Do Entropic Measurements of the Diversity of AI-generated Images Match Human Judgement? This paper proposes that the ability to generate diverse outputs in response to a single prompt is necessary for text-to-image models to become more effective creativity support tools. It formalises the problem of measuring the diversity of generated text and images, with an emphasis on interactive, exploratory use in open-ended and creative tasks. It suggests, motivated by research in the psychology of creativity, that diversity should sit alongside image quality and fit-to-prompt as critical measures in this setting. The paper adapts several diversity measures from the literature to this task, then explores how they compare to human diversity ratings. These evaluations show that algorithmic measures of diversity can be a useful proxy for human ratings, with both declining in accuracy as the difficulty of the task increases. The paper concludes with an exploratory qualitative analysis of the factors involved in human diversity judgments to guide future research in this emerging area.2026KGKazjon Grace et al.The University of SydneyGenerative AI (Text, Image, Music, Video)Explainable AI (XAI)Creative Collaboration & Feedback SystemsCHI
Enabling Partial Participation in Remote MeetingsWe propose and explore the concept of Partial Participation, facilitating remote collaborators to contribute to meetings in which they are not able to fully participate via an AI agent acting as a proxy. During the meeting, users can monitor LLM-generated real-time meeting updates and respond to questions posed by other attendees. Through a mixed-methods user study with 24 participants using our prototype, ProxyMe, we investigated how the frequency of updates (high vs. low) and the type of response style (multiple choice vs. text input) impact perceived presence and mental workload. Our findings reveal that no single setup is universally optimal, and the partial participation fosters a moderate level of social presence and attentional mental workload. Our contributions introduce partial participation as a new paradigm for remote collaboration and highlight how AI can mediate participation when full presence is not feasible.2026ZBZhongyi Bai et al.University of SydneyRemote Work Tools & ExperienceDistributed Team CollaborationHuman-LLM CollaborationCHI
Narratives and Perspectives: How AI Summaries Steer Users' Opinions and Engagement on Social MediaAI summaries on social media are reshaping how users form opinions about political topics, yet their influence remains largely unexamined despite their widespread deployment. This paper investigates how two types of AI summaries affect user opinions and engagement: textual summaries of discussion narratives and percentage breakdowns of agreement/disagreement. Through a 144-participant experiment on simulated online discussion threads, we found that displaying commenter agreement percentages amplified social conformity towards the majority views beyond reading comments alone. Conversely, AI narrative summaries created misperceptions of balance in polarised threads, reducing opinion change. While these summaries did not influence participants’ willingness to engage, toxic discussions deterred participation even when participants held majority views. Based on our findings, we provide critical design interventions for industry and researchers to mitigate these tools' polarising effects, paving the way for responsible AI deployment on social media platforms.2026JGJarod Govers et al.University of MelbourneConversational ChatbotsMisinformation & Fact-CheckingAI Ethics, Fairness & AccountabilityCHI
Gaze and Speech in Multimodal Human-Computer Interaction: A Scoping ReviewMultimodal interaction has long promised to make interfaces more intuitive and effective by combining complementary inputs. Among these, gaze and speech form a compelling pairing: gaze provides rapid spatial grounding, while speech conveys rich semantic information. Together, they offer rich cues for understanding user behaviour and intent. Yet despite decades of exploration, the research remains fragmented, making this synthesis timely as these inputs mature and are integrated into consumer-ready devices. This scoping review examined 103 studies published between 1991 and 2025, organised into \emph{explicit}, where users intentionally provide gaze and speech, and \emph{implicit}, where systems leverage users' natural behaviours to support interaction. Across both, we identified recurring ways for combining gaze and speech to resolve ambiguity, ground references, and support adaptivity. We contribute a synthesis of research on their combined use while highlighting challenges of temporal alignment, fusion and privacy, offering guidance for future research toward richer multimodal human-computer interaction.2026AKAnam Ahmad Khan et al.KAISTEye Tracking & Gaze InteractionVoice User Interface (VUI) DesignAffective Human-Computer DialogueCHI
