Finding: This scoping review (96 studies, 20 countries) finds that problematic internet use and anxiety in adolescents are bidirectionally linked, with emotion dysregulation and avoidance-based coping as the central shared mechanisms.
Why it matters: Tangentially relevant to behavior change design and user engagement/addiction-by-design topics, particularly the psychological mechanisms underlying compulsive technology use.
Method: Scoping review following Arksey and O'Malley's framework with PRISMA-ScR guidelines across three databases, 2014–2025.
Jargon: *Transdiagnostic process* = a psychological mechanism (here, emotion dysregulation) that cuts across multiple diagnostic categories rather than being specific to one disorder; *PIU* = Problematic Internet Use, characterized by loss of control over online activity with negative life consequences.
Eugénie Desmedt, Dominik Arnstorfer, Martina Mara • Semantic Scholar
Finding: In a field study with 117 participants, perceiving an AI poetry tool as a good team partner correlated with greater creativity support, and artistic background moderated authorship attribution—artists who felt strong team connection claimed *more* human authorship, while non-artists showed no such relationship.
Why it matters: Directly addresses HCI interface design, human-AI creative collaboration, authorship perception, and how user background shapes interaction—rich material for teaching creativity, AI's impact on creative work, and team dynamics with AI partners.
Method: Field deployment at a public arts festival using a custom instrument-like (non-conversational, embodied) LLM interface for poem creation; moderation analysis controlling for AI experience, creative self-efficacy, and perceived creativity support.
Jargon: *Syntactic Synthesizer*: a physical, parameter-manipulation interface for an LLM, deliberately non-anthropomorphized; *team experience*: perception of the AI as a collaborative partner rather than a tool; *creative self-efficacy*: belief in one's own creative ability.
⚠️ No control group mentioned | Single study (not replicated) | Limited statistical detail in abstract
Lukasz Bikowski, Khaled Elsayed, R. Daif et al. (4 authors) • Semantic Scholar
Finding: Firm-controlled cancellation barriers in digital subscription markets ("cancellation friction") erode consumers' practical ability to exercise formal cancellation rights, operating through exit-cost escalation and inertia exploitation, and the paper proposes a six-component entry-exit parity measurement framework plus five design standards.
Why it matters: Directly addresses dark patterns, choice architecture, and persuasive technology ethics — core topics in behavior change design and HCI — with concrete design standards that are teachable.
Method: Theoretical integration of transaction cost economics, behavioral economics, and HCI, with propositions evaluated by domain proximity and identification strategy; proposes (but does not validate) an operational measurement protocol.
Jargon: *Cancellation friction* = firm-designed barriers that increase the cost of canceling a subscription; *exit-capability erosion* = reduction in a consumer's practical (not just formal/legal) capacity to cancel; *passive renewal defaults* = subscriptions that auto-renew unless the user actively opts out; *entry-exit parity* = principle that canceling should be no harder than signing up.
D. Buzzo, Joep W. Frens, Kristina Andersen et al. (5 authors) • Semantic Scholar
Finding: This workshop paper proposes sharing "oblique" art-based pedagogical techniques (like walking exercises) used in HCI design education to surface hidden conceptual and critical thinking skills in students.
Why it matters: Directly relevant to HCI teaching practice and creative pedagogy—useful for a professor teaching HCI who may be looking for novel ways to teach design thinking, creativity, and critical reflection.
Method: Workshop format with shared practitioner examples; plans to produce a compendium of oblique teaching strategies.
Jargon: *Oblique approaches* — indirect, non-obvious instructional strategies that reveal underlying concepts through experience rather than direct instruction; *critical design pedagogy* — teaching design through questioning assumptions and power structures rather than just technical skills.
Finding: A field study of an affective computing system (EmoConnect) that uses facial expression analysis to provide continuous emotional feedback to call center managers, finding that facial-expression-derived fatigue correlated significantly with self-reported depression, negative mood, stress, and well-being.
Why it matters: Directly relevant to HCI/UX and behavior change design — this is a real-world deployment of persuasive/reflective technology aimed at promoting self-awareness and well-being, with ethical design considerations (consent, transparency, user-centered framing) that matter for teaching responsible technology design.
Method: Longitudinal field study (N=12 active users, 1–60 days), correlating continuous facial-expression-derived indicators with validated self-report scales (well-being, mood, stress, depression); naturalistic workplace setting distinguishes it from lab studies.
