📚 Research Digest

Finding: Physical coupling between soft Vine robots and rigid RESCUE roller robots enables heterogeneous robot teams to collaborate by sharing power, communication, and complementary capabilities through mechanical and electrical interfaces.

Why it matters: This demonstrates how different types of agents (robots) can overcome individual limitations through physical collaboration, offering insights for designing human-AI teams and collaborative systems.

Method: Experimental validation measuring pushing/pulling strength, buckling loads, and field demonstrations of collaborative behaviors in cluttered environments.

Jargon: Vine robot - soft, flexible robot that can navigate tight spaces; RESCUE rollers - small rigid robots that can form chains and provide precise actuation.

Finding: School management teams play a central role in developing school-based curriculum, but effective curriculum development requires collaborative co-construction with both internal stakeholders (teachers) and external stakeholders, with principals' understanding of curriculum being a key influencing factor.

Why it matters: This demonstrates how collaborative stakeholder involvement and team dynamics directly impact educational outcomes, providing insights into effective collaboration structures for curriculum design and implementation.

Method: Mixed methods explanatory sequential design combining quantitative surveys of 279 educators with qualitative interviews, analyzed using confirmatory factor analysis and thematic analysis.

Jargon: School-based curriculum (SBC) - locally adapted curriculum developed by schools; School management team (SMT) - administrative leadership team in schools.

Finding: This study analyzes 7,000 job postings to show that AI skills are now required in 27.8% of knowledge work positions (376% growth since ChatGPT's release) and proposes a three-dimensional performance evaluation model (AI Tool Mastery, Collaborative Work Quality, Human-AI Synergy) to measure human-AI collaborative work effectiveness.

Why it matters: This directly addresses how AI is transforming workplace performance measurement and collaboration patterns, providing concrete frameworks for understanding human-AI teamwork dynamics.

Method: Systematic analysis of LinkedIn job advertisements and Indeed salary data from 2022-2024 to track skill requirement changes post-ChatGPT.

Jargon: Human-AI Synergy - the combined effectiveness when humans and AI systems work together rather than independently.

Finding: This study analyzes how high school music achievement standards relate to digital literacy development using Bloom's Digital Taxonomy and explores instructional strategies that integrate digital tools into music education.

Why it matters: Provides insights into how traditional subject areas can be enhanced through digital integration, which is relevant to understanding technology's impact on learning and educational design.

Method: Expert analysis approach examining achievement standards through the lens of Bloom's Digital Taxonomy framework.

Jargon: Bloom's Digital Taxonomy - an updated version of Bloom's educational taxonomy that incorporates digital tools and skills into the traditional learning hierarchy.

Finding: AI-enabled UAVs can improve communication network coverage by 35% and reduce energy consumption by 20% in remote areas by dynamically managing IoT sensor connections across two-dimensional spaces.

Why it matters: This represents an application of AI to infrastructure challenges but has limited direct relevance to human-centered design, collaboration, or behavioral aspects of technology use.

Jargon: UAVs = Unmanned Aerial Vehicles (drones); IoT = Internet of Things (network of connected sensors/devices)

Finding: This study reveals how university students' motivation emerges from a complex negotiation between intrinsic desires (curiosity, competence, autonomy) and external pressures (grades, parental expectations), with six key motivational configurations identified through interviews with 25 Malaysian and Chinese students. The research integrates Self-Determination Theory with cognitive perspectives to show how timely, effort-focused feedback and autonomy support help students internalize extrinsic demands.

Why it matters: This directly informs teaching practices around motivation, feedback design, and goal-setting while providing cross-cultural insights into how different motivational factors interact in educational contexts.

Method: Semi-structured interviews with purposive sampling across institutions, analyzed using thematic analysis to identify emergent motivational patterns.

Jargon: Self-Determination Theory (SDT) - framework identifying autonomy, competence, and relatedness as core psychological needs; expectancy-value theory - cognitive model linking expectations of success and task value to motivation.

Finding: Managers sometimes evaluate subordinates with dark personality traits (Machiavellianism, psychopathy, narcissism) positively because they perceive these individuals as facilitating agentic goals, with dominance-oriented managers being most likely to show this bias.

