonal, (2) CTMs generate a sense of palpable anxiety; (3) divergent views between staff and patients. Clinical implications: The findings and implementation of co-produced changes to CTMs improved patients' experiences despite COVID-19 challenges. Factors beyond CTMs including the ward's power hierarchy, culture and language need addressing to facilitate shared decision-making. Sudden Gains in internet cognitive therapy for social anxiety disorder in routine clinical practice: Background: Research on sudden symptom improvements (sudden gains) in internet-delivered psychological interventions are limited in routine practice. Objective This was a study on sudden gain and its relationship to outcome and cognitive processes. The intervention examined was a therapist-guided internet Cognitive Therapy intervention for social anxiety disorder (iCT-SAD) delivered in primary mental health services in the UK. The aim was to replicate the approach of a previous exploratory study (Thew et al., submitted) where the data were drawn from a randomised controlled trial (Clark et al., 2022). Methods: The study used data from a larger implementation study (N = 193). Liebowitz Social Anxiety Scale was the primary outcome measure. Process variables examined included negative social cognition measured by the Social Cognition Questionnaire, self-focused attention (Social Phobia Weekly Summary Scale) and depressed mood (Patient Health Questionnaire). Results: Seventy sudden gains were found among 57 participants. The occurrence rate of sudden gain was 39%. Individuals who experienced sudden gains had a larger reduction in social anxiety symptoms at end of intervention and at three-month follow-up. There was evidence of a reduction in the frequency of negative social cognitions prior to the gain, whereas change in self-focused attention occurred simultaneously with the gain. Depressed mood did not show significant changes over these timepoints. Conclusions: Sudden gain was found in routine practice and was found to be related to meaningful symptom improvement.

We consider a smooth function f : M → R on a Riemannian submanifold M embedded in an ambient Euclidean space Rd. We provide theoretical guarantees for inferring the Morse index of a compact, dynamically isolated critical set of f using a pair of finite point clouds X and X− in Rd, and detail sufficient conditions which X and X− have to satisfy. We also outline a realisable computational procedure for obtaining X and X− and inferring the Morse index. In a separate study with Jacob Leygonie, we discuss how persistent homology and wavelet signatures can be applied to graph classification problems in machine learning. We construct a supervised learning model which backpropagates through persistent homology to optimise over the space of wavelet signatures. We then apply this model to several benchmark graph classification problems and obtain accuracies in line with state of the art models. The final portion of this thesis investigates how maximum diversity, a notion of effective cardinality in geometric measure theory, can be applied to industrial applications of interest to Elsevier, who sponsor this DPhil. In particular, we apply maximum diversity to assess the multi-disciplinarity of a collection of EPSRC grants, which takes relations between topics and grants into account. We also derive a generalised notion of Jaccard similarity such that we can measure the topical similarity between two grants, leveraging the pairwise similarities between their associated topics. We give recommendations on further industrial applications of maximum diversity in Elsevier's business.

In the field of paediatric neuro-oncology, the positions of tumours within the central nervous system of the patients makes the acquisition of solid tumour biopsies risky. For many tumour types, monitoring of treatment response is restricted to Magnetic Resonance Imaging (MRI), or cerebrospinal fluid (CSF) cytology in the cases with leptomeningeal dissemination. Both of these lack sensitivity, leaving room for improvement. Recent advances in molecular barcoding sensitivity and error suppression have made the sequencing of DNA derived from liquid biopsies possible. Liquid biopsies offer an alternative to solid biopsies, since the collection of bodily fluids is much less invasive by comparison, and liquid biopsies contain cell-free DNA (cfDNA). In cancer patients, it has been shown that a fraction of the cfDNA in multiple liquid biopsies, such as plasma and CSF, harbour the genetic alterations present within the tumour. This circulating tumour DNA (ctDNA) can be used as a biomarker for diagnosis, stratification, and surveillance of the tumour. The monitoring of treatment response, and the detection of minimal residual disease, is of particular importance in paediatric brain tumours, given the low sensitivity of existing methods. This project created a versatile system, utilising molecular barcoding, which was able to detect Single Nucleotide Variants (SNVs), Insertions/Deletions and Copy-Number Variants in a single assay. A wet-lab workflow was created and iteratively improved, such that it could handle a diverse range of liquid biopsy types, including plasma, cystic fluid and CSF. This workflow was coupled with a bioinformatic pipeline, designed to process the data for all three variant calling processes simultaneously. For SNV calling, a custom variant caller was created to aid in the suppression of errors in barcoded sequencing, and the system was used in the first documented tracking of Adamantinomatous Craniopharyngioma treatment response using cystic fluid liquid biopsies.