Restoration, Exploration and Transformation: How Youth Engage Character.AI for Fun, Feels and Finding themselvesYoung people are among the fastest adopters of generative AI, yet research emphasises adult-designed tools and experiments rather than playful, self-directed youth use. We analysed discourse from 4,172 users in Character.AI’s official Discord, finding that the most engaged users were predominantly adolescents (50% aged 13–17), female or non-binary (61.9%), with most (59%) creating their own characters. We contribute (1) a descriptive account of how highly-engaged youth on Character.AI's Discord use AI for playful, emotional, and creative practices that push the platform limits; (2) a framework of three engagement intents — Restoration (emotional regulation), Exploration (creative experimentation), and Transformation (identity development); and (3) a taxonomy of seven youth-created character archetypes. Together, these findings reveal how youth invent novel roles for AI, expose critical misalignments between youth use and current AI experiences, and provide frameworks for researchers and practitioners to design youth-centred AI futures.2026ABAnnabel Blake et al.The University of SydneyGenerative AI (Text, Image, Music, Video)Human-LLM CollaborationAI Ethics, Fairness & AccountabilityCHI
Sensemaking in Multi-Agent LLM Interfaces: How Users Interpret Transparency and Trustworthiness CuesAs multi-agent Large Language Models (LLMs) gain traction, designers must consider how to surface their internal reasoning in ways that foster appropriate trust. We present a design-led, qualitative, comparative structured observation study, exploring how users interpret and evaluate transparency in multi-agent LLMs. Participants interacted with five interface variants, each instantiating different combinations of transparency-related design dimensions, across two task types: information-seeking and logical reasoning. We surface participants’ mental models, the cues they interpret as signals of transparency and trustworthiness, and how they weigh the costs and benefits of increasing process visibility. Transparency needs were dynamic and context-sensitive, with the ideal "Goldilocks" (i.e., "just right" transparency) level shaped jointly by task demands, interface affordances, and user characteristics such as task expertise and dispositional AI trust. We highlight tensions between process visibility, information sufficiency, and cognitive effort, and synthesise these insights into design considerations for aligning transparency with user needs in future multi-agent LLM interfaces.2026SPSaumya Pareek et al.University of MelbourneHuman-LLM CollaborationExplainable AI (XAI)Privacy by Design & User ControlCHI
One Body, Two Minds: Alternating VR Perspective During Remote Teleoperation of Supernumerary LimbsRemote VR teleoperation with supernumerary robotic limbs enables distant users to operate in another’s local space. While a shared first-person view aids hand-eye coordination, locking the guest’s camera to the host’s head can degrade comfort, embodiment, and coordination. Based on a formative study (N=10) using a virtual supernumerary robotic limbs configuration to stress-test coordination, we propose guest-driven perspective switching from a shared first-person baseline (Shared Embodied View) to two alternatives: (a) a stabilized view with guest-controlled rotation (Embedded Anchored View), and (b) a fully decoupled third-person view (Out-of-body View). We ran a user study with 24 pairs (N=48), who switched between the baseline and proposed views as task demands changed. We measured performance, embodiment, fatigue, physiological arousal, and switching behaviors. Our results reveal role-dependent trade-offs: Out-of-body View improves navigation efficiency and reduces errors, while Embedded Anchored View supports embodiment. We conclude with guidelines: use Embedded Anchored View for hand-centric adjustments, Out-of-body View for navigation and object placement, and ensure smooth transitions.2026HZHongyu Zhou et al.The University of SydneyTeleoperation & TelepresenceSocial & Collaborative VRImmersion & Presence ResearchCHI
Is That You or The Machine? Translating Sociocultural Norms Across Distributed Spaces in Blended RealitiesWhen distributed mixed reality (MR) systems map physical spaces to enable co-presence of local and remote collaborators, they can unintentionally disrupt the sociocultural norms that give actions their meaning. For instance, a participant sitting at their own desk may be rendered as occupying their collaborator’s desk, inadvertently signalling an invasion of personal space. This paper examines the design tension between spatial information and sociocultural norms through a qualitative counterfactual cards activity with 20 participants, probing how they navigate these trade-offs across different collaborative contexts. Our findings show similarities between Expectancy Violations Theory (EVT) and the factors participants assess when they decide to uphold accurate spatial information or sociocultural norms during collaboration in MR. However, there were some departures from EVT, which we use to propose design implications and the development of MR-specific theories in the future.2026EWEmily Wong et al.The University of SydneyMixed Reality WorkspacesImmersion & Presence ResearchIdentity & Avatars in XRCHI
