Metacognitive self-regulation mediates the relationships between all three AI interaction mechanisms and both higher-order outcomes¶

Claim · All claims
Evidence · 1 study · associational r2 · q2
Currency · newest study 2026 · checked against its sources 2026-10-10 (automated, 1 of 1 entry) · not reviewed by a person

Subclaims¶

q2 i? Bias-corrected bootstrapped indirect effects through metacognitive self-regulation were significant for scaffolding (β = .012 and .013), verification literacy (β = .016 and .018), and offloading (β = −.031 and −.035) on both outcomes. → Davor I 2026

Evidence¶

Davor I 2026¶

Davor I, Larbi E and Boateng FO (2026). Artificial intelligence-supported learning and higher-order cognitive outcomes: the mediating role of metacognitive self-regulation. Front. Educ. 11:1855869. https://doi.org/10.3389/feduc.2026.1855869

q2 · i? · associational · r2

Mediation testing used bias-corrected bootstrapping with 5,000 resamples and 95% confidence intervals in the 533-student SEM; all six indirect effects were significant and "All confidence intervals excluded zero."

"metacognitive self-regulation significantly mediated the negative effects of cognitive offloading tendency on critical thinking (β = -.031, p = .016) and technical problem- solving (β = -.035, p = .006), supporting H6 and H9."

Discussion¶

Learner Variables¶

• Self-Regulation — predictor: learners who differ on it differ in outcomes
• Metacognition — predictor: learners who differ on it differ in outcomes

Related Claims¶

• Evidence suggests over-reliance on generative AI may negatively affect students' self-regulated learning behaviors, potentially harming retention and transfer to novel contexts — the general claim this finding bears on, with the studies behind it
• AI verification literacy does not directly predict critical thinking or technical problem-solving — related
• Cognitive offloading tendency is negatively associated with critical thinking, technical problem-solving, and metacognitive self-regulation — related
• Cognitive load partially mediates the relationship between AI-based educational technology use and critical thinking — related
• Self-efficacy and learning motivation serially mediate the link from AIGC affordance to self-regulated learning — related
• Perceived trust mediates the relationship between social influence and generative AI dependence — related
• AI dependency mediates the relationship between teacher support and learning burnout — related
• AI dependency mediates the relationship between academic self-efficacy and learning burnout — related
• AI literacy relates to AI integration intention through a serial indirect pathway via AI-TPACK and science teaching self-efficacy — related
• Perceived employability partially mediates the relationship between AI employment threat perception and career decision-making anxiety — related
• The structural model explains about half the variance in critical thinking and technical problem-solving and about a third in metacognitive self-regulation — related
• Higher reliance on AI tools is correlated with lower critical-thinking performance, mediated by cognitive offloading — related
• Dependent versus autonomous cognitive offloading to GenAI shows opposite motivational and cognitive outcome patterns — related
• Cognitive offloading to generative AI is associated with reduced learner engagement, lower critical thinking, and lower-quality reasoning — related
• Learning agency mediates the association between GenAI literacy and GenAI-assisted self-regulated learning behaviors — related
• Polychronicity moderates the relationships between effective AI use, AI over-reliance, and learning outcomes — related
• Self-regulated learning and critical AI evaluation promote effective AI use — related
• Among AI-exposed students, open collaboration predicted higher aggregate offloading, which was associated with lower Week 8 higher-order thinking (indirect estimate −0.34) — related
• How students and teachers want to interact with classroom AI largely depends on the teacher-student relationships that exist outside of AI use — related
• Strategic cognitive offloading can support transformative learning only when the offloading decision is intentional — related
• Effective intervention timing in AI-based instruction does not always mean immediate response; the interval between student difficulty and teacher intervention can allow independent problem solving — related
• GenAI's effects on higher-order thinking are mixed: support for reflection and idea expansion alongside cognitive offloading and no overall innovativeness gain — related
• Recurring mechanism clusters — self-regulation, metacognition, critical thinking, self-efficacy, trust, feedback literacy, AI literacy — link GenAI use to agency outcomes — a broader claim this one bears on
• GenAI can support learner agency through self-regulated learning, self-efficacy, and feedback literacy under appropriate conditions — related
• LLMs can deliver scalable, personalized, interactive feedback almost instantly, reducing dependency on teacher availability and supporting self-regulated learning in large courses — related
• Perceived usefulness of the AI assistant significantly and partially mediated the relationship between AI-assisted programming and collaborative outcomes and performance — related
• Metacognitive and strategic scaffold-use subgroups solved more problems correctly than moderate and low-use subgroups — related
• Learners perceive AI applications as useful for regulating their learning but not for regulating motivation — related
• GenAI literacy is positively associated with GenAI-assisted self-regulated learning behaviors among special education undergraduates — related
• Self-regulatory activities mediate the relationship between person, context, and achievement — a broader claim this one bears on
• GenAI learning outcomes depend more on use quality (capability, reflective engagement, retained agency) than on frequency of use — related
• Academic self-efficacy does not significantly predict academic stress, and social influence does not directly predict AI dependence — related
• Metacognition mediates the relationship between EF and SRL — related