This study aims to re-conceptualise volunteer tourism through wider theoretical questions and analytical approach in order to contribute to developing a more holistic theoretical framework for understanding volunteer tourism. This thesis is premised primarily on the question: 'why do individuals travel as volunteers?' Focusing on volunteer tourists from Hong Kong and Taiwan, this thesis seeks to unpack how volunteer tourism becomes a social trend. As a choice of travelling, taking part in volunteer tourism is not just a discrete choice of a single volunteer tourist. It is a prevailing choice from the overarching, complex ethical terrain in which an individual chooses to do things from a particular instilled position. This instilled position emerges from the interplay of neoliberal constructions and responsibilisation through institutionalised service obligation in schools that encourage students to extend care to distant others alongside the culturally rooted ethical dispositions that inform developing oneself by taking up more responsibilities. This social trend also manifests the search for rebuilding social ties that can be actualised through being part of moral communities and new ways of being through distancing from the 'familiar' and critical self-reflection in the 'unfamiliar'. This moves the conceptualisation of volunteer tourism beyond the absolute dichotomy between altruism and self-interest to the broader discussion around ways of being and becoming a moral self. This study also offers a 'post-normative' analytical framework to understand how a good self is cultivated, experienced and re-invented through volunteer tourism. On the one hand, it draws together the concepts to examine the active and conscious technologies of the self, unconscious dispositions and the notion of quality which play out for the development of the self towards an ethical subject. On the other hand, it links up the concepts of community, liminality and geographies of care to investigate the meaning and formation of the moral self in a collective setting. This study further contributes to this theoretical framework by widening the discussion to a non-Western context. The finding of this research shows that in Chinese societies in East Asia, developing a good self is the choice as well as the outcome, which is slightly different from the Western responsibilisation. This study contributes to the current responsibilisation as well as alternative tourism or ethical tourism literature by looking at the complex ethical and cultural framework in which individuals choose to do things from particular dispositions rather than positing individuals to make discrete choices. The way that these Asian volunteers conceive volunteer tourism, as informed by their traditional values, helps to re-orientate from sending aid to change poor people's lives to sharing and exchange which may also help to disorientate volunteers' subjectivities. This responds to the urge for 'Asianising the field' through more research on the emerging phenomenon of Asian tourists/tourism within Asia.

Ubiquitination is a post-translational modification involving the attachment of ubiquitin molecules onto target substrates. The most common function of ubiquitination is proteasomal degradation, which is commonly facilitated by the ubiquitination of target proteins at Lys-48. E3 ligases are crucial to the process of ubiquitination, and include WWP2 ubiquitin Ligase, an E3 enzyme that has been shown to interact with Smad mediators of the TGF family and might subsequently have a role in TGFβ-dependent oncogenic processes such as EMT. To elucidate the function of WWP2 in TGFβ-signalling and increase our understanding of this protein, is therefore the forefront of our research. Here, we found that the overexpression of WWP2 WW3-4 recognition domain led to an increase in Smad3-dependent TGF gene expression, providing strong evidence that in the absence of TGF , WW3-4 binds to the Smad3 mediator. We also predicted and validated the expression of two novel WWP2 isoforms, WWP2-ΔHECT and WWP2-N-ΔC2. Results suggest that the incomplete HECT domain within WWP2-ΔHECT is catalytically non-functional. Furthermore, the transcription of WWP2-ΔHECT has been linked to a 2kb region within intron 9/10 with its expression positively regulated by EGF stimulation. On the contrary, when the transcriptional mechanism of the existing WWP2-C isoform was investigated, a 0.5kb region at the end of intron 10/11 was identified to enhance promoter activity responsible for expression, which was negatively regulated by the SOX9 transcription factor. It was also found that ESRP splicing factor is involved in the negative regulation of not only the existing isoform WWP2-N, but also the novel WWP2-ΔHECT isoform. Using expression studies, we found that WWP2-N and ΔHECT were expressed at higher levels in epithelial cells, therefore suggesting their role as guardians of the epithelial phenotype during EMT. Overall, the data provided here can help clarify the role of WWP2 in TGFβ-dependent oncogenic signalling.

In Hong Kong, little is known on the role of government retraining in helping older and low-skilled workers cope with unemployment. The aim of this study is to explore the relationship between Hong Kong's government retraining and psychological well-being among unemployed older and low-skilled trainees, and the effects of demographic factor (age and gender) on coping mechanisms. The results are based on qualitative interviews with 15 Chinese re-employed security guards aged 50+ after participation in the government retraining programme. The results showed that participants used both problem-focused coping (Hong Kong's government retraining) and emotion-focused coping (social support, escape, relaxation and leisure activities). Such retraining helped participants to meet their economic and psychosocial needs, and develop jobs skills for better well-being. Participants tended to be self-reliant and to rely on financial support from family members to save face. The results also indicated that men were less likely to use social support than women to share their emotional distress with others, and relied more on relaxation and leisure activities. The lower educated tended to take part in less social leisure activities than the better educated to regulate stress, and preferred relaxation activities at home. Following re-employment, the better educated felt more underemployed than the lower educated, whereas those in former high-status jobs used higher-level self-categorisation than those in former low-status jobs as a way to escape from their low-status security job. Overall, the findings show that features from the theories of Jahoda (1982), Fryer (1986) and Warr (1987) relate to well-being, and demographic factors (age, gender and education) and cultural factors (face-saving and self-reliance) affect coping mechanisms. These findings implicate the need to adjust the existing theoretical framework to be appropriate for Hong Kong Chinese society and to offer practical implications for policy and practice in the areas of retraining policy, gendered labour market, unemployment benefits, age-friendly working environments, retraining programmes and career counselling service aiming to enhance well-being and facilitate work re-entry of older workers.

