Finding: Immediate reward (vs. delayed reward) causally reduces procrastination, with strongest effects for high-trait procrastinators, mediated by motivation—without sacrificing task completion rates.
Why it matters: Provides rare experimental causal evidence for a core behavioral science intervention relevant to goal-setting, habit formation, persuasive technology design, and motivation architecture in productivity tools and learning platforms.
Method: Preregistered experiment (n=611) using a novel Multi-day Online Reading Task (MORT) that captures full time-course of self-paced work over 7 days—strong ecological validity for a lab paradigm.
Jargon: *Trait procrastination* = stable individual disposition to delay tasks; *reward-performance relationships* = different structural mappings between task completion and payment timing.
⭐ Pre-registered ⚠️ No control group mentioned | Single study (not replicated) | Limited statistical detail in abstract
Finding: A normative (dynamic programming) theory of procrastination unifies isolated findings, predicting three temporal work patterns and identifying how discounting rate, reward magnitude, task aversiveness, and intervention design (immediate rewards, intermediate deadlines) affect procrastination severity and outcomes.
Why it matters: Directly relevant to goal-setting, motivation, behavior change design (deadlines, incentive structures, nudges), and habit formation — all core to the professor's teaching areas.
Method: Dynamic programming model deriving optimal work schedules over time, with empirical validation of the perfectionism component.
Jargon: *Temporal discounting* = tendency to value immediate rewards over future ones; *dynamic programming* = mathematical optimization across sequential decisions; *effort-cost function* = how the subjective cost of effort scales with amount of work done.
⚠️ No control group mentioned | Single study (not replicated) | Limited statistical detail in abstract
Finding: A Self-Determination Theory-informed gamification framework for preventive wellness apps maps autonomy, competence, and relatedness to specific design features; 60% of participants were overwhelmed by fully open goal-setting, supporting structured guidance as essential for autonomy support.
Why it matters: Directly relevant to gamification design, behavior change, goal-setting scaffolding, and persuasive technology — with an empirical finding that challenges naive "user control = good" assumptions in UX and habit formation design.
Method: Design science research combining systematic literature review, co-design with 20 young adults, and expert validation (n=6 multidisciplinary panel).
Jargon: *SDT (Self-Determination Theory)* — motivational theory positing three core needs: autonomy (sense of control), competence (mastery), and relatedness (social connection); *Design science research* — methodology focused on building and evaluating artifacts as contributions to knowledge.
Shankar Subramanian Iyer, Raman Subramanian, R. Arora et al. (4 authors) • Semantic Scholar
Finding: Digital payment systems exploit cognitive biases (mental accounting, present bias) to reduce the "pain of paying," increasing spending frequency by 34% and impulse purchases by 28%; the study introduces "Spendception" to describe how reduced transaction visibility drives unconscious overspending.
Why it matters: Directly relevant to behavioral science, dark patterns, choice architecture, and the ethics of persuasive technology design — core topics in behavior change design curriculum.
Jargon: *Pain of paying* — the psychological discomfort of spending that acts as a natural spending brake, reduced in cashless transactions. *Mental accounting* — people mentally categorize money differently depending on form/source. *Present bias* — overweighting immediate gratification over future consequences.
Xiaoling Li, Xu Xian, Peitao Du • Semantic Scholar
Finding: Students who use AI reflectively (plan first, evaluate output, revise in their own voice) retain psychological ownership of their work and experience less competency anxiety, while those who outsource to AI feel the work is "polished but distant" and report shallower learning and stronger anxiety about their own abilities.
Why it matters: Directly relevant to how AI tools are reshaping learning, skill acquisition, and academic identity—with implications for designing AI-assisted learning environments and assessing authentic human contribution.
Method: Qualitative: 50 student interviews + 10 faculty interviews + 128 critical incident records from Chinese universities, analyzed via hybrid deductive-inductive thematic analysis in NVivo 14.
Jargon: *Psychological ownership* — the felt sense that "this is mine," even when co-created; *competency anxiety* — fear that one lacks genuine ability, distinct from performance anxiety; *dependent AI outsourcing* vs. *reflective human-AI collaboration* — the study's core distinction between passive delegation and active engagement with AI output.