Better Assumptions, Stronger Conclusions: The Case for Ordinal Regression in HCIDespite the widespread use of ordinal measures in HCI, such as Likert-items, there is little consensus among HCI researchers on the statistical methods used for analysing such data. Both parametric and non-parametric methods have been extensively used within the discipline, with limited reflection on their assumptions and appropriateness for such analyses. In this paper, we examine recent HCI works that report statistical analyses of ordinal measures. We highlight prevalent methods used, discuss their limitations and spotlight key assumptions and oversights that diminish the insights drawn from these methods. Finally, we champion and detail the use of cumulative link (mixed) models (CLM/CLMM) for analysing ordinal data. Further, we provide practical worked examples of applying CLM/CLMMs using R to published open-sourced datasets. This work contributes towards a better understanding of the statistical methods used to analyse ordinal data in HCI and helps to consolidate practices for future work.2026BSBrandon Victor Syiem et al.University of SydneyUser Research Methods (Interviews, Surveys, Observation)Computational Methods in HCIResearch Ethics & Open ScienceCHI
Weight-Induced Consumed Endurance (WICE): A Model to Quantify Shoulder Fatigue with Weighted ObjectsFatigue is a major challenge in mid-air interactions, often resulting in a sensation of heaviness––particularly when users carry weighted objects on their arms. Existing models for characterising shoulder fatigue were primarily developed for bare-hand scenarios, limiting their applicability in situations involving encumbrance. In this paper, we introduce Weight-Induced Consumed Endurance (WICE), a novel model that accurately estimates shoulder fatigue when additional weight is attached at various locations on the arm. WICE enhances the calculation of instantaneous shoulder torque by incorporating information about the attached weight, integrates individual arm mass for more personalised fatigue estimation, and uses a Bayesian framework to simulate the distribution of shoulder fatigue. Our evaluation shows that WICE strongly correlates with both experimentally measured endurance time and subjective Borg CR10 ratings, demonstrating its reliability as an objective fatigue metric in both encumbered and no-weight conditions. We further demonstrate how WICE can be applied to examine the effects of controller and haptic devices on user fatigue. WICE provides a foundation for developing fatigue-aware systems that can sense and adapt encumbrance, allowing for more tailored ergonomic MR interactions.2025TLTinghui Li et al.Force Feedback & Pseudo-Haptic WeightFull-Body Interaction & Embodied InputBiosensors & Physiological MonitoringUIST
How your Physical Environment Affects Spatial Presence in Virtual RealityVirtual reality (VR) is often used in small physical environments, requiring users to remain aware of their environment to avoid injury or damage. However, this can reduce their spatial presence in VR. Previous work and theory lack an account of how the physical environment (PE) affects spatial presence. To address this gap, we investigated the effect on spatial presence of (1) the degree of spatial knowledge of the PE and (2) knowledge of and (3) collisions with obstacles in the PE. Estimates from Bayesian regression models suggest that limiting spatial knowledge of the PE increases spatial presence initially but amplifies the detrimental effect of obstacle collisions. Repeatedly avoiding obstacles further decreases spatial presence, but removing them from the user's path yields a partial recovery. Our work contributes empirical evidence to theories of spatial presence formation and highlights the need to consider the physical environment when designing for presence in VR.2025TGThomas van Gemert et al.University of Copenhagen, Department of Computer ScienceMixed Reality WorkspacesImmersion & Presence ResearchContext-Aware ComputingCHI
Theorising in HCI using Causal ModelsAlthough the literature on Human-Computer Interaction (HCI) catalogues many theories, it offers surprisingly few tools for theorising. This paper critiques dominant approaches to engaging with theory and proposes a working model for theorising in HCI. We then present graphical causal modelling as an effective theorising tool. This includes a step-by-step guide to building causal models and examples of their use in different stages of the research process. We explain how causal models help develop method-agnostic representations of research problems using directed acyclic graphs, identify potential confounders, and construct alternative interpretations of data. Finally, we discuss their limitations and challenges for adoption by the HCI community.2025EVEduardo Velloso et al.University of Sydney, School of Computer ScienceExplainable AI (XAI)Computational Methods in HCICHI