This thesis explores Robert Schumann's influence on the two very different versions of Johannes Brahms' Piano Trio in B major, Op. 8 – the first version of 1854 and the revised version of 1889 (or Op. 8a and Op. 8b, respectively) – by establishing a biographical and compositional relationship with Schumann's Piano Trio in D minor, Op. 63 (1847), with particular reference to his writings. This comparative study firstly examines Schumann's important contribution to the piano trio genre and the musical aesthetics inherent in his work. His impact on the genesis of Brahms' Op. 8a beyond the biographical is considered in depth here for the first time in a scholarly study. By identifying Schumann's compositional methods from his middle-late period and his musical aesthetics as manifested in relevant compositions, my hypothesis argues that the young Brahms modeled his Op. 8a on Schumann's Op. 63; it also suggests a parallel performing tradition between the two works. In contrast to the continuing musicological enquiries concerning the two versions of Op. 8, performers have paid little attention to Op. 8a owing to a number of misconceptions. The reconstruction of a performing tradition of the first version did not get under way until the 1980s, when the first recording appeared. Although it would seem reasonable to seek clues as to its interpretation from the performing tradition of Brahms' Op. 8b, the more widely known version, my hypothesis instead seeks to align Brahms' Op. 8a with Schumann's seminal work in the same genre, while also reframing Op. 8b in this context. The notion of the 'Schumannesque' with respect to the composer's middle-late period will be examined, alongside a set of musical aesthetics based on Schumann's early Davidsbündler ideology. The ways in which these issues relate to Brahms' Op. 8a are reflected in my proposed performance guidelines for the work, which are substantiated through methods of practice-based research.

The aim of this thesis was to contribute to the development of an objective, evidence-based international classification system for para-swimmers by quantifying the effect of physical impairment on passive and active drag. The thesis comprises five studies. Study 1 identified a significant relationship between normalised passive drag and the para-swimmers’ International Paralympic Committee (IPC) Class, but an inconsistent difference in normalised passive drag between adjacent classes. High within-class variability in passive drag indicates that the current classification system does not always differentiate clearly between swimming groups. Study 2 found that anthropometric features of para-swimmers, such as height and body mass, differed significantly between IPC Classes, whereas Shoulder Width, Chest Depth, Shoulder Girth and Torso Girth did not. A weak correlation existed between para-swimmers’ anthropometry and their passive drag, which indicates that other factors, such as impairment type, may be more important predictors of passive drag than anthropometry. Study 3 revealed that certain impairments, such as double-leg amputation above knee level, may predispose a para-swimmer to a relatively high passive drag which disadvantages them in competition. Study 4 compared two methods of estimating active drag during front crawl swimming: the Naval Architecture Based Approach (NABA) and the Active Towing Method (ATM). The means were not statistically different. Using a sensitivity analysis, the NABA was identified as the more reliable method of assessing active drag. Study 5 found that active and passive drag of elite para-swimmers are highly correlated but no relationship existed between active drag and International Paralympic Committee S Class (IPC S Class), indicating that factors other than impairment level may be more important in determining active drag. The relationships discovered between drag, IPC Class, anthropometry and impairments will contribute to the development of the future IPC Classification system.

The current thesis aims to explore issues regarding the development and support of second languages when two individuals from different linguistic backgrounds come together in romantic, intimate, and meaningful relationship. Based on a thorough conceptualisation of multilingual couplehood and its place within the interdisciplinary topic of intercultural relationships and marriage, this exploratory study seeks to highlight the issues pertaining to communication, and even more precisely, language, how this impacts individual and couple identity, and its significance in multilingual couple contexts The study and presentation of the research is built on the foundation of a relational philosophy of being and knowing, particularly in line with Gergen’s (2009) understanding of ‘being-in-relation’. Arguments are presented for a relational ontology and epistemology in hopes of identifying the merits of this perspective, and to promote what I consider a more fitting understanding of identity, relationships, and second language education. On this theoretical basis, the study adopts the qualitative methodological approach of narrative inquiry to investigate the meaning of language, relationship, and identity in the contexts of four self-identified multilingual couples currently living in the UK. Data was collected by means of three stage interviews: one with each individual of the couple and a final interview with both partners present. The audio-recorded interviews were transcribed and subjected to a two-stage narrative analysis. The findings have uncovered a wide variety of issues regarding intimate relationships and matters related to the development and support of second languages. While it is clear that each couple has differing attitudes and perceptions of multilingual couplehood, and ways of addressing them, there are some commonalities that have emerged, particularly in regards to the negotiation and sharing of languages, which has appeared to be a very complex issue in terms of the manifestation of language support and learning motivation within the couplehood. It has also been found that there are various ways in which partners develop and create their multilingual couple identities, which affects and also is affected by the linguistic and cultural identities of the individuals themselves. It is clear that second language development and support is a fundamentally relational process in the intimate context of couple relationships, as the detailed narratives of the couples and their linguistic journeys are highly revealing of what is clearly a complex, intricate, and relational ongoing process.