Why it matters: This reveals how manager biases and goal orientations can systematically distort performance evaluations and team composition decisions, directly impacting team dynamics and organizational behavior.

Method: Two studies with 1,209 managers evaluating job candidates and rating actual subordinates, testing dual-strategies theory predictions about dominance vs. prestige-oriented leadership styles.

Jargon: Agentic goals = achievement-focused objectives emphasizing individual success and task completion; Dark personality traits = antisocial characteristics including manipulation, lack of empathy, and grandiosity.

Finding: This literature review identifies four key factors (ethical leadership, digital competence, work motivation, and organizational commitment) that influence human resource readiness for organizational transformation in the Indonesian National Police, emphasizing how digital competence and psychological empowerment support adaptation to technological change.

Why it matters: The findings on digital competence, work motivation, and organizational commitment during technology adoption provide insights relevant to understanding how teams adapt to AI and digital tools in work environments.

Method: Literature review synthesizing findings from national and international publications over the last decade.

Jargon: Polri Presisi - Indonesian National Police precision/modernization transformation program; psychological empowerment - giving employees autonomy and meaningful work to increase intrinsic motivation.

Finding: LLMs demonstrate 110-300 times greater resistance to emotional framing effects in high-stakes decisions compared to humans, showing near-zero bias (Cohen's h = 0.003) versus substantial human biases (Cohen's h = 0.3-0.8) across healthcare, law, and finance domains.

Why it matters: This reveals critical insights about AI's potential to complement or de-bias human decision-making, with major implications for human-AI collaboration design and understanding cognitive biases in high-stakes contexts.

Method: Novel controlled perturbation framework testing emotional framing effects across 162 scenarios in three professional domains, comparing LLM versus human responses.

Jargon: Cohen's h - effect size measure for comparing proportions; sycophantic alignment - AI tendency to agree with user preferences regardless of correctness.

Finding: Work arrangement autonomy (flexibility in when/where to work) has a negative direct effect on job performance but a positive indirect effect through perceived procedural justice, organizational commitment, and intrinsic motivation.

Why it matters: This explains the psychological mechanisms behind remote work effectiveness and provides evidence for designing workplace policies that balance autonomy with performance outcomes.

Method: Survey of 322 US employees using sequential mediation modeling to test direct and indirect pathways from work autonomy to performance.

Jargon: Self-determination theory (SDT) - theory that autonomy, competence, and relatedness drive intrinsic motivation; Social exchange theory (SET) - framework explaining reciprocal relationships between employees and organizations.

Finding: This study develops an AR framework for teachers based on Gagne's nine events of instruction and self-determination theory, identifying five key constructs (motivation, technology skills, instructional design, AR development tools, and AR application types) needed to enhance student motivation through AR learning experiences.

Why it matters: This directly addresses how emerging technology (AR) can be designed to support motivation and learning, providing practical frameworks for educators working at the intersection of HCI and education.

Method: Uses design and development research (DDR) methodology with mixed methods, surveying 35 technology teachers through purposive sampling with closed and open questionnaires.

Jargon: DDR (Design and Development Research) - a systematic approach to creating educational interventions and frameworks; Self-Determination Theory - psychological theory focusing on intrinsic motivation through autonomy, competence, and relatedness; Gagne's Nine Events - instructional design framework covering attention, objectives, recall, content, guidance, practice, feedback, assessment, and retention.

Finding: This mixed-methods study found that technology acceptance factors (performance expectancy, effort expectancy, social influence, facilitating conditions, hedonic motivation, and habit) significantly influenced EFL students' intention to use generative AI for English learning, while self-determination theory components (autonomy, competence, relatedness) predicted actual usage but not behavioral intention.

Why it matters: This demonstrates how psychological needs and technology acceptance models interact differently at intention versus behavior stages, providing crucial insights for designing AI learning tools that sustain engagement.

Method: Used explanatory sequential mixed-methods with PLS-SEM analysis of 462 student surveys followed by thematic analysis of 15 semi-structured interviews.

Jargon: UTAUT2 = Unified Theory of Acceptance and Use of Technology 2 (framework for predicting technology adoption); SDT = Self-Determination Theory (motivation theory based on autonomy, competence, relatedness needs); PLS-SEM = Partial Least Squares Structural Equation Modeling (statistical analysis method).