Jargon: *Affective computing* — technology that detects or responds to human emotional states; *ecological validity* — how well findings from a study reflect real-world conditions; *ethical-by-design* — building privacy and consent into the system architecture from the start rather than retrofitting.
Alan Aronica, Jamie Harper, Ed Finn et al. (8 authors) • Semantic Scholar
Finding: This workshop uses participatory co-design and Solarpunk narratives to explore sustainable AI deployment in natural environments, citizen ownership of AI systems, and data governance, using gamified activities including future scenarios, storyboards, and speculative design artifacts.
Why it matters: Directly relevant to HCI, participatory design, gamification as a design method, and AI's societal impact—all core teaching areas.
Method: Structured gamified co-design workshop with future scenario planning, narrative construction, and storyboarding as participatory design tools.
Jargon: *Solarpunk*: a speculative fiction/design movement imagining harmonious coexistence of humanity, technology, and nature. *EdgeAI*: AI processing done locally on devices at the "edge" of a network (e.g., in the field) rather than in centralized cloud servers. *Jokaisenoikeudet*: Finnish legal concept of universal public access to natural spaces, used here as a metaphor for democratic AI access.
⚠️ No control group mentioned | Limited statistical detail in abstract
Christopher J. Soto, Christopher Napolitano, Madison N. Sewell et al. (7 authors) • PsyArXiv
Finding: Proposes the Cumulative Practice Model (CPM), arguing that social, emotional, and behavioral (SEB) skills are better targets for intentional self-change than personality traits, developed through cycles of motivation, instruction, practice, and feedback.
Why it matters: The CPM framework maps closely onto learning sciences, skill acquisition, habit formation, and behavior change design—particularly relevant for designing interventions, coaching tools, or gamified systems that scaffold deliberate practice.
Jargon: *Volitional personality change* = intentional effort to change one's own personality; *SEB skills* = social, emotional, and behavioral capabilities framed as learnable rather than fixed traits; *deliberate practice* = structured, feedback-driven repetition aimed at skill improvement.
Laura Kanschat, J. Kuch, Daksitha Senel Withanage Don et al. (6 authors) • Semantic Scholar
Finding: A web-based voice design tool (NPVoice) for tabletop RPG Game Masters was prototyped and evaluated, revealing that AI voice generation works better for prepared static content (speeches) than dynamic improvised dialogue due to latency issues.
Why it matters: Directly relevant to HCI/UX design for creative tools, games/gamification, and AI's impact on creative collaborative work—specifically how AI augments human creativity in live storytelling contexts.
Method: Formative survey (60 participants) followed by prototype evaluation via 5 semi-structured interviews with experienced GMs.
⚠️ Very small sample (n=28) | Self-reported data only | No control group mentioned | Single study (not replicated) | Limited statistical detail in abstract
Finding: Proposes "Life Control" as a 10-dimension psychological framework integrating agency, self-regulation, motivation, resilience, and resource allocation, with a key "Non-Outsourcing Principle" predicting that assistance weakens capacity when it replaces rather than supports participation.
Why it matters: The Non-Outsourcing Principle has direct relevance to AI-assisted work, behavior change design, and learning sciences—specifically the question of when AI tools undermine vs. support human skill development and agency.
Jargon: *Non-Outsourcing Principle*: the idea that repeated delegation of tasks to external agents erodes the capacities those tasks would otherwise develop; *psychologically formative activities*: participation-based activities that build or maintain psychological capacities.
Annika Frommholz, S. Lepa, S. Weinzierl et al. (4 authors) • Semantic Scholar
Finding: Presents a data-driven framework for modeling how UX sounds communicate system status, affect, and brand identity, identifying recurring audio design patterns across 805 real-world sounds using unsupervised sequence modeling and explainable ML.
Why it matters: Relevant to HCI and UX design, specifically the underexplored domain of sonic interaction design, which could inform multimodal interface design principles.
Method: Unsupervised sequence modeling on 805 UX sounds coded across five design layers (texture, timbre, accentuation, rhythm, melody), with explainable ML linking patterns to semantic expression.
M. Lárusdóttir, Å. Cajander, Johanna Viitanen et al. (8 authors) • Semantic Scholar
Finding: This workshop paper argues that HCI considerations (usability, user involvement, work environment impacts) are underspecified in public IT procurement processes, particularly for AI and healthcare systems, and proposes frameworks for integrating HCI earlier in system lifecycles.