In this thesis, surface science techniques were employed to study the chemistry of rutile TiO2(110). Scanning tunnelling microscopy (STM) and ultraviolet photoemission spectroscopy (UPS) have been used to determine the origin of the band-gap state in rutile TiO2(110). By employing electron bombardment to vary the Ob-vac density while monitoring the band-gap state with UPS, we demonstrate that Ob-vac make dominant contribution to the photoemission peak and that is magnitude is directly proportional to the Ob-vac density. CO adsorption on the Pd/TiO2(110) surface was investigated with synchrotron radiation spectroscopies and STM. The Pd islands, which were grown by physical vapour deposition (PVD) of Pd onto the TiO2(110) substrate at ~800 K, had a pseudo-hexagonal shape and were not encapsulated with Ti^n+ (n<4) species from the substrate. In addition, it was found that CO molecules bond vertically and form various ordered overlayers on the Pd(111) islands. O2 adsorption on the cross-linked TiO2(110)-(1x2) surface was investigated with XPS, UPS and STM. The introduction of a small amount of O2 leads to a drastic reduction in the number of the Ti3+ species at the topmost surface layers and the band-gap state intensity, as well as a noticeable rise in the surface workfunction. In STM, O2 and its related molecules were found to preferably adsorb at the centre of the (1x2) strands. Current imaging tunnelling spectroscopy (CITS) was also performed on the same surface at 78 K. It was found that the densities of the two occupied states, one at -0.7 V and and another at -1.3 V, vary between different features on the surface. Moreover, whilst having less-populated occupied states, the cross-links possess an empty state at 1.2 V which cannot be detected anywhere else. This work will be compared with theoretical calculations to elucidate the geometric structure of the cross-link TiO2(110)-(1x2) surface.

This thesis examines two approaches to Galois correspondences in formal logic. A standard result of classical first-order model theory is the observation that models of L-theories with a weak form of elimination of imaginaries hold a correspondence between their substructures and automorphism groups defined on them. This work applies the resultant framework to explore the practical consequences of a model-theoretic Galois theory with respect to certain first-order L-theories. The framework is also used to motivate an examination of its underlying model-theoretic foundations. The model-theoretic Galois theory of pure fields and valued fields is compared to the algebraic Galois theory of pure and valued fields to point out differences that may hold between them. The framework of this logical Galois correspondence is also applied to the theory of pseudoexponentiation to obtain a sketch of the Galois theory of exponential fields, where the fixed substructure of the complex pseudoexponential field B is an exponential field with the field Qrab as its algebraic subfield. This work obtains a partial exponential analogue to the Kronecker-Weber theorem by describing the pure field-theoretic abelian extensions of Qrab, expanding upon work in the twelfth of Hilbert’s problems. This result is then used to determine some of the model-theoretic abelian extensions of the fixed substructure of B. This work also incorporates the principles required of this model-theoretic framework in order to develop a model theory over substructural logics which is capable of expressing this Galois correspondence. A formal semantics is developed for quantified predicate substructural logics based on algebraic models for their propositional or nonquantified fragments. This semantics is then used to develop substructural forms of standard results in classical first-order model theory. This work then uses this substructural model theory to demonstrate the Galois correspondence that substructural first-order theories can carry in certain situations.

Within the context of parent education, the existing parenting programmes in Hong Kong are either drawn from contexts that are less sensitive to the local one or imbued with Confucian and High Context culture (HC) values of social hierarchy and group harmony. Within such cultural values educators and learners are expected to behave in adherence to the expected social role wherein the educator is the knowledge transmitter whereas the learner is the knowledge absorber or the approaches to parenting that permeate the programmes are inappropriate to the context. The impracticality and cultural insensitivity of these existing parent education programmes, together with news related to adaptation of both new arrivals and Hong Kong locals in terms of adapting to the increasingly diverse sociocultural population in Hong Kong, alert the parent education field to a dire need to develop a model for parent education programmes that caters for different family unique needs within the diverse sociocultural context. With an aim to explore the essential components that are needs and cultural sensitive, the study adopted narrative inquiry as the research methodology to explore the three research questions: (I) To what extent do sociocultural factors affect group interaction among members from diverse backgrounds in Hong Kong? (2) What mechanisms affect group interaction among people from diverse backgrounds in parent education programmes in Hong Kong? (3) How can one formulate a model of parent education programme to suit families having diverse backgrounds, values, and beliefs in the Hong Kong context? The findings of these three research questions emphasise the vital role played by the group facilitator in creating a mutual supportive and need sensitive context to engage learners in mutual learning. Also, a critical analysis of the Subjective Group Dynamics model (SGD) is discussed in the light of group formation among members with diverse sociocultural background within the Confucian and High Context culture (HC) oriented context culture in Hong Kong. The findings of the research questions and analysis of SOD build the foundation for the Family-centred Mutual Support Model of parent education programmes in Hong Kong that emerged from this study.