Finding: This study examines how China's "Double Reduction" policy creates opportunities to shift high school education from external control to student autonomous development, identifying key barriers (insufficient intrinsic motivation, path dependency in management, lack of teacher guidance skills) and proposing a five-dimensional framework for cultivating student self-management competency.

Why it matters: This directly addresses motivation, goal-setting, and behavioral change in educational contexts, offering insights into how institutional policies can support or hinder student autonomy and self-regulation skills.

Method: Case study analysis of a provincial exemplary high school's reform practices combined with theoretical framework development.

Jargon: "Double Reduction" policy - Chinese education reform aimed at reducing homework burden and after-school tutoring; "connotative development" - internal quality improvement rather than external expansion.

Finding: The paper presents eBrainISA, a hardware architecture designed to run hybrid brain-inspired computing models (combining artificial and spiking neural networks) efficiently on resource-constrained edge devices, achieving significant memory and energy savings.

Why it matters: This technical architecture work has minimal direct relevance to HCI, creativity, teams, or behavioral science - it's primarily a computer engineering solution for hardware efficiency rather than addressing human factors or collaborative aspects of AI systems.

Jargon: ISA = Instruction Set Architecture (the interface between software and hardware); SNNs = Spiking Neural Networks (brain-inspired models that process information through discrete spikes); HNNs = Hybrid Neural Networks (combining traditional and spiking approaches).

Finding: Survey of 380 Chinese undergraduate students reveals strong student readiness and preference for metaverse-based blended learning in English education, with students favoring collaborative and simulative learning environments despite limited familiarity with metaverse technologies.

Why it matters: Provides empirical evidence for designing immersive learning experiences and understanding student preferences for collaborative, technology-enhanced educational environments.

Method: Cross-sectional survey measuring learning preferences and technological readiness among undergraduate students in Southwest China.

Jargon: MBBL (metaverse-based blended learning) - combining traditional instruction with immersive virtual world technologies; EFL (English as a foreign language) - English instruction for non-native speakers.

Finding: FACTY, a GPT-based chatbot for teaching algebraic factoring, provided effective process-level and self-regulation feedback to engineering students, with success varying based on students' foundational knowledge gaps.

Why it matters: This demonstrates how AI tutoring systems can complement human instruction through specific feedback mechanisms, directly relevant to AI's impact on learning and educational technology design.

Method: Mixed-methods study with 10 first-year engineering students across four stages including assessments, autonomous practice, and interviews, analyzed through a mathematics feedback framework.

Jargon: Process-level feedback = guidance on procedures and strategies; Self-regulation-level feedback = helping students monitor and direct their own learning.

Finding: The DIVE Model provides a six-phase framework for creating interactive videos that transform passive video consumption into active experimental learning through embedded tasks, clickable elements, and instant feedback, specifically designed for online science education.

Why it matters: This directly addresses how educational technology design can enhance learning through interactivity and engagement, providing a concrete framework for designing better online learning experiences.

Method: Develops a systematic design model grounded in Cognitive Load Theory, Dual Coding Theory, and Constructivism for creating interactive educational videos.

Jargon: DIVE Model - six-phase framework for Designing Interactive Videos for Experimentation; Cognitive Load Theory - how working memory limitations affect learning; Dual Coding Theory - processing visual and verbal information through separate channels.

Finding: An online flipped classroom using Interactive Response Systems (IRS) produced significantly greater vocabulary learning gains than traditional paper-based flipped classrooms among low-proficiency EFL students, though students reported internet connectivity issues and learning anxiety with the digital approach.

Why it matters: This demonstrates how interactive technology can enhance learning outcomes in flipped classroom designs, providing evidence for educational technology decisions and understanding student experience trade-offs in digital learning environments.

Method: Two-cycle classroom action research with 46 students comparing paper-based vs. online IRS flipped classrooms, using pre/post vocabulary tests and thematic analysis of student responses.

Jargon: Interactive Response System (IRS) - digital platforms that allow real-time student participation and feedback during lessons; Flipped Classroom (FC) - pedagogical model where students engage with content outside class and do "homework" activities in class.

Finding: Data-Driven Learning (DDL) approaches significantly outperformed conventional teaching methods for ESL grammar instruction among Filipino secondary students, but implementation faces barriers including limited technological access and inadequate teacher training.

