📚 Research Digest

Finding: Social proof is the most influential behavioral economics principle in digital marketplace decision-making among young professionals, with online shopping frequency (not demographics) driving significant differences in consumer behavior patterns.

Why it matters: Directly covers behavioral science topics central to the professor's interests: nudge theory, choice architecture, loss aversion, anchoring, default bias, and dark patterns/ethical persuasion in digital contexts.

Method: Sequential explanatory mixed-methods (n=163 survey + 10 qualitative informants; Pearson correlation, ANOVA, Colaizzi's phenomenological analysis).

Jargon: *Loss aversion*: tendency to weight losses more heavily than equivalent gains; *Default bias*: preference for pre-set options; *Choice overload*: decision paralysis from too many options; *Colaizzi's method*: phenomenological approach to extracting themes from lived-experience interviews.

Finding: Casino interior design systematically exploits cognitive biases (via Prospect Theory) and environmental manipulation (no clocks, maze layouts, sensory stimulation) to distort time perception and risk evaluation, weakening rational self-regulation and prolonging engagement.

Why it matters: Directly relevant to behavior change design topics—dark patterns, choice architecture, nudges, addiction by design, and ethical persuasion—offering a rich architectural case study of manipulative design at scale.

Method: Theoretical synthesis with comparative case studies (Las Vegas vs. Singapore) examining how regulatory and cultural contexts moderate behavioral design outcomes.

Jargon: *Prospect Theory* — behavioral economics framework where people evaluate gains/losses asymmetrically, overweighting losses; *rational self-regulation* — conscious cognitive control over impulses and decision-making.

Finding: A large pre-registered replication (n=3,323) of a classic social contagion study found no evidence that witnessing ingroup dishonesty increases individual cheating more than witnessing outgroup dishonesty—strong Bayesian null evidence challenges the original 2009 finding.

Why it matters: This is a direct replication failure of a foundational behavioral science result about social norms and dishonesty contagion, with implications for understanding how group identity shapes ethical decision-making and norm-following behavior relevant to behavioral science and persuasive technology design.

Method: Registered report with conceptual replication using both minimal and natural groups, incentivized coin-flip cheating task, and Bayesian analysis (BF₀₁) to quantify null evidence—a methodologically rigorous approach.

Jargon: *Ingroup contagion* = tendency to copy behavior of in-group members; *Outgroup differentiation* = tendency to behave oppositely from out-group members; *Minimal groups* = arbitrary lab-assigned categories (vs. natural groups like university affiliation); *BF₀₁* = Bayes Factor favoring the null hypothesis over the alternative.

Finding: In group learning contexts, peer support and rating fairness reduce social loafing (free-riding behavior) and thereby improve learning achievement, while self-efficacy and motivation operate indirectly through self-expectation rather than directly inhibiting social loafing.

Why it matters: Directly addresses team dynamics, collective performance, and motivation structures in collaborative learning—relevant to teaching about group work, goal-setting, and behavioral science of effort allocation.

Method: Survey of 600 university students with structural equation modeling to test mediation paths.

Jargon: *Social loafing* = tendency for individuals to exert less effort in group tasks than alone; *Achievement Goal Theory* = framework distinguishing mastery goals (learning-oriented) from performance goals (outcome-oriented).

Finding: AI enhances teaching efficiency through automation and adaptive learning but cannot replicate the affective, ethical, and relational dimensions of teaching; thus, human-AI collaboration rather than replacement is the likely outcome.

Why it matters: Directly relevant to AI's impact on work and human collaboration, and touches on learning sciences and skill acquisition in the context of evolving professional roles for educators.

Method: Non-empirical qualitative literature review synthesizing international scholarship — no original data collected.

Jargon: *Adaptive learning*: AI systems that adjust instructional content to individual learner needs in real time; *Hybrid intelligence*: frameworks combining human and AI capabilities to outperform either alone.

Finding: XAI explanations (SHAP-derived) do not automatically improve decision quality over prediction-only AI output for program managers—the assumption that transparency aids decisions lacks empirical grounding, and this pilot (n=18) provides preliminary effect size estimates for a full study.

Why it matters: Directly relevant to AI impact on human decision-making and work, plus behavioral science (dual process theory, bounded rationality) applied to a practical AI-augmented context—useful for teaching how explainability intersects with cognitive biases and decision quality.

Method: Three-condition controlled experiment (unaided, prediction-only AI, XAI with SHAP explanations) measuring decision accuracy, targeting precision, latency, and communication quality.

Jargon: *SHAP* = SHapley Additive exPlanations, a method for attributing prediction outcomes to individual features; *Dual Process Theory* = System 1 (fast/intuitive) vs. System 2 (slow/deliberate) thinking framework; *XAI* = AI systems designed to make their reasoning interpretable to humans.

Finding: In-service teachers hold more creativity myths but higher self-efficacy for teaching creativity; accurate creativity knowledge (endorsing facts over myths) positively predicts teaching self-efficacy.

Why it matters: Directly relevant to creativity education and skill acquisition—useful for teaching about how misconceptions about creativity affect educators' ability to foster it in learners.

Method: Survey-based regression analysis with 233 Italian teachers comparing pre-service vs. in-service groups on creativity knowledge and self-efficacy scales.

Jargon: *Creativity myths vs. facts* = common misconceptions about creativity (e.g., creativity is innate/fixed) versus empirically supported understandings; *teaching for creativity self-efficacy* = a teacher's confidence in their ability to foster student creativity.

Finding: Proposes thought experiments (TEs) as a formal methodological tool for HCI conceptual work, arguing they help researchers interrogate normative implications of design choices and anticipate technology harms.

