Finding: Behavioral biases (overconfidence, loss aversion, herding, mental accounting, anchoring) negatively affect household financial decision-making, with financial literacy partially mediating this relationship.
Why it matters: Tangentially relevant to behavioral science and decision-making biases, but narrowly focused on household finance rather than HCI, behavior change design, or workplace/learning contexts.
Method: PLS-SEM with 386 respondents in a cross-sectional survey design.
⚠️ Self-reported data only | Cross-sectional design (no longitudinal data)
Abhishek Dedhe, Samuel G. B. Johnson, Igor Grossmann • PsyArXiv
Finding: Proposes a two-level cognitive architecture of practical wisdom for ill-defined decisions, where metacognitive processes regulate a repertoire of culturally-transmitted decision strategies under radical uncertainty.
Why it matters: Directly relevant to behavioral science and decision-making (especially under uncertainty), with implications for team cognition, AI collaboration, and how wisdom develops — all touchpoints for an HCI/behavioral science curriculum.
Method: Theoretical/conceptual synthesis drawing on cognitive science to build a formally tractable architecture, with testable predictions derived at the end.
Jargon: *Ill-defined decisions* — problems lacking clear probabilities or commensurable outcomes; *metacognitive processes* — thinking about one's own thinking/strategy selection; *object-level decision strategies* — the actual heuristics or rules applied to a problem.
Ming Pong, Sokha Dara, Niruba Sarath Jayasundara • Semantic Scholar
Finding: ChatGPT usage positively predicts problem-solving ability in undergraduates, and achievement motivation moderates this relationship — higher motivation amplifies the benefit of ChatGPT use (β = 0.157, p < 0.05), explaining 44.6% of variance in problem-solving.
Why it matters: Directly relevant to AI's impact on learning and skill acquisition, and touches on motivation as a moderator of technology-aided learning — useful for teaching about AI tools in education and individual differences in technology adoption.
Method: PLS-SEM with measurement invariance testing (MICOM) across two national samples (Thailand/Cambodia, n=115); small sample, self-report survey, cross-sectional — significant methodological limitations to flag in class.
Jargon: *PLS-SEM* = Partial Least Squares Structural Equation Modeling, a variance-based technique for testing path models; *MICOM* = Measurement Invariance of Composite Models, tests whether constructs mean the same thing across groups before pooling data.
⚠️ Self-reported data only | Convenience sample (e.g., MTurk, students)
Finding: Proposes a conceptual four-layer framework combining cognitive bias identification, strategic planning models (MOST/TOWS), AI-based mitigation via LLMs, and quantitative validation methods like Monte Carlo simulation to reduce bias in strategic decision-making.
Why it matters: Directly relevant to behavioral science, cognitive biases, and AI impact on decision-making — especially for teaching how AI tools might counteract known human biases in organizational and strategic contexts.
Method: Exploratory LLM prompt architecture (system instructions, role definition, context, task) applied to detect authority, optimism, and framing biases in a real policy narrative.
Finding: A desktop VR serious game (BD-VRG) built on constructivist principles improved color composition learning in first-year design students (d = 0.82), with good usability and technology acceptance scores.
Why it matters: Directly relevant to serious games, gamification in education, and learning sciences—especially the intersection of VR, design education, and skill acquisition for abstract perceptual concepts.
Method: 42 first-year art/design students; pre/post-test design with Technology Acceptance Model and System Usability Scale measures alongside qualitative feedback.
Jargon: *ADDIE model* = Analyze, Design, Develop, Implement, Evaluate—a standard instructional design framework; *Desktop VR* = VR experienced on a standard monitor (no headset), lowering cost and access barriers; *TAM* = Technology Acceptance Model, a standard framework measuring perceived usefulness and ease of use.
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-improvement than personality traits, developing through cycles of motivation, instruction, practice, and feedback.
Why it matters: Relevant to learning sciences and skill acquisition, and has implications for behavior change design and coaching frameworks used in HCI/productivity tool design.
Jargon: *Volitional personality change* = intentional efforts to alter one's own personality; *SEB skills* = social, emotional, and behavioral skills treated as learnable capabilities rather than fixed traits; *CPM* = their proposed cyclical model of skill development.
