📚 Research Digest

Finding: Pre-service teachers producing higher-quality instructional designs with GenAI showed a distinct evaluation-revision feedback cycle ("planning → decision-making → generation → evaluation → revision → re-evaluation"), while low-quality producers were stuck in fragmented, ineffective behavioral loops—with large effect sizes for the evaluation-revision cycle (r = 0.606) and mental demand (r = 0.647).

Why it matters: Directly informs how to teach AI-assisted design workflows, human-AI collaboration patterns, and cognitive load management—all highly relevant to HCI, learning sciences, and AI's impact on creative/instructional work.

Method: Lag sequential analysis compared behavioral transition patterns between high- and low-quality groups across two tasks (within-sample replication), combined with multidimensional cognitive load measurement.

Jargon: *Lag sequential analysis* — a method for identifying statistically significant transitions between behavioral states in sequential data; *instructional design* — the systematic planning of teaching/learning activities; *cognitive load* — mental effort required by a task, often decomposed into intrinsic, extraneous, and germane dimensions.

Finding: Head-mounted AR with gamification and virtual pet features significantly enhanced enjoyment, affective valence, attentional dissociation, and remembered pleasure during brief vigorous physical activity bouts, while motivational prompts alone had no effect.

Why it matters: Directly relevant to gamification design, persuasive technology for behavior change, and HCI—showing which AR-based engagement mechanisms actually shift exercise experience and memory (peak-end effects).

Method: Counterbalanced within-subjects design (n=24) comparing three AR conditions vs. control using linear mixed models; includes both in-the-moment and remembered affect measures.

Jargon: *Affective valence*: subjective feeling of pleasure/displeasure during activity. *Attentional dissociation*: distraction from discomfort by redirecting attention outward. *Remembered pleasure*: post-activity affective recall, which drives future behavior more than in-the-moment experience. *VILPA*: brief vigorous physical activity embedded in daily life rather than structured exercise.

Finding: This chapter synthesizes research on teacher achievement goals, examining how teachers' personal achievement goals shape their instructional behavior, professional learning, well-being, and the classroom goal structures that influence student motivation.

Why it matters: Directly relevant to goal-setting theory (achievement goals, mastery vs. performance orientations) and learning sciences, and provides a dual-role framework useful for teaching about motivation design in educational and organizational contexts.

Jargon: *Achievement goal theory* — framework distinguishing goals oriented toward mastery/learning vs. performance/demonstrating competence; *classroom goal structures* — the motivational climate teachers create for students; *goal-complex* — the combination of a goal with the underlying reasons/motives for pursuing it.

Finding: Motivational incentives (time pressure vs. performance-contingent monetary rewards) have differential effects on attentional strategy optimization, with monetary rewards producing modest improvements but time pressure failing to shift strategy use.

Why it matters: Relevant to behavioral science and persuasive technology design—specifically how incentive structures can (or fail to) shift cognitive strategy adoption, which connects to goal-setting and behavior change design.

Method: Adaptive Choice Visual Search (ACVS) paradigm, allowing measurement of spontaneous vs. optimal attentional strategy selection across two experiments with different motivational manipulations.

Jargon: *Attentional control strategies* — deliberate selection of which visual features to prioritize when searching for a target; *performance-contingent reward* — payment tied directly to speed/accuracy of task performance rather than participation.

Finding: Seeding (providing CO₂ anchor values) improved carbon numeracy but did not increase meat tax acceptance or shift food choices; perceived personal price impact was the strongest predictor of policy acceptance.

Why it matters: Relevant to behavioral science and nudge/choice architecture research — specifically the limits of information-based interventions on policy acceptance and behavior change, which connects to persuasive technology and nudge design.

Method: Two pre-registered experiments (N≈1,000 each) with behavioral food choice measure (lottery-based food box selection).

Jargon: *Seeding* — providing reference values to anchor estimates; *carbon numeracy* — ability to estimate carbon footprints of foods without explicit calculation.

Finding: Schumpeter's creative destruction framework remains valid in the AI era but requires reformulation: AI automates "formative creativity" while "transformative creativity" remains dependent on human judgment, and entrepreneurial agency becomes distributed across human-algorithmic configurations.

Why it matters: The formative vs. transformative creativity distinction and the concept of distributed human-algorithmic agency are relevant to understanding AI's impact on creative work and collaboration.

Jargon: *Creative destruction* = Schumpeter's term for innovation displacing existing economic structures; *Intellectual monopoly* = market dominance through control of knowledge/IP rather than physical capital; *Formative vs. transformative creativity* = routine/combinatorial innovation vs. genuinely novel discontinuous innovation.

Finding: AI-generated facial expressions are more stereotypically intense than human-posed expressions ("Hyper-Expressions"), with nearly identical perceptual structures between AI and human faces.

Why it matters: Marginally relevant to AI's societal impact and potential HCI implications around AI-generated imagery and emotional design, but the core focus is emotion perception psychology, not HCI, behavior change, or learning.