A study examined the role of grammar instruction in second language (L2) learning by investigating: (1) whether L2 learners significantly improved their language skills after formal grammar instruction, (2) to what extent grammatical knowledge predicted a learner's next level of overall language proficiency, (3) to what extent grammatical knowledge is related to the four language skills (listening, speaking, reading, writing), and (4) to what degree learners at different levels of overall language proficiency differ in the way they benefit from grammar instruction. Subjects were 242 young adult students enrolled in a university English-as-a-Second-Language (ESL) program. Results suggest that the students improved significantly after formal instruction, and that grammatical knowledge is a significant predictor of the students' readiness for the next course level. In addition, grammatical knowledge seems to be a significant component in mastering the four language skills, and students at the beginning level of overall language proficiency gain more on the overall test score than do students at the intermediate or advanced level. However, the grammar competence level and interaction between the overall language proficiency level and grammar competence level could not explain the gain in overall test score. (Author/MSE)

The present study aimed to explore how the students and the teachers in Hong Kong conceptualised “effective teaching”. 106 Form three students and 3 teachers were asked to rank the significance of forty teaching traits and evaluate the teachers’ teaching performance. A focus group interview and individual interviews were conducted with the students and the teachers respectively. Classroom observations together with follow-up talks were also arranged for the teachers. It was found that both the participating groups had similar though not identical views on the effective aspects of teaching. They generally thought that an effective teacher should use various strategies to motivate students, involve all of the students in the lesson, use a variety of activities/ learning methods, give clear instructions and explanations, and keep students on task throughout the lesson. They also thought that an effective teacher should have a good relationship with his/ her students and make the best use of reinforcements. They agreed with most of the reviewed literature about teaching effectiveness except that they preferred the teacher-centred approaches to the student-centred approaches. This could be mainly due to the cultural differences. Although the teachers claimed that students should take care of their own learning, they did not actively try to encourage their students to do so in reality. For students, a majority group tended to be passive and rely a lot on their teachers while a minority group tended to be more proactive and learn more independently. Since the western theories might not be compatible with the eastern beliefs, it was suggested that teachers should test the theories before use. This could be encouraged by a more bottom-up educational system. It was also suggested that students and teachers should have a more active role to play in teacher evaluation. They should be given more chance to talk and share their perceptions of effective teaching. It is argued that this sharing of perceptions would enable greater consensus to emerge between teachers and students. Last but not least, it would be sensible for teachers to treat teaching as a combination of a science, a craft and an art so that they were better able to meet the expectations of different types of students.

Using a range of complementary methods (miscue analysis, interviews, and questionnaires), this thesis aims to explore the reading processes of some Korean primary school students in English. The questionnaire was collected from 78 Korean primary school students in Seoul. Of those students, 12 were selected to provide miscue data. Interview data was collected from the miscue participants as well as six adult participants (three state school teachers, one private lesson teacher, one parent, one member of staff at a children's bookshop). Established miscue analysis techniques were adapted to accommodate the research context of Korean primary school students; participants were asked to read the same text twice without any significant pause, and, after the second oral reading, were asked to translate what they read in English into Korean. Data analysis was carried out quantitatively as well as qualitatively. The number of miscues were counted and categorised into five types: substitutions, insertions, omissions, repetitions, and reformulations. In the qualitative analysis, translation and interview data were used to provide more information about reading processes. The analysis shows that more than thirty percent of miscues, mostly substitution miscues, were repeated across the two oral readings. They also show that a lot of non-word substitutions were produced and that most of the real-word substitutions produced were based on graphophonic cues, rather than syntactic or semantic cues. The translation data suggest that many students experience difficulty in using syntactic cues. The mismatches between miscue and translation data show that the students' difficulties in decoding do not necessarily reflect difficulties in reading comprehension, or vice versa. The analysis of the interview transcripts show that many students consider reading to be an oral activity and associate reading difficulties mainly with pronunciation and vocabulary. The results suggest that they approach reading in English in a disconnected way, failing to integrate cues from different levels. The discussion suggests that miscue analysis can perform a helpful role in allowing researchers to gain a greater understanding of readers' expectations of reading. It is also suggested, however, that Goodman's 'window' metaphor for miscue analysis may be misleading and that miscue data should be used with a degree of caution, without the assumption that it can serve as a transparency that will reveal all aspects of reading processes.

The obligate intracellular parasite Plasmodium falciparum causes the most severe forms of human malaria. Apical rhoptry organelles of invasive merozoites contain proteins that are instrumental to the intraerythrocytic development of the parasite. Our group has determined that the RhopHl protein of the high molecular mass rhoptry complex (RhopH) is encoded by genes of the highly-conserved clag (cytoadherence-linked asexual gene) family. In this study, the characterisation of the RhopH 1/Clag proteins is continued. It was ascertained that the family has five members in P. falciparum 3D7 (clag2, -3.1, -3.2, -£and -9), each of which is transcribed. Specific antibodies were raised to demonstrate that all Clags are expressed at the apical end of merozoites, but in an apparently differential distribution. Both Clag3.1 and Clag9 are found in the basal bulb of the rhoptries as part of the RhopH complex. However, Clag2, -3.2 and -8 appear to remain outside the complex and their distribution is distinctly more apical, potentially in the region of the rhoptry neck. It was determined that RhopH complexes contain only individual Clag members in a mutually exclusive association, thereby suggesting the existence of multiple complexes in which specificity is conferred by RhopH 1/Clag. Those Clag proteins that are part of the RhopH complex are carried into the newly invaded red blood cell where they persist. There is evidence to suggest that Clag3.1 and Clag9 are trafficked across the parasitophorous vacuolar membrane, and that their function is in the developing parasite. Clag9 is the most diverse member of the family and was originally proposed to be involved in the pathogenic process of cytoadhesion from the surface of the infected red blood cell. However, we have now identified it as a rhoptry protein. This may suggest that Clag9 is indirectly involved in cytoadherence within the cycle that follows erythrocyte invasion.