Finding: Pro-environmental outcomes are discounted more steeply as effort increases compared to personal outcomes, suggesting people's cost-benefit analysis weighs effort costs more heavily when the benefit is environmental rather than personal.
Why it matters: Directly relevant to behavior change design and persuasive technology — understanding effort discounting shapes how to design nudges and choice architecture that reduce friction for pro-social behaviors.
Method: Within-subjects online experiment (N=400) using a binary choice task (accept/decline effort-reward combinations), allowing clean separation of effort, reward magnitude, and outcome type effects.
Jargon: *Effort discounting* — the reduction in perceived value of an outcome as the effort required to obtain it increases, analogous to temporal discounting but for effort costs rather than delay.
⭐ Longitudinal design ⚠️ Self-reported data only | No control group mentioned | Limited statistical detail in abstract
Kathleen N. Heeter, J. T. Macy, A. Chow et al. (8 authors) • Semantic Scholar
Finding: In a service-learning health course, performance and avoidant goal orientations (not mastery) predicted academic performance, and self-efficacy did not mediate this relationship — contradicting typical patterns found in traditional academic settings.
Why it matters: The contextual moderating finding (goal orientation effects differ by learning environment) is mildly relevant to goal-setting theory and learning sciences, but the specific domain (health professions service-learning) is narrow.
Method: Mediation analysis with bootstrapped confidence intervals (5,000 resamples), n=74 — small sample limits generalizability.
Jargon: *Mastery orientation* = motivated by learning/competence; *performance orientation* = motivated by outperforming others; *avoidant orientation* = motivated by avoiding failure or looking incompetent.
Finding: The PRISM model proposes that psychedelics don't change cognitive *capacity* but rather *motivation for cognitive effort* — reducing effort costs acutely and increasing salience of meaningful goals post-acutely.
Why it matters: The motivational reallocation framing has peripheral relevance to goal-setting and behavioral science (effort cost perception, goal salience, motivation), but the primary domain is pharmacology/neuroscience, not HCI or behavioral design.
⚠️ No control group mentioned | Single study (not replicated) | Limited statistical detail in abstract
Marie-José van Tol, Ernst H. W. Koster, Matthias Pillny et al. (5 authors) • PsyArXiv
Finding: Proposes a behavioral economics-inspired model of cognitive control allocation to explain persistent cognitive deficits in severe mental illness, framing task engagement failures as maladaptive decision-making.
Why it matters: The behavioral economics / decision-making framing has peripheral overlap with cognitive biases and behavior change, but the application is clinical psychiatry/rehabilitation, not HCI, learning, or workplace contexts.
Tanay Katiyar, George Gillett, Adam Hunt et al. (8 authors) • PsyArXiv
Finding: Peer-to-peer transmission of ADHD information through sequential sharing increases self-diagnosis rates, with effects moderated by format (list vs. narrative) and motivation (likes vs. accuracy).
Why it matters: Tangentially relevant to behavioral science and information diffusion, but focused on psychiatric self-diagnosis rather than HCI, creativity, learning, or behavior change design.
Method: Transmission chain paradigm — participants sequentially modify and pass along information, allowing researchers to track how content and its psychological effects evolve across social transmission.
⚠️ No control group mentioned | Single study (not replicated) | Limited statistical detail in abstract
Finding: Children show stable individual differences in how they trade off external resource sampling versus internal working memory use across tasks, with "heavy-loaders" consistently relying more on memory and "light-lifters" sampling more frequently.
Why it matters: Has peripheral relevance to learning sciences and cognitive load considerations in instructional design, but is narrowly focused on developmental working memory strategy research in children.
Method: Preregistered developmental study using two isomorphic tablet tasks (Shopping Game, Zoo Game) with N=121 children ages 6-10, measuring test-retest stability of memory strategies.
Jargon: *Heavy-loaders/light-lifters*: individuals who prefer internal memory vs. external resource sampling; *CRT-D*: Cognitive Reflection Task (developmental version), measuring deliberative thinking over intuitive responses.
⭐ Pre-registered | ⭐ Longitudinal design ⚠️ No control group mentioned | Single study (not replicated)
Brendan Conway-Smith, Ruby Potash, Matthew Sacchet • PsyArXiv
Finding: Advanced meditation develops through automatization of metacognitive control—explicit instructions become less necessary as practitioners shift from effortful to fluent, automatic metacognitive regulation directed at specific phenomenological targets.