Why it matters: This demonstrates how technology-enhanced learning methods can improve educational outcomes while revealing critical implementation challenges around teacher preparation and digital equity that are relevant to HCI and learning sciences research.

Method: Quasi-experimental pretest-posttest design comparing DDL vs. conventional instruction, supplemented by focus groups with teachers and students to understand practical implementation barriers.

Jargon: Data-Driven Learning (DDL) - pedagogical approach using corpus linguistics and real language data to teach grammar patterns; Global South - developing countries, primarily in Africa, Asia, and Latin America.

Finding: Outdoor mathematics instruction significantly improved grade 7 students' understanding of measurement concepts compared to conventional classroom instruction, and students reported positive perceptions of the outdoor activities.

Why it matters: This demonstrates how alternative learning environments can enhance engagement and skill acquisition, relevant to learning sciences and educational innovation.

Method: Two-group quasi-experimental design comparing outdoor vs. conventional classroom instruction using a 40-item test and perception measures.

Jargon: Quasi-experimental design - research method that compares groups but lacks random assignment, making it less controlled than true experiments.

Finding: LLM Shepherding uses large language models to provide short "hints" (10-30% of full response) to small language models rather than complete answers, achieving 42-94% cost reduction while maintaining accuracy on math and coding tasks.

Why it matters: This demonstrates a novel human-AI collaboration paradigm where AI systems provide guidance rather than solutions, with implications for educational scaffolding, creative support tools, and cost-effective AI assistance in learning and work contexts.

Method: Develops a two-stage predictor that determines when hints are needed and optimal hint length, tested on mathematical reasoning and code generation benchmarks.

Jargon: SLM = Small Language Model (cheaper, less capable); LLM = Large Language Model (expensive, more capable); routing/cascading = existing methods for combining different AI models.

Finding: This paper introduces a benchmark for testing AI agents in enterprise systems, revealing that frontier LLMs fail to predict cascading side effects of their actions in complex workflows due to "dynamics blindness."

Why it matters: The findings about AI limitations in complex systems and the need for better world modeling relate to AI's impact on work environments and human-AI collaboration challenges.

Jargon: Dynamics blindness - AI's inability to predict indirect consequences of actions in complex systems; World modeling - mentally simulating system state changes.

Finding: This paper presents a data-driven method to define Operational Design Domains (ODDs) for safety-critical AI systems using kernel-based representations, moving away from traditional expert-based domain definition to automated extraction from existing data.

Why it matters: This addresses a key challenge in AI impact on work and society - how to certify and deploy AI systems safely in complex real-world environments where human expertise may be incomplete or insufficient.

Method: Uses multi-dimensional kernel-based representation with Monte Carlo validation and real-world aviation collision-avoidance case study to demonstrate the approach.

Jargon: ODD (Operational Design Domain) - the specific conditions and environment in which an AI system is designed to operate safely; Safety-by-Design - methodology ensuring safety considerations are built into system design from the start.

Finding: Spatially localized voice and environmental Foley audio can create "auditory embodiment" for conversational agents, increasing user sense of co-presence but paradoxically reducing perceptions of the agent's attention and social connection.

Why it matters: This reveals important HCI design tradeoffs where realistic audio cues enhance presence but may hurt social rapport, critical for designing effective AI assistants and virtual collaboration tools.

Method: Within-subjects experiment (n=24) testing 2x2 conditions of spatialization (monaural vs. spatialized) and environmental audio (none vs. Foley sounds) during agent conversations.

Jargon: Foley audio = artificially created sound effects that mimic real environmental sounds; spatialization = audio positioning that creates sense of directional location.

Finding: Teams with blind/low-vision and sighted workers must constantly transform inaccessible information (PDFs, diagrams) into accessible formats, revealing four coordination patterns and highlighting how current systems create additional labor burdens rather than supporting inclusive collaboration.

Why it matters: This directly informs HCI design for inclusive teamwork and reveals how accessibility needs drive team dynamics and collaborative processes that most design frameworks overlook.

Method: Week-long diary study with 23 professionals across mixed-visual ability teams, documenting 36 real transformation cases with follow-up interviews.