The thesis investigates the role of parent managers in school management committees in Hong Kong and explores the policy development and intentions of the Government of the Hong Kong Special Administrative Region of the People's Republic of China in introducing the reform to enhance parent participation. This study focuses on the impact of the reform and the match/mismatch between the reform aims and the expectations of the parents. It comprises a survey of the role of parent managers in school management committees in government and aided schools and interviews with principals and parent managers in three selected schools. The findings show that the reform has had little impact on the aided schools and there is strong opposition from the sponsoring bodies. There is also a mismatch between government aims and parent expectations. The government considers the most effective form of parent participation to be in school governance, while parents care mainly about their children's learning. Parents lack the knowledge, skills and the time to perform the monitoring role and hold schools accountable. They have little influence in decision-making and they have strong trust in the school heads. Partnership has not developed as parents are regarded as unequal partners and they are willing to perform roles assigned to them by the heads. Parents' attitude towards participation in decision-making is found to be more positive and they can even perform a better job than the principals in explaining school policies to other parents and gaining parent support. The findings suggest that the government should re-establish the harmonious working relations with the sponsoring bodies and match government aims with parent expectations. Schools should develop shared goals with parents and identify issues that require their active participation.

Multilayers of STO/YBCO thin films on Ce02/YSZ/CeO2 buffered curved Nibased rolling-assisted biaxially textured substrates (RABiTS) have been grown successfully by pulsed laser deposition in order to demonstrate the feasibility of fabricating an in situ multi-turn superconducting coil on a cylindrical heated former by continuous deposition of sequential layers of superconductor and insulator. Such an approach would offer a radical alternative to the conventional winding of coils from pre-formed long lengths of wires and tapes. AC susceptibility, X-ray diffraction (θ - 20 ω - scans, ψ - scans), scanning electron microscopy (SEM) and electron backscattered diffraction (EBSD) were used to characterise the superconducting, crystallinity, morphology and microtextural properties of the grown samples. The microstructure and chemical analysis of the interfaces of the curved samples were further investigated by cross-sectional transmission electron microscopy (XTEM) and scanning transmission electron microscopy (STEM). The superconducting and crystallographic properties of the curved composite were similar to those achieved on flat surfaces. The optimum thickness of each YBCO layer in the multilayer architecture was below I gm on a buffered curved surface. In the multilayer structure of six YBCO layers insulated by six STO layers (STO/YBCO)6 on a buffered curved surface, biaxial texture can be still transmitted from the substrate to all the layers and all the YBCO layers had c-axis orientation. Moreover, no interfacial diffusion or chemical reaction was observed at all the interfaces of the multilayer. However, defects such as pores and outgrowths were commonly found which affected the superconducting properties of the curved samples.

Two-liquid phase dispersion in mechanically stirred vessels is a key step in many operations in the chemical, biochemical, food and pharmaceutical industries. Examples of such operations include emulsification, solvent extraction, biotransformation and fermentation. In all these cases, the size and size distribution of drops are amongst the most important parameters determining the stability of the dispersion and the efficiency of the contacting operation. The stable size and size distribution of drops and the rate at which they are achieved is a delicate balance between drop breakage and coalescence. Prediction of the mean drop size and size distribution therefore requires knowledge of these two rate processes. In this study models are developed for breakage rate based on fluid dynamics considerations. These are incorporated into a general population balance equation which is then solved numerically in order to simulate both the transient and steady state size distributions. Measurements of drop size distributions as a function of agitation time are carried for the dispersion of xylene in water in the presence of a surfactant to minimise drop coalescence, in a 0.15m diameter vessel agitated by a standard Rushton turbine. The volume concentration of xylene in these experiments are kept to less than 50%, conditions for which the mixture viscosity remains relatively low and Newtonian and agitation in the vessel remains turbulent. Limited comparative experiments are also reported in turbulent flow by using an up-and-down moving impeller. These experiments are aimed primarily towards establishing the influence of operating conditions and physical properties of the dispersion on drop breakage. The experimental drop size distributions are analysed and compared with the simulations and good agreement is obtained between the two. Additionally, experimental data are reported for xylene-in water dispersions in the mechanically stirred vessel under high xylene phase concentration, conditions for which the rheology of the dispersion is Newtonian and flow in the vessel is laminar and/or transitional. The population balance equations are modified to account for these effects and solved using a two-zone model. Good agreement is found between experimental data and simulations. The Sauter mean drop diameter of the steady-state drop size distributions for both the up-and-down moving impellers and the Rushton turbine are analysed using established models of drop breakage under turbulent and laminar flow regime and new method is presented for the analysis of power input data for up-and-down moving impellers. Recommendations were made to model coalescence and breakage simultaneously, to study vibromixer at high dispersed phase volume and to improve the discretization techniques so as to manipulate the grid size during simulation.