Why it matters: Tangentially relevant to learning sciences and skill acquisition frameworks, but the domain (meditation/phenomenology) is narrow and unlikely to transfer directly to HCI, creativity, teams, or behavior change teaching contexts.
Nirit Binyamini Ben-Meir, Hong Haoran, Miyuki Oka et al. (8 authors) • Semantic Scholar
Finding: Digital mediation of human-nonhuman species encounters creates psychological distances (temporal mismatch, perceived authenticity issues) even as it tries to bridge spatial/temporal gaps, with design strategies proposed for hybrid techno-ecological environments.
Why it matters: Tangentially relevant to HCI and design, but focused on ecological/environmental interaction design rather than the professor's core areas of creativity, teams, learning, AI, or behavioral science.
Ha Le, Akshat Choube, Varun Mishra et al. (4 authors) • Semantic Scholar
Finding: GLOSS4HAR, a multi-agent LLM system, can correct self-reported activity annotations by combining passive sensing data with lightweight self-reports, achieving 75-92% F1 alignment with human annotations.
Why it matters: Tangentially relevant to HCI and human-AI collaboration in research methodology, but focused narrowly on activity recognition/health tracking data pipelines rather than the professor's core teaching areas.
⚠️ Self-reported data only | No control group mentioned | Single study (not replicated) | Limited statistical detail in abstract
Lenneke Kuijer, Joep W. Frens, Piet de Koning et al. (4 authors) • Semantic Scholar
Finding: A workshop methodology for exploring "post-growth" design paradigms in HCI, where participants create a "sandbox" context to develop design propositions for post-growth living.
Why it matters: Marginally relevant as an HCI workshop design method, but the core focus on ecological economics and post-growth ideology is peripheral to the professor's specific interests in creativity, behavioral science, teams, or AI.
⚠️ No control group mentioned | Single study (not replicated) | Limited statistical detail in abstract
Finding: Traditional cultural elements in digital games (costumes, environments, characters, narratives) serve as tools for cultural preservation and cross-cultural communication, enhancing player engagement and emotional resonance.
Why it matters: Tangentially relevant to game design and cultural dimensions of player engagement, but lacks empirical rigor and doesn't connect to HCI, behavioral science, or learning sciences in meaningful ways.
Jargon: Cultural totems = symbolic objects representing group identity; immersive experience = sense of presence/absorption in a game world.
Miran Kim, Junkyu Lee, Hosung Nam • Semantic Scholar
Finding: App development as a learning activity in teacher education increased perceived digital competence and "pedagogical authorship" (teachers as creators rather than consumers of digital tools), with gains more pronounced for undergraduates than graduates.
Why it matters: Marginal relevance to learning sciences and skill acquisition, but narrowly focused on English teacher education rather than HCI, creativity, or behavioral science broadly.
⚠️ Very small sample (n=16) | Self-reported data only | No control group mentioned | Single study (not replicated) | Convenience sample (e.g., MTurk, students)
Finding: Introduces the Prompt-Mediated Curriculum Development (PMCD) framework, positioning "prompt literacy" as the core competency for educators collaborating with AI systems in curriculum design, with a three-level taxonomy (Foundational, Pedagogical, Agentic Orchestration).
Why it matters: Tangentially relevant to AI impact on work and human-AI collaboration, but focused narrowly on curriculum administration rather than HCI, creativity, learning sciences, or behavioral science.
Jargon: *Prompt literacy*: ability to formulate, refine, and evaluate AI prompts for pedagogically relevant outcomes; *Agentic AI*: AI systems that autonomously take multi-step actions toward goals.
N. Mukhtar, Kabiru Adamu, Datti Adam Suleiman et al. (4 authors) • Semantic Scholar
Finding: Jigsaw IV cooperative learning significantly improved calculus retention among polytechnic students compared to conventional lecture methods (mean 42.70 vs. 33.30).
Why it matters: Marginally relevant to team dynamics and collaborative learning, but the context (Nigerian polytechnic calculus instruction) and narrow pedagogical focus make it a weak fit for the professor's core interests.