Tuwhera Research Repository

The Tuwhera Research Repository provides sustainable open access and archiving to a broad range of AUT produced research - theses, research outputs and datasets.

Theses & dissertations

Deposit your thesis

Datasets

Deposit your dataset

Research outputs

Deposit your research output
via Research Elements

Recent Submissions

Access status: Open Access ,
Beyond 'Good Enough for a Girl'. Understanding How Mainstream Media Representations of the Black Ferns (Ngā Mamaku) Have Shifted Over a Decade
(Auckland University of Technology, 2026) Peti, Fereni; Treadwell, Gregory
The game of rugby has historically been seen as a man’s game, particularly in Aotearoa New Zealand where the nation’s identity is centred around the mythology that is the All Blacks. As a result, women in rugby and in particular the Black Ferns or Nga Mamaku have typically been sidelined in the landscape of sports media. Mainstream media have created an illusion that this pattern of marginalisation is no longer present and that equity in coverage has already been achieved for the Black Ferns. However, it is far from the truth. Rooted in patriarchal and colonial systems, the Black Ferns face systemic barriers that intersect to protect the masculine identity of rugby and maintain male dominance at the top of the sports media hierarchy. The present research investigates how mainstream media have constructed representation of the Black Ferns across two distinct periods (2014-2015 and 2024-2025) which both includes a non-World Cup year and World Cup year. In particular, the study examines how visibility, framing patterns, and representation has shifted across the decade and what these shifts reveal about media attitudes and broader structures. Relevant literature and secondary sources of online articles have been used in this study to understand these shifts and expose what gaps still exists for media in their coverage of the Black Ferns. The study concludes that although mainstream media have made progress to move away from gendered stereotypes and limited visibility for the Black Ferns, there are still significant gaps in media coverage of the team. It uncovers subtle traces of gendered ideologies that frame women in a stereotypical manner, masked patterns of conditional visibility, and fragmented portrayal of the players indigenous identities. Mainstream media still have a long way to go before they claim the illusion of equity in media coverage as a reality for the Black Ferns. It is recommended that media change the institutional structure to allow for more routine coverage of the team, build competency around the use of gendered stereotypes, and tackle more stories that speak to the cultural identity of a team that is predominantly Maori and Pasifika.
Access status: Open Access ,
When Source Disclosure Matters: User Perceptions of Generative AI Authorship in Landing Page Content
(Auckland University of Technology, 2026) Mirza, Hira; Guinibert, Matthew
The widespread adoption of generative AI (GenAI) in digital content marketing has outpaced the understanding of how users evaluate it. While marketers deploy GenAI to produce personalised content at scale, user perceptions of authenticity, credibility, and value remain poorly understood. One particular under-researched area is conversion-oriented environments where content must inform, persuade, and establish trust simultaneously. Existing literature on user evaluations of digital content addresses GenAI content perception broadly, but little empirical work examines how source disclosure and functional context together shape user preferences. This study investigates user preferences for GenAI-authored versus human-authored content on website landing pages, a persuasion-critical environment where content credibility directly influences decision-making. Using a 2x2 mixed comparative experimental design, 95 participants recruited via CloudResearch Connect evaluated landing page content under both disclosed and undisclosed authorship conditions to assess how source awareness affects content evaluations. The findings reveal a central paradox. Users evaluate GenAI-authored content favourably when it is informative and well-structured, yet prefer human authorship once the source is disclosed. Disclosure was found to activate relational scrutiny, such as authenticity and human-likeness, without significantly degrading functional evaluations of content quality, persuasiveness and credibility. Rather than confirming AI aversion, results show that human favouritism operates alongside, not instead of, positive GenAI evaluations. The study proposes a dual evaluative framework to account for this. Findings are interpreted through this lens as reflecting functional competence when GenAI content is judged on its informational merit, and relational authenticity when human authorship is valued for its perceived intentionality and authority. Critically, user acceptance of GenAI content is conditional rather than categorical, shaped by task, type, digital environment, and the level of user trust at stake. These findings offer empirical grounding for marketers navigating GenAI adoption and for policymakers designing authorship disclosure standards.
Access status: Open Access ,
Image-Based Classification of Bolus Phases in Robotic Mastication Using Hybrid Machine Learning Analysis
(Elsevier BV, 2026-09-22) Akarawita, I; Chen, B; Dhupia, JS; Stommel, M; Xu, W