The laminar annular flow field formed in the gap of two concentric rotating cylinders with and without a small degree of superimposed axial flow can be used effectively to manipulate the residence time distribution of various species in the gap in order to promote radial mixing with negligible axial dispersion, conditions that are considered important for good heat and mass transfer and reaction processes. This thesis is concerned with an experimental and theoretical study of laminar Couette flow stability in concentric devices. Numerical simulations based on the solution of Navier-Stokes equations are presented for Newtonian and non- Newtonian liquids, showing the effects of operating and geometrical parameters on the transition of laminar to Taylor vortex flow for induced rotational-axial flow in the gap of a pair of rotating cylinders. These simulations indicate that annular rotational flow becomes more stable in the presence of a small degree of axial flow and as the gap width increases. The effect of rotational speed on the breakdown of laminar flow is more complex and for given radius ratio and axial flow rate depends on both the angular speed ratio and the direction of the rotation of the cylinders, counter-rotating flow generally producing a more stable flow than co-rotating. Experimental data are provided on the residence time distribution for flow of several Newtonian water-glycerol solutions and non-Newtonian carboxymethyl cellulose solutions and xanthan gum solutions through the gap of two concentric rotating cylinders operating over a range of conditions. The equipment used consisted of a pair of horizontal cylinders 1 m long with a stationary outer perspex shell and an inner rotating shaft with a variable speed control. The results from these experiments indicate that the residence time distribution of the species in the gap is a complex function of the flow structure in the gap which in turn depends on the operating and geometrical variables. The experimental findings are successfully analysed and discussed using the simulations studies. Recommendations are made for future work.

Much of what is known about the acquisition of sociolinguistic competence in second language Spanish is based on studies of English-speaking learners. This body of research has demonstrated common pathways in the acquisition of multiple variable structures, as well as how second language learners come to use (or not) Spanish in linguistically and socially appropriate ways. Nevertheless, our understanding of the acquisition of sociolinguistic variation in second language Spanish is limited by the lack of systematic investigation of learners from distinct individual and social backgrounds. To that end, this dissertation explored Korean learners' acquisition and use of three variable structures in Spanish--the copula with adjectives, subject expression, and intervocalic stops. The study compared Korean-speaking learners' (N = 66), English-speaking learners' (N = 65), and native speakers' (N = 11) production of each variable structure under study in a sociolinguistic interview, a picture book description task, and a picture identification task. Additionally, an acoustic analysis of all participants' intervocalic stop production is offered. Results indicate that Korean-speaking learners' development of the copula with adjectives is associated with an increase in estar "to be" use and a decrease in omission. Subject expression development is associated with an increase in the production of null subject pronouns. Intervocalic stop development is linked to a decreased use of [f] for /p/ and of voiceless stops for /b d ?/. Inferential statistical analyses revealed various linguistic and extralinguistic factors to significantly constrain Korean-speaking learners' variable use of each linguistic structure across levels. Findings for Korean-speaking learners are directly compared to the English-speaking learners and native speakers sampled. Quantitative findings are complemented by a qualitative report of classroom observations conducted at the Korean research site. Results are discussed in relation to previous research, and contributions, limitations, and future directions are outlined and addressed. [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.]

Economical operation of future satellite systems for mobile communications can only be fulfilled by using dedicated on-board processing satellites, which would allow both cheap earth terminals and lower space segment costs. With on-board modems and codecs, the up-link and down-link can be optimized separately. An attractive scheme is to use FDMA/SCPC on the up-link and TDM on the down-link. This scheme allows mobile terminals to transmit a narrow band, low power signal, resulting in smaller dishes and HP As with lower output power. On the up-link, there are hundreds to thousands of FDM channels to be demodulated on-board. The most promising approach is the use of all-digital multicarrier demodulators (MCDs), where analogue and digital hardware are efficiently shared amongst channels, and digital signal processing is used at an early stage to take advantage of VLSI implementation. A MCD consists of a channelliser for separation of FDM channels, followed by individual modulators for each channel. Major research areas in MCDs are in multirate digital signal processing, and the optimal estimation for synchronization, which form the basis of the thesis. Complex signal theories are central to the development of structured approaches for the sampling and processing of bandpass signals, which are the foundations in both channelliser and demodulator design. In multirate DSP, novel polyphase theories replace many ad-hoc, tedious and error-prone design procedures. For example, a polyphase-matrix-DFT channelliser includes all efficient filter bank techniques as special cases. Also, a polyphase-lattice filter is derived, not only for sampling rate conversion, but also capable of sampling phase variation, which is required for symbol timing adjustment in all-digital demodulators. In modulation schemes, a systematic survey is reported, based on two expressions that includes all formats in linear and constant envelope modulation. In synchronization techniques, classifications according to the criterion of statistical optimization, the data dependency, and the method of parameter extraction, reflect the inherent complexity and performance of numerous existing algorithms. The designs of two new algorithms are presented: a differential decision frequency error detector that is simple and fast; a dual-comb-filter frequency/timing error detector that is targeted at VLSI implementation. The real-time implementation of a complete 4x16 kb/s MCD for the T-SAT project is described in detail, which proved many of the structured design concepts developed in the thesis. The requirements of software tools for various levels of simulation in multirate DSP and communications are analysed. This led to the implementation of a data-flow oriented simulation system, which was used in all research work in the thesis.