⚠️ No control group mentioned | Single study (not replicated) | Convenience sample (e.g., MTurk, students) | Limited statistical detail in abstract
Christopher J. Cagna, Jamil Bhanji, Darian Raizberg et al. (5 authors) • PsyArXiv
Finding: A novel "Feedback Value Training" paradigm attempted to increase learners' willingness to seek out performance feedback; while overall feedback-seeking didn't significantly increase, participants showed greater valuation of *informative* feedback, and "need for cognition" moderated responsiveness to the training.
Why it matters: Directly relevant to learning sciences, motivation, and behavior change design—specifically how to make feedback-seeking intrinsically valued rather than avoided, with implications for educational technology and skill acquisition design.
Method: Willingness-to-pay paradigm to quantify feedback value—participants decided whether to "purchase" feedback during an associative memory task, providing a behavioral (not self-report) measure of feedback valuation.
Jargon: *Need for cognition*: individual trait reflecting tendency to enjoy and engage in effortful thinking. *Associative memory task*: learning task requiring pairing of stimuli. *Feedback value training (FVT)*: experimental manipulation of learning context designed to increase perceived worth of feedback.
⚠️ Small sample (n=40) | No control group mentioned | Single study (not replicated)
Finding: Mandatory AI implementation in SMEs simultaneously augments human capital (via task enablement) and erodes it (via job insecurity), with no direct effect on employee experience—effects operate through dual psychological pathways.
Why it matters: Directly addresses AI's paradoxical impact on workers, offering a behavioral science framework (JD-R theory) relevant to teaching how AI reshapes work, motivation, and human collaboration.
Method: Two-wave survey (n=461) with PLS-SEM, fsQCA, and 45 semi-structured interviews across China and Europe—notable for multi-method rigor and cross-cultural comparison.
Jargon: *JD-R theory* = Job Demands–Resources theory, which models how workplace demands deplete and resources restore employee wellbeing; *fsQCA* = fuzzy-set qualitative comparative analysis, identifies combinations of conditions that produce outcomes; *IPMA* = importance-performance map analysis, ranks variables by both their effect size and current performance level; *IC* = intellectual capital (human, structural, and relational capital assets of a firm).
Emily Xing, Yuning Liu, Ravi Iyer et al. (4 authors) • PsyArXiv
Finding: Longitudinal fixed-effects analysis of 1,030 U.S. adults shows platform-specific differences in self-reported social media experiences—Twitter/X and Facebook associate with harm, Snapchat with social connection, YouTube with learning.
Why it matters: Relevant to behavioral science and persuasive technology research on how platform design choices (affordances) differentially shape user outcomes, including harmful engagement patterns.
Method: Within-person fixed effects panel analysis across six waves controls for stable individual differences, isolating platform usage changes from confounds.
Jargon: *Fixed effects analysis*: statistical method that controls for all time-invariant individual characteristics by using each person as their own control across time points.
Finding: Future time perspective reduces training procrastination in athletes, partially mediated by self-control, with mastery-approach goals robustly amplifying this effect while performance-avoidance goals' moderating role disappears when both are modeled simultaneously.
Why it matters: The goal-setting and self-control mechanisms here (future orientation, mastery vs. avoidance goals, procrastination) are directly relevant to behavioral science, motivation research, and goal-setting pedagogy (e.g., OKRs, achievement goal theory).
Method: Structural equation modeling with PROCESS Model 5 moderated mediation, N=1,674 athletes across 48 institutions — solid sample size for this kind of path model.
Jargon: *Future time perspective (FTP)* = how far into the future a person mentally extends their goals and planning. *Mastery-approach goals* = motivation oriented toward developing competence. *Performance-avoidance goals* = motivation oriented toward avoiding appearing incompetent. *PROCESS Model 5* = a moderated mediation model testing conditional indirect effects.
⭐ Longitudinal design ⚠️ Self-reported data only | No control group mentioned | Single study (not replicated) | Cross-sectional design (no longitudinal data)
Nahomi Maki, K. Kanai, Taruto Fuyama • Semantic Scholar
Finding: Asymmetric sensory distribution across players in a cooperative VR game produces qualitatively distinct experiences—visual players report higher engagement/immersion while auditory players show greater reflective insight.
Why it matters: Relevant to team dynamics and collaborative game design, specifically how role asymmetry shapes complementary experiences and collaboration quality.
Method: Two-player VR prototype study comparing roles with intentionally divided sensory modalities (one visual, one auditory).
