Finding: A Bayesian Item Response Theory framework quantifies human-AI synergy by separating individual vs. collaborative ability and controlling for task difficulty, finding that perspective-taking ability predicts collaborative (but not solo) performance with AI.
Why it matters: Directly relevant to understanding AI's impact on human collaboration and how user social-cognitive factors shape human-AI teaming — useful for teaching AI + teamwork intersections.
Method: Bayesian Item Response Theory applied to human-AI benchmark data (n=667), decomposing additive synergy from non-linear interaction dynamics.
Jargon: *Item Response Theory (IRT)* — psychometric framework modeling the relationship between latent ability and task performance, controlling for item/task difficulty; *additive synergy* — performance gain from collaboration beyond what either party achieves alone.
Finding: Individual differences in decision strategies during AI-assisted decision-making are obscured by aggregate-level analyses, meaning interventions to improve "appropriate reliance" on AI fail because they assume homogeneous behavior across users.
Why it matters: Directly relevant to understanding how humans collaborate with AI systems and why behavioral interventions for human-AI teaming have mixed results — important for teaching AI's impact on work and decision-making.
Method: Three-study design combining simulation modeling, a behavioral face-matching task with varying AI confidence signals, and simulation to demonstrate ecological fallacy in group-level analysis.
Jargon: *Appropriate reliance* = using AI advice when it's correct and ignoring it when it's wrong; *decision strategies* = individual-level rules (e.g., always follow AI, weight by confidence) governing how a person integrates AI advice with their own judgment.
⭐ Includes replication ⚠️ No control group mentioned | Limited statistical detail in abstract
Finding: Employee-led AI collaboration produces higher job satisfaction than GenAI-led collaboration, mediated by AI self-efficacy, with regulatory focus moderating the effect (promotion-focused employees thrive under employee-led; prevention-focused under GenAI-led).
Why it matters: Directly relevant to AI's impact on human work and collaboration dynamics, with implications for designing human-AI workflows and understanding motivational orientations in AI-augmented teams.
Method: Three experiments (two scenario-based, one field experiment across hotels with different star ratings) using a 2×2 between-subjects design.
Jargon: *Regulatory focus theory* — promotion-focused individuals pursue gains/aspirations; prevention-focused individuals avoid losses/obligations. *AI self-efficacy* — an employee's belief in their ability to work effectively with AI tools.
⭐ Includes replication ⚠️ Limited statistical detail in abstract
Finding: Proposes a theoretical upper bound of Cohen's d = 0.50 for experimentally induced behavioral changes, arguing identity-preservation mechanisms naturally constrain effect sizes rather than methodological flaws being solely responsible.
Why it matters: Has marginal relevance for interpreting effect sizes in behavior change and persuasive technology research, but the core argument is meta-theoretical about psychological experimentation rather than about HCI, games, creativity, or any of the primary topic areas.
Jargon: *Cohen's d* = standardized effect size measure; *assimilation/accommodation* = Piagetian terms for integrating new info into existing schemas vs. changing schemas to fit new info.
Finding: This scoping review maps how the Behavior Change Wheel (BCW) framework has been applied to design physical activity interventions for children and adolescents, finding that most applications focus on individual self-regulation (goal-setting, self-monitoring) rather than environmental/opportunity-level change.
Why it matters: The BCW and COM-B model are relevant to behavior change design broadly, but this paper's focus is narrowly on pediatric physical activity interventions, not HCI, games, or the digital/persuasive technology contexts the user cares about.
Jargon: *BCW (Behavior Change Wheel)*: a framework linking behavioral diagnosis to intervention strategies; *COM-B*: Capability, Opportunity, Motivation-Behavior model underlying the BCW; *PRISMA-ScR*: reporting guidelines for scoping reviews.