69238 research outputs found
Sort by
How to change the welfare state from a Taxation to a Savings based model
No full textThe future of public welfare states is in doubt as costs trend up due to population ageing. Yet there is little agreement about how to reform them. We show how tax cuts can be designed to establish mandatory savings accounts so that a (mostly) publicly funded and provided welfare system can be changed into one that relies largely on private funding and private suppliers. Greater competition in the provision of health-care services offers the potential for significant efficiency gains. The government retains sufficient revenues to act as ‘insurer of last resort’ for those individuals unable to meet welfare costs out of their savings accounts. To our knowledge, showing how both a tax and welfare reform can be jointly designed to enable the transition to this new type of system to occur in a potentially politically feasible way has not been done before
Person-centred Planning to Support Employment Prospects of People with Intellectual Disabilities
No full textEmployment offers individuals with intellectual disabilities (ID) opportunities for independence, self-determination, and improved quality of life. However, they often face barriers such as employer biases, societal stigma, and skill deficits. Supported employment programmmes provide tailored training, but it is unclear if these align with personal interests. Person-centred Planning (PCP) may enhance self-determination by focusing on individual goals and preferences. This study explored the effectiveness of PCP in improving job-related skill acquisition for adults with ID in a supported employment setting. It examined whether PCP could increase learning opportunities aligned with personal goals and enhance task performance. The research combined PCP with Preference Assessments (PA) to identify meaningful job- related goals. A multiple-probe design was used with four participants in a training café, measuring task accuracy, learning opportunities, and the social validity of PCP. Results showed that PCP and PA effectively identified preferred tasks and increased goal-aligned learning opportunities. Three of the four participants significantly improved their task performance. The study emphasised the importance of involving participants in decision-making, enhancing autonomy and engagement. However, challenges like environmental constraints, limited resources, and insufficient support from participants’ circles hindered PCP implementation. These findings suggest PCP is a viable approach for improving skill acquisition and employment outcomes for individuals with ID. By providing a structured framework for goal-setting and skill development, PCP fosters better communication between trainees and support staff, leading to more tailored training programmmes. Future research should address barriers, explore PCP’s long-term sustainability, and extend its application to diverse employment contexts
Aged Residential Care Deterioration Early Warning System (DEWS) feasibility study
No full textThis report details the feasibility research study undertaken between January and October
2024. The study examined whether the Deterioration Early Warning System (DEWS) could
be implemented in age-related residential care (ARC) in Aotearoa New Zealand and, more
importantly, whether it should be implemented in ARC
Variation in the realisation of vowels in Māori ꞌAvaiki Nui Cook Islands Māori
No full textCook Islands Māori / Māori ꞌAvaiki Nui (CIM) has the typical vowel paradigm found in Polynesian
languages - five vowels /i e ɐ o u/ with phonemic length distinction. However, a point of contention in
language talisation often observed in the Pacific context is the claim that “in our language the vowels
are always pronounced the same”. This can lead to stress in learners who have anxiety about meeting
this pronunciation criteria. We hope that by showing the systematic variation within “native” speaker
vowel realisation we can reassure learners that this is a normal and harmless phenomenon. The
differences between varieties of CIM are usually discussed by the language community in terms of
lexical and prosodic variation. In this presentation we will show another type of variation in CIM:
variation of vowel quality. We do this to document the diversity of the language and its potential vowel
shifts, as well as to support teachers in promoting a descriptivist approach.
Each Island in the Cook Islands is considered to have its own variety. Some linguists and community
members consider the varieties spoken in the Northern island of Manihiki Rakahanga and Tonareva
(Penrhyn) to be distinct languages (Nicholas 2018). Staying agnostic on this question, we collected data
from several of the islands in the archipelago and used a machine-learning model to describe their
variation. We analysed 11 speakers from the islands of Rarotonga (n=6), Atiu (n=1), Mauke (n=3) and
Mitiaro (n=1), for a total of 4 hours 4 minutes of conversational speech. The data comes from the Te
Vairanga Tuatua corpus (Nicholas 2012), and was transcribed by expert speakers of CIM using ELAN
(2024) and then force-aligned using a custom-trained MFA model (McAuliffe et al. 2017). We used
Praat (Boersma & Weenink 2024) to extract the data of 70894 vowels, specifically the F1, F2 and F3
at 50% of the duration of each vowel. We converted the hertz frequencies to barks, and then z-scored
them for each speaker. Figure 1 shows this data by island and by type of vowel.
There are some general trends visible in the data. First, the position of the /u/ and /u:/ vowels in
Rarotonga is more fronted than in any of the other islands. This might be due to intense contact with
Pacific (including New Zealand) and American varieties of English, which might be transforming /u/
into a [ʉ] (Bauer 2015, Labov et al. 2006). In the other islands, the /u/ is further back from the short /o/.
These islands have many fewer tourists, and less contact with English overall. A second pattern is the
difference in vowel quality between long and short vowels: The /a/~/a:/ vowels are consistently different
across the islands, but vowels like /e/~/e:/ can have very different qualities (as in Atiu), or be almost
identical (as in Maꞌuke).
