Accessible & Collaborative Digital Practices
Date: Tuesday 11 Nov 2025
Room: S/3.20 Turing Suite
| Time | Presentation | Abstract | Authors | Contribution type |
|---|---|---|---|---|
| 15:30-15:40 | Strategic Pausing in Digital Reading: A Pilot Evaluation of SmartPause on Cognitive Performance | Digital reading environments are prone to premature disengagement, which can impair comprehension and cognitive performance. This pilot study introduces a simplified version of SmartPause—a lightweight, timing-sensitive intervention that encourages readers to continue until a natural breakpoint. Unlike earlier conceptual designs that included reflection prompts and note-taking, this version isolates the effect of exit timing alone, offering a more minimal and scalable interaction. By removing additional scaffolding and controlling for individual variability, we test whether timing cues by themselves support cognitive performance. Fourteen participants read two digital passages in a fixed order: one passage ended at a natural breakpoint, while the other was exited partway through. Exit timing was counterbalanced across participants to compare cognitive outcomes associated with early versus full completion. Measures included selective attention, free recall, perceived workload, engagement and affective state. Preliminary results suggest directional benefits in favour of natural breakpoint-aligned exits, particularly for memory, attentional control, and affective tone. These trends tentatively support the role of cognitively aligned pausing in enhancing digital reading outcomes without adding user burden. This exploratory pilot offers early evidence for context-aware disengagement as a design principle and motivates future development of adaptive reading systems aligned with cognitive rhythms. | Naile Burcu Hacioglu, Maria Chiara Leva and Hyowon Lee | Late Breaking Work |
| 15:40-15:50 | StemA11y: An AI-Driven Mobile System for Non-Visual and Multisensory Access to STEM Content | Blind and low-vision (BLV) learners encounter persistent barriers in accessing STEM educational materials. Traditional accommodations, including manual Braille transcription and tactile graphic production, are labor-intensive and time-consuming, often resulting in significant delays that impede timely engagement with STEM content. To address this gap, we present StemA11y, an AI-powered iOS application that automatically converts STEM materials into accessible formats and delivers them via speech, audio cues, haptics, and VoiceOver support. We evaluated StemA11y with 11 BLV participants across two content types–math worksheets and adapted SAT papers. Our results revealed that System Usability Scale (SUS) scores exceeded the industry benchmark of 68 for both math worksheets (M=75.2) and SAT papers (M=83.2). Technology Acceptance Model (TAM) ratings indicated high perceived ease of use, ease of learning, and overall satisfaction. Further, qualitative feedback demonstrated the importance of customizable math verbosity, consistent graphic cues, and context-aware navigation. Drawing on these insights, we discuss design considerations for developing multisensory AI-powered mobile systems to improve STEM accessibility for BLV learners, including configurable content verbosity, strategies to reduce redundant cues and cognitive load, and enhanced support for exploring graphical information. | Hari Palani and Rudaiba Adnin | Late Breaking Work |
| 15:50-16:00 | Towards Human-AI collaborative methods in UX: a qualitative case study for discovery research platforms | This paper introduces human and Artificial Intelligence (AI) collaboration for the preparation and analysis of empirical qualitative data. Through a case study, we explore this collaboration with user experiences of researchers across four research disciplines: (1) Molecular Dynamics; (2) Earth Systems; (3) Social Sciences and Humanities; and (4) Mathematics, alongside commercial data providers. Generative AI and Large Language Models (LLMs) conducted processes within the thematic analysis of empirical interview transcripts led by researchers. Over a six-month period, we iteratively translated user insights into design artefacts – user personas, scenarios, and tool functionalities. Here, we report on the focus of the analysis stages combining AI-process and critical human decision-making. This case study demonstrates how using generative AI can streamline traditional qualitative methods in discovery tool design, whilst maintaining quality of research outcomes. We highlight the stages required for delivery of meaningful, collaborative, user-centred design artefacts, and reflect on our Human-AI interactions. | Diane Morrow, Franjo Pehar, Crystal Silver, Kathy-Ann Fletcher and Stefano De Paoli | Late Breaking Work |
| 16:00-16:10 | Understanding Human Centred AI in Digital Health: Policies and Privacy protection for African users | This study examines the data protection and the application of human-centred artificial intelligence (AI) in digital health across 10 African countries, emphasizing informed consent and privacy. With increasing AI use in healthcare, particularly in low-income nations, concerns around data privacy and inconsistent regulations have emerged. Using systematic review methods, this study reviews data protection laws of 10 African countries enacted between 2015 and 2025. These 10 countries were selected because they tailored their data protection laws to incorporate specific protections for sensitive data, which emphasize data security, accountability, and data subject rights. A comparative analysis was conducted using Africa’s Malabo Convention, the U.S. code of federal regulation (CFR), and Europe’s General Data Protection Regulation (GDPR). Major issues identified include fragmented laws, weak enforcement, and insufficient provisions for user consent and data erasure. The study emphasizes the value of aligning regulatory efforts with the African Union’s framework, recognizing distinct African ethical standards rather than relying solely on international norms. Additionally, it explores how human-centred AI, particularly user autonomy, can ethically guide digital health implementation. The research contributes by outlining how harmonized regulations and AI designs adapted to local contexts can effectively address privacy challenges. | Chizoba Agwaonye and Neda Azarmehr | Late Breaking Work |