Jargon: Mixed-visual ability teams = teams including both blind/low-vision (BLV) and sighted workers; representational transformations = converting information from inaccessible to accessible formats.

Finding: The paper proposes per-parameter task arithmetic (PerTA) for selectively removing private information from large language models while preserving overall model utility, using parameter-specific weights to avoid over-forgetting during the unlearning process.

Why it matters: This directly addresses AI's impact on privacy and data governance in work contexts, showing how to maintain useful AI capabilities while protecting sensitive information—a critical concern as AI tools become ubiquitous in collaborative work environments.

Method: The approach rescales task vectors using gradients or Fisher information to estimate parameter importance, then selectively adjusts parameters for forgetting versus retention rather than applying uniform changes.

Jargon: Task vector (TV) = parameter difference between a fine-tuned model and original model; Unlearning = removing specific information from trained models; Fisher information = measure of parameter sensitivity to model output changes.

Finding: Generic "improved" prompt templates can hurt LLM performance on specific tasks - extraction pass rates dropped from 100% to 90% and RAG compliance from 93% to 80% when switching from task-specific to generic prompts, demonstrating that universal prompt recipes don't exist.

Why it matters: This directly impacts how humans should collaborate with AI tools, showing that iterative evaluation and customization are essential for effective human-AI workflows rather than relying on one-size-fits-all solutions.

Method: Introduces a Define-Test-Diagnose-Fix workflow and Minimum Viable Evaluation Suite (MVES) for systematic LLM application testing, with reproducible experiments using local models.

Jargon: RAG = Retrieval-Augmented Generation (AI systems that combine information retrieval with text generation); LLM-as-judge = using language models to evaluate other language model outputs; Agentic workflows = AI systems that can use tools and take actions autonomously.

Finding: Vidmento is a tool that helps video creators blend captured footage with AI-generated content through context-aware expansion, enabling them to fill narrative gaps and enhance creative expression in video storytelling.

Why it matters: This demonstrates how AI tools can augment rather than replace human creativity, providing insights into designing collaborative AI systems that support creative workflows and enhance human creative expression.

Method: The research combined interviews with 8 video creators to understand workflow needs, followed by a user study with 12 creators testing the developed tool.

Jargon: Context-aware expansion - AI generation that considers surrounding content to create stylistically and narratively coherent additions; Hybrid video stories - videos combining human-captured and AI-generated content.

Finding: AI agent inference creates new computational bottlenecks beyond traditional compute limitations, particularly in memory capacity and bandwidth, requiring specialized heterogeneous computing architectures to efficiently support agentic workflows like coding, web use, and computer interaction.

Why it matters: This technical infrastructure analysis is relevant for understanding how AI agents will scale in real-world collaborative and work contexts, affecting the feasibility of AI-human collaboration scenarios.

Jargon: Operational Intensity (OI) - ratio of compute operations to memory access; Capacity Footprint (CF) - memory requirements; KV cache - key-value memory storage for transformer models; GQA/MLA - grouped query attention/multi-head latent attention architectures.

Finding: This paper proposes "Agentic Environmental Simulations" that use AI to model spatial environments as cognitive agents rather than physics particles, introducing "Cognitive Friction" as a metric to identify ambiguous design elements ("Phantom Affordances") in built spaces.

Why it matters: This directly advances HCI by reimagining how we design and evaluate spatial interfaces and environments, offering new methods to identify usability issues through AI-detected cognitive ambiguities.

Method: Uses Large Multimodal models for episodic spatial reasoning and formalizes AI hallucinations as diagnostic tools for measuring cognitive friction in environmental design.

Jargon: Phantom Affordances = misleading visual cues in environments; Cognitive Friction = measurable confusion/ambiguity users experience; Episodic Spatial Reasoning = event-driven rather than time-based simulation progression.

Finding: The paper introduces the "Presence Without Belonging" framework, arguing that autistic loneliness often occurs during social participation rather than isolation, requiring recognition, access, and sustainability for true belonging.

Why it matters: This reframes inclusion from mere participation to meaningful belonging, offering crucial insights for designing inclusive interfaces, collaborative environments, and educational experiences that support diverse cognitive styles.

Jargon: Epistemic justice - being recognized as a credible knower and having one's experiences validated; Neurodiversity - viewing neurological differences like autism as natural variations rather than deficits.