Addressing the pressing global need for clean energy amid accelerating climate change, the research highlights the urgency of reducing CO₂ emissions by transitioning from fossil fuels to renewable energy sources. Current global temperatures are above long-term averages, emphasizing the need for non-fossil energy solutions. Hydrogen is increasingly recognized as a pivotal clean energy candidate. Although most hydrogen is currently produced from fossil fuels, advancements in hydrogen production and its expanding applications in electric vehicles and energy grids would offer promising avenues for mitigating global warming. Hydrogen fuel cells and water electrolysis are identified as key technologies for harnessing hydrogen's potential. However, these technologies face significant economic and technical challenges that must be overcome for widespread adoption. Alkaline Water Electrolysis (AWE) is the most prevalent method but suffers from low current density. Alternative approaches like Solid Oxide Electrolysis Cells (SOEC), Proton Exchange Membrane Water Electrolysis (PEMWE), and Anion Exchange Membrane Water Electrolysis (AEMWE) encounter barriers in commercialization due to high costs and technical complexities. Recent research has turned towards polymer 8 membrane technologies as a promising solution. These membranes offer lower costs and operate at more manageable temperatures compared to conventional methods. Proton Exchange Membrane (PEM) technologies hold promise for electric vehicles and energy grids due to high energy density, despite requiring expensive catalysts and specialized fluorinated membranes. In contrast, Anion Exchange Membranes (AEMs) present cost advantages but face challenges with membrane decomposition and low conductivity. This study aims to guide the optimal design of Anion Exchange Membranes (AEMs) to enhance their conductivity and stability. Using molecular simulations, this research predicts membrane performance and identifies polymer structures that could improve system performances. Classical molecular dynamics (MD) simulations are employed to explore structural changes and their impact on polymer properties and performance, offering a systematic approach to polymer design at the molecular level. The primary objectives of this work include (i) identifying descriptors critical for a better understanding of polymer decomposition and (ii) developing methods capable of predicting the equilibrium sorption isotherms of water equilibrium in these polymeric systems. The study centers on proposing a design guide based on a quantitative analysis of the structure-performance relationship of poly(2,6-dimethyl-1,4-phenylene oxide)-trimethylammonium (PPOTMA) as a reference polymer. Moreover, the outcome would contribute to propelling the development of AEMs towards widespread commercial viability in the economic landscape.

In view of the increasing prevalence of neurodevelopmental disorders, the burdens they inflict on affected patients, family and society, as well as their unclear etiology, identifying their risk factors is of tremendous importance. We need to better determine the etiologic window of intervention and identify high-risk populations who would benefit from public health interventions. Given the early onset of neurodevelopmental disorders, scientists have focused on risk factors in early childhood and in utero, based on the Developmental Origins of Health and Disease hypothesis. However, the third generational effects have rarely been explored, although pregnant women’s exposures to internal and external factors can affect not only their offspring but their grandchildren via the germline. This dissertation comprises three studies that investigated the multigenerational associations of grandmothers’ exposures experienced periconceptionally with their grandchildren’s neurodevelopmental deficits and adverse birth outcomes. Neurodevelopmental disorders are a group of heterogeneous conditions that affect the development of the neurological system and brain, leading to a delay or disturbance in the acquisition of motor, social language and cognitive skills. While some neurodevelopmental disorders are highly heritable, most have complex and multiple contributors, including genetic, biological, psychosocial and environmental factors. Adverse birth outcomes, including preterm birth or low birth weight, have also been proposed as risk factors for pervasive developmental disorders as well as many chronic diseases later in life. While previous studies have focused on the relationship between maternal exposure during pregnancy and health outcomes in the offspring, the third generational effect has been neglected, mainly due to a lack of data on three generations in human, particularly for neurodevelopmental deficits. However, with access to a national, longitudinal healthcare professional’s cohort study data (Nurses’ Health Study II), this research possessed a unique setting to investigate associations in a population-based study across three generations. In Chapters 1 and 2, grandchildren’s risk of attention-deficit/hyperactivity disorder (ADHD), one of the most prevalent neurodevelopmental disorders, was assessed according to grandmothers’ smoking during pregnancy, pre-pregnancy obesity status, and gestational weight gain. In Chapter 3, we revisited a previous study that examined the relationship between grandmaternal use of diethylstilbestrol during pregnancy and risk of preterm birth and low birth weight, by further addressing a more comprehensive list of confounding factors as well as indication bias. Overall, the findings from this research demonstrate that grandmothers’ periconceptional health risk behaviors, unhealthy weight status, and exposure to an endocrine disruptor were associated with their grandchildren’s risk of ADHD or preterm birth/low birth weight, independent of maternal factors., Epidemiology