Why it matters: Relevant to HCI teaching and research ethics; offers a structured approach to critiquing design frameworks (like Value-Sensitive Design) that connects to behavioral science and ethical persuasion concerns.

Method: Illustrative/philosophical method using two worked examples (Value-Sensitive Design critique and Nissenbaum's contextual integrity) to demonstrate TE structure and application.

Jargon: *Value-Sensitive Design* — a design framework that accounts for human values throughout the technical design process; *Contextual integrity* — Nissenbaum's principle that information flows appropriately when they match the norms of the context in which data was originally shared.

Finding: Non-routine tasks improve employee performance and citizenship behaviors in hybrid workplaces, mediated by "interactive lurking behavior" on digital collaboration platforms, with paradox mindset as a moderator.

Why it matters: Relevant to HCI (hybrid work, digital collaboration platforms) and behavioral science (task design effects on motivation and behavior), with some connection to goal-setting and work design theories.

Method: Three experimental studies across USA and UK comparing routine vs. non-routine task conditions.

Jargon: *Interactive lurking behavior* = passively observing others' interactions on digital platforms without actively participating; *Paradox mindset* = cognitive orientation that embraces rather than avoids contradictory demands; *TMKEPs* = technology-mediated knowledge exchange platforms (e.g., Slack, Teams); *Job Characteristics Theory (JCT)* = framework linking task properties (autonomy, skill variety) to motivation and performance outcomes.

Finding: Teachers implementing a motivation-based academic intervention (Pathways-to-Success) achieved above-threshold fidelity across varied school contexts when given ongoing implementation support, with fidelity linked to better student identity-based motivation and academic outcomes.

Why it matters: Touches on behavior change, motivation, and learning sciences—specifically how fidelity of implementation affects a goal/identity-based motivation intervention—relevant to behavioral science and learning sciences interests.

Method: Video-based fidelity assessment across 92 schools, 10 years, 8,456 students, with AI-based fidelity coding being explored for scalability.

Jargon: *Fidelity of implementation* = degree to which an intervention is delivered as designed; *identity-based motivation* = motivation grounded in students' sense of who they are and who they could become.

Finding: About 26% of social/personality psychology articles make real-world application claims, with 1 in 20 including such claims in the abstract, raising questions about whether the evidentiary bar is high enough given replication challenges in the field.

Why it matters: Directly relevant to behavioral science pedagogy—when teaching replication crises, cognitive biases research, or persuasive technology findings, instructors need to critically evaluate how well lab findings translate to real-world interventions.

Jargon: *Application claims* = explicit statements that findings should or can be applied outside the study context; *Altmetric score* = measure of online attention/social media engagement a paper receives.

Finding: A blended learning framework for retail managers in South Africa improved nine management competencies (including goal-setting and decision-making) by ~11-12% across learner and manager ratings in a pre-post design with 94 participants.

Why it matters: Touches goal-setting, decision-making, and learning sciences, but is narrowly applied to vocational training design in a specific retail/HR context with limited theoretical contribution to the professor's core areas.

Method: Quantitative pre-post survey with learner and line-manager ratings; factor analysis and ANOVA for reliability/validity across nine competency constructs.

Jargon: *NQF* = National Qualifications Framework (South African credential standard); *SETA* = Sector Education and Training Authority (accreditation body).

Finding: Racial microaggressions in clinical settings disrupt self-efficacy and decisional balance for cardiovascular health behavior change, with effects moderated by whether race is conceptualized as biological or social.

Why it matters: Tangentially relevant to behavior change and self-efficacy mechanisms, but the focus is on racial health disparities and clinical communication rather than HCI, design, teams, or the core behavioral science topics the professor teaches.

Finding: Suboptimal devaluation protocols in habit research cause goal-directed behavior to be misclassified as habitual, meaning prior human habit research may be methodologically flawed.

Why it matters: Tangentially relevant to behavioral science and habit formation research—a foundational area for behavior change design—but this paper is primarily a methodological critique of laboratory habit measurement paradigms, not applied to HCI, persuasion, or real-world behavior change.

Method: Pre-registered conceptual replication of a free-operant task with improved outcome devaluation, plus reanalysis of prior datasets.

Jargon: *Outcome devaluation*: making a reward undesirable (e.g., by satiation) to test whether behavior persists habitually or stops due to goal-directedness. *Free-operant task*: a paradigm where participants freely perform actions to earn outcomes, used to study habit formation.

Finding: Computational models in cognitive science can "go rogue" by breaking out of intended constraints, generating artifacts that produce systematically false conclusions about human behavior.

Why it matters: Tangentially relevant to behavioral science methodology — researchers using computational models to study decision-making or cognition need to be aware of these validity threats, but this is a narrow methodological critique aimed at cognitive scientists rather than HCI, design, or applied behavioral science practitioners.

Finding: Peer-assisted distance learning (APDL) outperforms conventional modular distance learning for 10th-grade math performance, with peer collaboration specifically enhancing problem-solving abilities.

Why it matters: Tangentially relevant to collaborative learning and team dynamics in educational settings, but narrowly focused on K-10 math in a Philippine pandemic context with a small sample.

Method: Quasi-experimental crossover design (n=40) with paired/unpaired t-tests; crossover controls for carry-over effects.

Finding: Pre-service teachers rated Canva highly effective for supporting visual and kinesthetic learning styles in English language teaching, based on Likert-scale surveys of 60 teachers.

Why it matters: Tangentially relevant to HCI/UX and learning sciences, but focuses narrowly on a single tool in ELT contexts with weak methodology (self-report perceptions, learning styles framework is largely discredited in cognitive science).

Jargon: *Learning styles (VAK)*: Visual-Auditory-Kinesthetic model, which lacks strong empirical support in learning sciences literature.