Finding: This review proposes a temporal cascade framework showing how moral decision-making unfolds across three brain networks - salience network for detecting moral cues, central executive network for conflict monitoring, and default mode network for value-based evaluation - with distinct neural signatures at each stage.

Why it matters: This framework directly informs understanding of decision-making processes and cognitive biases in moral contexts, which is fundamental to behavioral science and has implications for designing systems that support ethical decision-making.

Method: Integrative review synthesizing neuroimaging, EEG, and neurostimulation studies to map temporal dynamics of moral decision processes.

Jargon: ERP = event-related potentials (brain electrical responses); P200/P260/N200/P300/LPP = specific neural components occurring at different millisecond intervals; theta/alpha/beta/delta = different brain wave frequencies.

Finding: Two experimental paradigms for measuring habits (Outcome Devaluation and Response Remapping) produce opposite effects under time pressure, revealing that habits involve multiple distinct cognitive mechanisms rather than a single phenomenon.

Why it matters: Understanding how time pressure affects automatic vs. deliberate decision-making is crucial for designing interfaces, work environments, and interventions that either leverage or counteract habitual behaviors.

Method: Direct experimental comparison of two habit measurement paradigms with systematic manipulation of response time constraints across two experiments.

Jargon: Outcome Devaluation task - measures habits by testing if people continue learned behaviors even when rewards are removed; Response Remapping task - measures habits by testing automatic action selection under new stimulus-response mappings.

Finding: Social fitness platforms like Strava can boost motivation and community but create significant psychological risks for users prone to social comparison and low self-compassion, with women being particularly vulnerable to negative effects.

Why it matters: This research directly informs how social comparison mechanisms in digital platforms affect user motivation and well-being, providing crucial insights for designing healthier social features in fitness apps and understanding broader HCI design implications.

Method: Mixed-methods study combining quantitative survey of 114 Strava users examining social comparison tendencies with qualitative interviews of long-time and new users.

Jargon: Social comparison orientation - tendency to evaluate oneself relative to others; self-compassion - being kind to oneself during difficulties rather than harshly self-critical.

Finding: Social threats (angry expressions) speed up movement decisions but create a speed-accuracy trade-off where people initially move more slowly and centrally when choosing to avoid threats, suggesting they delay commitment until gathering enough evidence for optimal threat avoidance.

Why it matters: This reveals how perceived urgency affects decision-making processes and could inform design of interfaces that need to balance quick responses with accuracy under stress or threat conditions.

Method: Used a "go-before-you-know" paradigm tracking finger-reaching kinematics as participants chose between chairs in a waiting room with neutral vs. threatening (angry/fearful) individuals.

Jargon: Go-before-you-know paradigm - experimental setup where participants must begin moving before they have complete information to make their decision.

Finding: A multi-agent AI simulator successfully trained 72 mental health professionals in Single-Session Interpersonal Psychotherapy, showing significant improvements in knowledge, self-efficacy, and willingness to deliver interventions through a system where trainees practice with virtual patients while receiving AI-powered supervision.

Why it matters: This demonstrates how AI can scale professional skill acquisition in high-stakes domains, addressing the core challenge of learning complex interpersonal skills when human supervisors are scarce.

Method: Mixed-methods pre-post study using a three-agent architecture (Trainee-Patient-Supervisor) that provides real-time and post-session feedback during simulated therapy sessions.

Jargon: Single-Session Interventions (SSIs) - brief, one-time therapeutic interventions designed to be effective in a single meeting; IPT-SSI - Interpersonal Psychotherapy adapted for single sessions.

Finding: This paper presents an automated pipeline for neuropsychological assessment that creates "digital twins" of individuals' executive functions and metacognition, standardizing administration and scoring of cognitive tasks like the Wisconsin Card Sorting Test across the lifespan.

Why it matters: The standardized assessment of executive functions and metacognition is relevant to understanding cognitive processes that underlie decision-making, learning, and behavioral patterns in HCI and educational contexts.

Method: Uses computational modeling to create multi-layer digital representations of individual cognitive profiles, with automated scoring and trial-by-trial analysis of neuropsychological tasks.