There are several horizons of future work for this project. The first is to analyse information from
additional islands. For example, Aitutaki has a large tourism footprint, and so it might be a candidate
for more contact effects with English. On the other hand, islands like Mangaia in the Southern Group,
or Rakahanga and Tongareva in the Northern Group might have more prototypical vowel triangles. The
second important goal is to convert these results into a format that is useful for school teachers and
diaspora speakers, and produce educational materials that explain the variation in the vowels in a
context that is positive towards language revitalisation
Introduction to go-alongs as a qualitative research method in applied linguistics
No full textThe go-along method, a place-based research approach that integrates the strengths of interviews and observations, presents opportunities for exploring the intersections of language and place, such as language use and performance (e.g., people's engagement with the linguistic landscape or people's choice of language in certain contexts or for particular tasks), and language teaching (e.g., situational language teaching, out-of-class language learning, language teaching through linguistic landscapes). Initially developed in sociology and subsequently adopted in other disciplines, go-alongs remain underutilised in applied linguistics research. Grounded in the sociocognitive approach and public pedagogy (out-of-class learning), this study demonstrates how go-alongs can conceptually link language and place by illustrating the three functions they serve: descriptive, pedagogical, and diagnostic (or evaluative). The descriptive function facilitates the documentation of participants’ language use across various places and settings. The pedagogical function supports go-along language teaching by assigning participants tasks tied to specific real-life places, and the data of place-related language use collected during task-doing is developed into materials for future teaching. The diagnostic (or evaluative) function allows for testing participants’ language skills/performances at different places. Go-alongs can also contribute to triangulating interview and observation data, thereby strengthening methodological validity and reliability. Drawing on our two empirical projects that used go-alongs as the primary research tool, we propose a set of procedural guidelines aimed at researchers or practitioners using go-alongs in applied linguistic research. We close with limitations and suggestions for future studies
LIBRA: Measuring Bias of Large Language Model from a Local Context
No full textLarge Language Models (LLMs) have significantly advanced natural language processing applications, yet their widespread use raises concerns regarding inherent biases that may reduce utility or harm for particular social groups. Despite the advancement in addressing LLM bias, existing research has two major limitations. First, existing LLM bias evaluation focuses on the U.S. cultural context, making it challenging to reveal stereotypical biases of LLMs toward other cultures, leading to unfair development and use of LLMs. Second, current bias evaluation often assumes models are familiar with the target social groups. When LLMs encounter words beyond their knowledge boundaries that are unfamiliar in their training data, they produce irrelevant results in the local context due to hallucinations and overconfidence, which are not necessarily indicative of inherent bias. This research addresses these limitations with a Local Integrated Bias Recognition and Assessment Framework (LIBRA) for measuring bias using datasets sourced from local corpora without crowdsourcing. Implementing this framework, we develop a dataset comprising over 360,000 test cases in the New Zealand context. Furthermore, we propose the Enhanced Idealized CAT Score (EiCAT), integrating the iCAT score with a beyond knowledge boundary score (bbs) and a distribution divergence-based bias measurement to tackle the challenge of LLMs encountering words beyond knowledge boundaries. Our results show that the BERT family, GPT-2, and Llama-3 models seldom understand local words in different contexts. While Llama-3 exhibits larger bias, it responds better to different cultural contexts. The code and dataset are available at: https://github.com/ipangbo/LIBRA
Designing Trust in Cosmetic Medical Tourism: The SIGN Framework for Empathy and Cultural Sensitivity in the Pre-Visit Patient Experience
No full textThis thesis explores how visual design and strategic communication influence patient trust in cosmetic medical tourism. As international healthcare experiences become increasingly digital and culturally diverse, trust emerges as a critical factor, particularly during the pre-visit, website-driven decision phase. While existing literature emphasises content quality and medical credentials, it often overlooks culturally adaptive visuals, emotional resonance, and user journey design.
To address this gap, the study adopts a mixed-methods, design-led approach to evaluate how cosmetic clinics construct trust through digital interfaces. The research began with a benchmark analysis of three foundational frameworks: Cultural Sensitivity Cultivation (CSC), Multimodal Communicative Model (MCM), Medical Tourism System & Service Provision Frameworks (MTSF & MTSPF). Additionally, more than twenty industry frameworks, toolkits, and models from service design, patient experience, healthcare communication, social sciences, human-centred design, and cultural competence were examined. These insights were synthesised through affinity mapping into the SIGN framework, a structured trust model comprising four pillars: Sensitisation, Involvement, Guidance, and Navigation.
The SIGN framework guided the development of a website audit matrix, structured around rubric-based scoring and visual heatmaps. Nine cosmetic surgery clinic websites from South Korea, Turkey, and Colombia were evaluated across five key sections: homepage, about/team, international patient support, services/treatments, and testimonials. Evaluation methods were drawn from the Pragmatic Usability Rating by Experts (PURE) system and visual analytics inspired by Hotjar. These audits validated the SIGN pillars by highlighting strengths and gaps in website layout and navigation flow.