Why it matters: Tangentially relevant to AI impact on work and HCI design, but focuses narrowly on procurement policy and public-sector administration rather than the behavioral, creative, or cognitive dimensions central to the user's teaching areas.
Method: Workshop format drawing on the AI-PROCARE project, bringing together researchers, industry, and public-sector stakeholders to co-develop guidelines.
⚠️ No control group mentioned | Single study (not replicated) | Limited statistical detail in abstract
W. Zhou, Martina Vincoli, Willem van der Maden et al. (11 authors) • Semantic Scholar
Finding: This workshop paper examines challenges IxD/HCI educators face in keeping pace with GenAI, using participatory methods to surface shared resources and future directions.
Why it matters: Tangentially relevant to AI's impact on education and HCI teaching, but it's a workshop proposal focused on educator professional development rather than empirical findings on learning, behavior, or design.
Method: Participatory workshop design with collective knowledge-building activities.
⚠️ No control group mentioned | Single study (not replicated) | Limited statistical detail in abstract
A. Giachetti, Marco Emporio, H. Wannous et al. (4 authors) • Semantic Scholar
Finding: This tutorial covers methods and challenges for hand gesture recognition and generation in multimodal interactive systems, including HCI, extended reality, robotics, and embodied AI.
Why it matters: Gesture-based interfaces are relevant to HCI and UX design, particularly for extended reality and embodied interaction contexts, but this paper is deeply technical (tracking algorithms, recognition pipelines) rather than user-centered or behavioral.
Method: Tutorial format covering vision-based, sensor-based, and wearable hand tracking approaches with learning-based recognition methods.
⚠️ No control group mentioned | Single study (not replicated) | Limited statistical detail in abstract
Yang Yang, Akina Umemoto, Senne Braem et al. (4 authors) • PsyArXiv
Finding: Fatigue accumulation and psychiatric symptom dimensions (anxious-depression vs. compulsivity) predict distinct patterns of cognitive effort avoidance and task execution, suggesting motivational dysfunction is multi-staged rather than a simple reluctance to exert effort.
Why it matters: Has peripheral relevance to effort discounting and motivation research (similar to the "Effort discounting" paper the user liked), but focuses heavily on clinical/psychiatric populations rather than HCI, behavioral design, or learning contexts.
Method: Computational modeling of fatigue, effort selection, and task performance across time in a 75-min online experiment (n=529) using transdiagnostic symptom dimensions.
⭐ Longitudinal design ⚠️ No control group mentioned | Single study (not replicated) | Limited statistical detail in abstract
Ilaria Maria Antonietta Benzi, Francesca Ciardo • PsyArXiv
Finding: Epistemic stances (trust, mistrust, credulity toward social information sources) predict how people evaluate humanoid robots socially, but don't affect perceptual judgments like eeriness or likeability when viewing robot faces.
Why it matters: Marginally relevant to HCI and human-robot interaction, but the focus on individual difference psychometrics and robot evaluation is quite narrow and distant from the user's core interests.
Finding: Future Workshops with pregnancy loss community users identified unmet needs around harmful engagement patterns, community parameters, and community discovery, yielding sociotechnical design recommendations for online support spaces.
Why it matters: Tangentially relevant to HCI and behavior change design (dark patterns, ethical persuasion, user engagement), but the domain is highly specific and the design insights are narrow.
Method: Future Workshops (a participatory design method) + reflexive thematic analysis with N=14 participants.
⚠️ Very small sample (n=14) | No control group mentioned | Single study (not replicated) | Limited statistical detail in abstract
A. Benedetti, Lina Eklund, S. Grimme et al. (9 authors) • Semantic Scholar
Finding: This workshop paper proposes using speculative co-design methods to explore collaborative technologies supporting early parenthood across diverse stakeholders, critiquing the individual/mother-centric focus of current HCI and personal informatics tools.
Why it matters: Tangentially relevant to HCI design and collaborative technology, but the domain focus (early parenthood) is too narrow and applied to connect meaningfully to teaching interests in creativity, teams, behavior change, or broader HCI.
Finding: A model (LGSM) can infer an individual's engagement and leadership status solely from observing their teammates' multimodal behavior, without directly sensing the target person.
Why it matters: Tangentially relevant to team dynamics and engagement measurement, but the focus is heavily on sensor/ML methodology rather than behavioral science, HCI design, or collaboration insights.
Method: Latent Group State Modeling encodes teammates' multimodal signals into a shared latent representation, then performs target-conditioned inference evaluated on a three-person team dataset.