Jargon: Digital twins - computational models that represent individual cognitive profiles; Executive functions - higher-order cognitive processes like working memory and cognitive flexibility; Metacognition - awareness and understanding of one's own thought processes.

Finding: The Polar H10 chest strap shows strong validity for measuring both individual heart rate and heart rate synchrony between pairs of people, matching gold-standard ECG systems in accuracy while being much more practical for real-world use.

Why it matters: This enables researchers to study team dynamics, collaboration, and interpersonal connection through physiological synchrony in natural settings rather than being confined to lab environments.

Method: Dyads watched emotionally evocative audiovisual stimuli while wearing both Polar H10 devices and ECG systems, with validation at both individual and synchrony measurement levels.

Jargon: HR synchrony - when two people's heart rates become coordinated or aligned, often indicating emotional connection or shared engagement.

Finding: This tutorial addresses a critical gap in behavioral and cognitive research by teaching researchers how to properly analyze circular/periodic data (like time of day, screen orientations, or angular measurements) that conventional Euclidean statistics handle incorrectly.

Why it matters: Many HCI studies involve circular data (e.g., gaze directions, touch angles, temporal patterns) that researchers often analyze incorrectly, leading to flawed conclusions about user behavior and interface design.

Method: Comprehensive tutorial covering mathematical representations, visualization techniques, circular statistics (central tendency, dispersion), and circular distribution families as alternatives to standard statistical approaches.

Jargon: Circular data - measurements with periodic/wraparound properties where 0° equals 360°; Euclidean space - standard mathematical space where conventional statistics apply.

Finding: This study proposes social climate as a fifth core domain of physical literacy and uses network analysis to show that regulatory motivation, social support from friends, and autonomous self-regulation are key connectors in youth physical activity engagement.

Why it matters: The network approach to understanding interconnected motivational and social factors could inform design of gamified fitness applications and collaborative physical activity interventions.

Method: Cross-sectional study of 771 youth using validated questionnaires and ActiGraph accelerometers, analyzed with psychological network analysis to identify key connectors between domains.

Jargon: Physical literacy = holistic framework integrating motivation, confidence, physical competence, and knowledge for sustained physical activity; MVPA = moderate-to-vigorous physical activity.

Finding: Saccadic eye movements temporarily disrupt the attentional filter that controls what information enters visual working memory, with filtering effectiveness recovering after ~400ms but remaining impaired at 50ms post-saccade.

Why it matters: This reveals a fundamental constraint in human visual-cognitive processing that affects how users can attend to and remember information during natural viewing, directly informing UX design for interfaces requiring visual attention and memory.

Method: Used a continuous report paradigm where participants remembered target colors while ignoring distractors, comparing performance with no saccades versus saccades followed by short (50ms) or long (400ms) delays.

Jargon: Visual working memory (VWM) = temporary storage system for visual information; saccades = rapid eye movements between fixation points; attentional filter = cognitive mechanism that selects relevant information for processing.

Finding: NFL fans on Reddit use more collective pronouns ("we") on game days, with losing fans actually showing higher collective language use than winning fans, despite experiencing more negative emotions after defeats.

Why it matters: This demonstrates how digital platforms shape identity expression and group belonging behaviors, providing insights into how people use language to maintain social connections even during negative experiences.

Method: Analyzed 4 million social interactions from 32 NFL team subreddits using automated psychological text analysis to examine language patterns around game outcomes.

Jargon: Collective pronouns - words like "we," "us," "our" that indicate group membership and shared identity.

Finding: Study used the COM-B behavioral model to identify why adolescents engage or disengage with digital mental health interventions, finding that limited awareness, privacy concerns, and negative reactions to algorithmic targeting create barriers while peer endorsements and trusted contexts facilitate engagement.

Why it matters: Provides actionable insights for designing user experiences and dissemination strategies that account for adolescent behavioral patterns and decision-making processes around digital health tools.

Method: Qualitative study using semi-structured interviews and focus groups with 19 adolescents, analyzed through theory-driven coding based on the COM-B (Capability, Opportunity, Motivation-Behavior) framework.

Jargon: COM-B model - behavioral framework examining how Capability (knowledge/skills), Opportunity (environmental factors), and Motivation (reflective/automatic processes) influence behavior; SSI - single-session interventions (brief, focused therapeutic activities).