Jargon: *Perceptual asymmetry* — deliberate mismatch in sensory information available to different players; *cross-modal* — involving multiple sensory modalities.
Finding: Frequent digital media use combined with cognitive biases (heuristics, impulsivity, low risk perception) and high online trust increases susceptibility to cybercrime and fraud, explained through Routine Activity Theory and behavioral decision-making frameworks.
Why it matters: The cognitive biases and behavioral decision-making angle (heuristics, trust heuristics, digital nudges gone wrong) touches behavioral science and persuasive technology ethics, but the cybercrime/fraud focus is tangential to the professor's core interests.
Jargon: *Routine Activity Theory* — criminological framework positing victimization occurs when a motivated offender, suitable target, and absent guardian converge; *digital guardianship* — protective behaviors/tools that reduce online vulnerability.
Piia Näykki, Tiina Törmänen, Jaana Isohätälä et al. (5 authors) • Semantic Scholar
Finding: Emotion regulation strategies in collaborative math tasks don't always match the challenges faced, leading to recurring group difficulties; pre-task evaluations predicted regulation level while post-task evaluations tracked challenge level.
Why it matters: Touches team dynamics and collaborative learning, but narrowly focused on emotion regulation in math education contexts with limited applicability to the professor's core HCI/creativity/behavioral science interests.
Method: 87 hours of videotaped group interaction with mixed-methods analysis (quantitative pre/post evaluations + qualitative case studies).
⚠️ Small sample (n=62) | No control group mentioned | Single study (not replicated) | Limited statistical detail in abstract
Finding: Trait anxiety during quarantine reduces positive affect and increases negative affect, and greater variability in negative affect lowers patch-exit thresholds in a foraging decision task, suggesting emotional instability leads to persisting in depleted options.
Why it matters: The decision-making under uncertainty angle (affect influencing sequential choice behavior) has weak connections to behavioral science and cognitive biases, but the social isolation and anxiety framing is too clinical/psychological to be actionable for HCI, teams, or behavior change design contexts.
Method: 10-day longitudinal experience sampling study (~31,000 momentary assessments) combined with a virtual patch-foraging task to measure decision thresholds.
Jargon: *Patch-foraging task*: a decision paradigm where participants must decide when to leave a diminishing resource for a potentially better one, used to model real-world explore/exploit tradeoffs. *Trait anxiety*: a stable personality disposition toward experiencing anxiety across situations.
⭐ Longitudinal design ⚠️ No control group mentioned | Single study (not replicated) | Limited statistical detail in abstract
Zara Ersozlu, Susan Ledger, Mark Babic et al. (4 authors) • Semantic Scholar
Finding: An AI-supported moderation workflow for teacher performance assessment reduced marking time and cognitive load while improving feedback consistency, with humans retaining final decision authority.
Why it matters: Marginally relevant to AI's impact on work and human-AI collaboration, but narrowly focused on educational credentialing assessment rather than learning sciences, HCI, or behavioral science proper.
Sofia Rasyidah Salsabila, Rumaysha Latifah • Semantic Scholar
Finding: AI can serve as a "co-counselor" in e-counseling by assisting with assessment, emotion detection, documentation, and monitoring, but faces challenges around privacy, algorithmic bias, and cultural/empathy limitations.
Why it matters: Tangentially relevant to AI's impact on human collaboration and work, but focused narrowly on mental health counseling services rather than HCI, creativity, teams, or behavioral science.
Jargon: *Co-counselor*: AI acting as a collaborative assistant alongside a human counselor rather than replacing them; *algorithmic bias*: systematic errors in AI outputs due to skewed training data.
Finding: Bibliometric analysis of 1,487 papers on AI and transversal skills (critical thinking, collaboration, digital competence, ethics) in higher education (2020–2025), finding five major research themes including AI literacy and human-AI collaboration readiness.
Why it matters: Marginally useful as a literature map for AI's impact on work and education, but it's a meta-level mapping study with no empirical findings about actual learning, behavior, or AI use.
Method: Bibliometric analysis using Bibliometrix and VOSviewer on Scopus-indexed publications — a citation/keyword mapping approach, not an experiment or observational study.
Jargon: *Bibliometric analysis*: quantitative study of publication patterns, citations, and co-authorship networks rather than primary data collection.