Non-invasive, continuous monitoring of food bolus transformation during robotic mastication remains a significant challenge, as existing methods rely on physical extraction of the bolus or single-modality image features that lack robustness across the continuous and non-linear nature of food breakdown. This paper investigates whether hybrid machine-extracted image features can reliably classify chewing phases during in vitro robotic mastication without physical intervention. A hybrid feature-extraction framework is proposed that integrates deep convolutional features from VGG16 with classical texture descriptors, Grey-Level Co-occurrence Matrix (GLCM), image lightness, and variance, applied to a dataset of 1024 segmented chewed biscuit bolus images captured across two chewing trajectories: vertical motion (T1) and combined vertical and lateral motion (T13). Supervised classifiers, Support Vector Machines, Linear Discriminant Analysis (LDA), and Random Forests were evaluated using trial-level grouped cross-validation. Using a 3-phase schema conceptually aligned with the Hutchings and Lillford Mouth Process Model, LDA achieved accuracies of 75.9% for T1 and 82.3% (T13), with the hybrid feature set outperforming the ablation study. Improved separability with T13 demonstrates that shear effect produces more distinctive bolus image features. The unchewed class was consistently well separated, while structured misclassification patterns occur mainly between adjacent phases highlight the challenging nature of the study. These results establish a non-contact, vision-based measurement framework for classifying bolus evolution as a proof-of-concept foundation for robotic chewing systems.
Access status: Open Access ,
Single-Channel EEG Denoising and Artifact Removal: A Systematic Scoping Review of Methods, Benchmarking, and Reproducibility
(IOP Publishing, 2026-10-06) Shaikh, Usman Qamar; Kalra, Anubha; Lowe, Andrew; Khan, Imran
Objective. Single-channel and wearable electroencephalography (EEG) reduce acquisition burden but limit the spatial information available to conventional multichannel artifact-removal methods. We conducted a systematic scoping review of single-channel-compatible EEG artifact removal, linking method operation with artifact coverage, evaluation design, generalisation, deployment, comparison, and reproducibility. Approach. Database and citation searches through 15 August 2026 identified 178 articles describing 199 method variants. For learned methods, 303 reported model–dataset evaluations were also examined for data independence, generalisation, and deployment. Main results. Most variants were reference-free at inference (179/199), while 13 used a non-neural auxiliary reference. Methods spanned seven operating paradigms, including decomposition, transformed source separation, adaptive correction, and learned reconstruction. Ocular and muscle artifacts were evaluated most often, and 38/178 articles tested simultaneous or overlapping mixtures. Across 380 dataset uses, evaluation centred on controlled signal reconstruction. Known-target fidelity was reported for 170/199 variants, compared with 21 for clean-input or non-artifact signal comparison, 32 for contaminated downstream utility, and three for clean-input downstream non-degradation. Among 81 learned variants, 18 reported source-trained application to an external dataset, device, or site; only 31/303 evaluations explicitly separated test windows or epochs, clean sources, and artifact sources from training. No measured deployment evidence was reported for 122/199 variants. Public method resources were available for 21 variants, while trained weights and explicit licences were uncommon. Significance. The review provides a framework for selecting datasets, metrics, comparators, and implementations, and for distinguishing reconstruction, neural preservation, transfer, and wearable deployment evidence.
Access status: Open Access ,
Construction Supply Chain Management for Circular Economy Performance: Procurement Economics and Configuration Evidence From New Zealand
(MDPI AG, 2026-10-09) Tookey, John; Sultanbekov, Kanat; Rotimi, Funmilayo Ebun; Dhawan, Kamal
Supply chain management (SCM) optimisation is a critical but under-examined pathway for delivering circular economy outcomes in the construction sector, particularly in geographically remote and small- and medium-sized enterprise-dominated markets such as New Zealand (NZ). This study investigates how SCM practices, such as digital tools, early contractor involvement, strategic partnerships, and reverse logistics, can be optimised to improve circular performance in NZ-built environment projects. A pragmatist, abductive mixed-methods design combined 75 case studies, a 28-case quantitative subset, 15 elite semi-structured interviews and documentary evidence. Reflexive Thematic Analysis was integrated with descriptive statistics, cross-tabulation, non-parametric testing, and configuration archetype analysis. Coordination and partnerships are the most frequently observed SCM lever, appearing in 46.7% of the 75 cases and in all four cases in the highest waste-diversion band, while digital tools are seldom deployed as the primary lever. Two operational metrics summarise these configurations: a Waste Hierarchy Index (WHI) of 77.6%, calculated across the 66 cases with an assignable waste hierarchy position, and a Digital SCM Adoption Rate (DSAR) of 15.1%, calculated across the 73 cases with a stated primary lever. Together they indicate an upstream but digitally uneven pattern of optimisation. The study contributes a configuration-based method for comparing SCM levers in circular construction, together with the first NZ evidence base of this kind, and finds that procurement reform rather than digital investment is the key constraint on circular performance. Practical implications include embedding early contractor involvement and integrated design in public procurement, funding sector-wide digital interoperability rather than firm-level tools, and extending site waste management plan reporting. Priority future research is longitudinal, scale-stratified testing of the configuration archetypes identified in this paper.