"It’s Not the AI’s Fault Because It Relies Purely on Data": How Causal Attributions of AI Decisions Shape Trust in AI SystemsHumans naturally seek to identify causes behind outcomes through causal attribution, yet Human-AI research often overlooks how users perceive causality behind AI decisions. We examine how this perceived locus of causality—internal or external to the AI—influences trust, and how decision stakes and outcome favourability moderate this relationship. Participants (N=192) engaged with AI-based decision-making scenarios operationalising varying loci of causality, stakes, and favourability, evaluating their trust in each AI. We find that internal attributions foster lower trust as participants perceive the AI to have high autonomy and decision-making responsibility. Conversely, external attributions portray the AI as merely "a tool" processing data, reducing its perceived agency and distributing responsibility, thereby boosting trust. Moreover, stakes moderate this relationship—external attributions foster even more trust in lower-risk, low-stakes scenarios. Our findings establish causal attribution as a crucial yet underexplored determinant of trust in AI, highlighting the importance of accounting for it when researching trust dynamics.2025SPSaumya Pareek et al.University of Melbourne, School of Computing and Information SystemsExplainable AI (XAI)AI Ethics, Fairness & AccountabilityPrivacy by Design & User ControlCHI
Responsibility Attribution in Human Interactions with Everyday AI SystemsHow do individuals perceive AI systems as responsible entities in everyday collaborations between humans and AI? Drawing on psychological literature from attribution theory, praise-blame asymmetries and negativity bias, this study investigated the effects of perspective (actor vs observer) and outcome favorability (positive vs negative) on how participants (N=321) attributed responsibility for outcomes resulting from shared human-AI decision-making. Both Bayesian modelling and reflexive thematic analysis of results revealed that, overall, participants were more likely to attribute greater responsibility to the AI systems. When the outcome was positive, participants were more likely to ascribe shared responsibility to both Human and AI systems, rather than either separately. When the outcome was negative, participants were more likely to attribute responsibility to a single entity, but not consistently towards the human or the AI. These results build on the understanding of how individuals cast blame and praise for shared interactions involving AI systems.2025JBJoe Brailsford et al.The University of Melbourne, School of Computing and Information SystemsAI Ethics, Fairness & AccountabilityAlgorithmic Fairness & BiasCHI
Spatial Heterogeneity in Distributed Mixed Reality CollaborationCollaborative Mixed Reality (MR) enables embodied meetings for distributed collaborators working across a variety of locations. However, providing a coherent experience for all users regardless of the spatial configurations of their respective physical environments is a central challenge. We present the Spatial Heterogeneity Framework, which breaks the problem into four core components: the activity zones, heterogeneity ladder, blended proxemics, and MR solutions matrix. We explain the interplay between these components, demonstrating their interconnectivity via a case study. Our framework enables researchers to navigate differences and trade-offs between solutions for distributed MR collaboration. It also supports designers to think about the role of space, technology, and social behaviours in MR collaboration. Ultimately, our contributions advance the field by conceptualising the challenges of spatial heterogeneity and strategies to overcome them.2025EWEmily Wong et al.The University of Sydney, School of Computer Science; The University of Melbourne, School of Computing and Information SystemsMixed Reality WorkspacesContext-Aware ComputingCHI
Estimating the Effects of Encumbrance and Walking on Mixed Reality InteractionThis paper investigates the effects of two situational impairments---encumbrance (i.e., carrying a heavy object) and walking---on interaction performance in canonical mixed reality tasks. We built Bayesian regression models of movement time, pointing offset, error rate, and throughput for target acquisition task, and throughput, UER, and CER for text entry task to estimate these effects. Our results indicate that 1.0 kg encumbrance increases selection movement time by 28%, decreases text entry throughput by 17%, and increase UER by 50%, but does not affect pointing offset. Walking led to a 63% increase in ray-cast movement time and a 51% reduction in text entry throughput. It also increased selection pointing offset by 16%, ray-cast pointing offset by 17%, and error rate by 8.4%. The interaction effect on 1.0 kg encumbrance and walking resulted in a 112% increase in ray-cast movement time. Our findings enhance the understanding of the effects of encumbrance and walking on mixed reality interaction, and contribute towards accumulating knowledge of situational impairments research in mixed reality.2025TLTinghui Li et al.University of Sydney, School of Computer ScienceFull-Body Interaction & Embodied InputMixed Reality WorkspacesCHI