The study also identified standard tools to address trust gaps, such as patient journey maps, personas, emotional probes, clarity checklists, and visual storytelling templates. These were integrated into the SIGN framework, operationalising it into a practical, evidence-based design toolkit.
This research contributes both theoretically and practically. It positions culturally sensitive visual communication as foundational to digital trust. It offers a validated framework and actionable toolkit to assist clinics, designers, and healthcare communicators in building trust with international patients at the critical pre-visit decision stage
Beyond technical competencies: A critical analysis of global research on language teacher AI literacy
No full textThis article critically examines global research on teacher AI literacy, highlighting significant gaps in theory, assessment, and pedagogical integration. Focusing on language education, particularly Chinese as a Foreign Language, it argues for a shift beyond technical skills toward context-aware, ethically grounded, and pedagogically meaningful frameworks that enable a profound educational transformation
A pre-clinical drug study using Cysteamine/Everolimus combination therapy to treat cystinosis knock-out rats
No full textNephropathic cystinosis is a rare lysosomal storage disorder that is caused by mutations in the CTNS gene, encoding the cystine transporter cystinosin. Loss of cystinosin leads to cystine accumulation and crystal formation within the lysosomes of cells, resulting in tissue damage. Although cystinosis is a multi-systemic disorder, the kidneys are the first and major organ affected. Patients display symptoms around 6 months of age that include failure to thrive, excessive thirst (polydipsia) and urination (polyuria), and proximal tubule dysfunction in the form of Fanconi syndrome (excessive urinary loss of essential nutrients). Cystinosis eventually progresses to end-stage kidney failure, requiring a renal transplant within the first decade of life if left untreated. Currently, the only treatment available is the aminothiol, cysteamine, which depletes lysosomal cystine levels. However, cysteamine does not rescue all the renal manifestations of cystinosis and there is an urgent need for better therapies. We previously showed that activation of autophagy using the pharmacological mTORC1 inhibitor, everolimus, combined with cysteamine leads to a rescue of the cystinotic phenotype in vitro. In this thesis, we performed preclinical drug studies on a Ctns knockout (Ctns-/-) rat model to assess the therapeutic potential of the cysteamine and everolimus combination treatment, with the hypothesis that this will confer greater renoprotection compared with cysteamine monotreatment. Six-week-old male and female Ctns-/- rats were treated for 6 months with jelly pills containing 30 mg/kg of cysteamine given twice daily, 10 mg/kg of everolimus given twice weekly or the combination of the two drugs. We found that cysteamine monotreatment partially improved the renal cystinotic phenotype in both sexes, while everolimus monotreatment conferred some mild improvements but only in males. The combination treatment resulted in a superior reduction in tissue cystine levels, and reduced the severity of the polyuria, polydipsia, renal histological lesions and delayed the onset of the Fanconi syndrome in male rats. By contrast, female Ctns-/- rats showed a response to the combination-treatment that was largely equivalent to cysteamine monotreatment, consistent with a sex-dependent response to everolimus. We also found that the combination treatment adversely affected growth, lipid homeostasis and wound healing in both sexes, suggesting that further optimisation of drug dosage and timing may be necessary to improve efficacy and overcome the influence of sex. Taken together, this work demonstrates the potential of a cysteamine-everolimus combination therapy to enhance the treatment of cystinosis and paves the way for preclinical trials in humans
Essay on Mergers and Acquisitions, Firm Performance, and Target Valuation
No full textThis thesis investigates mergers and acquisitions (M&As) within the Chinese market by focusing on the influence of related-party M&As, network centrality, and policy changes on M&A outcomes. Chapter 1 presents an examination of the centrality of target firms within social networks, particularly through the connections of CEOs and board members. It asserts that a central position is significantly correlated with higher acquisition premiums, which can be attributed to reduced information asymmetry. This effect is stronger in non-state-owned enterprises (non-SOEs) than in state-owned enterprises (SOEs). Chapter 2 presents an investigation of the influence of RPTs on the long-term performance of acquiring firms. It asserts that RPTs generally provide significant benefits, particularly enhancing the post-M&A performance of non-SOEs over SOEs. The analysis reveals that related-party M&A alleviates financial constraints, leading to positive stock market reactions and improved long-term performance. Chapter 3 addresses the valuation of unlisted targets and evaluates the effectiveness of policy reforms introduced by the China Securities Regulatory Commission in 2014—“Regulatory Practice of Major Asset Restructuring for Unlisted Companies” and “Regulatory Practice of Acquisition of Unlisted Companies.” The purpose of these new policies is to increase transparency and market orientation in M&A processes involving unlisted companies. However, regardless of these intentions, the findings indicate that these policies have neither successfully enhanced transparency nor mitigate the discounts for unlisted targets. Collectively, these chapters provide a nuanced understanding of the strategic, financial, and policy dimensions shaping M&A activities in China. Moreover, the findings suggest that current policy strategies require reevaluation to better align with market dynamics and transparency objectives