How motivation and roles influence metacognitive engagement in student-GenAI interaction

Synthesis: This qualitative study analyzes 120 authentic student-GenAI interaction logs and retrospective interviews from 24 Chinese undergraduates to examine how motivational orientations and perceived GenAI roles shape metacognitive engagement. Using epistemic network and lag sequential analysis, the authors find that intrinsically motivated students predominantly positioned GenAI as an instructor or collaborator, while extrinsically motivated students treated it as a replacement tool. Intrinsic motivation and collaborator framing were associated with higher-order metacognitive engagement (evaluation, elaboration) and a complete metacognitive chain, whereas replacement-tool use produced no higher-order metacognitive connections. The authors conclude that the metacognitive consequences of GenAI use are conditioned by motivation and role perception rather than inherent to the technology.

Core Finding

The quality and structure of metacognitive engagement during GenAI interaction is systematically conditioned by students' motivational orientation and the role they assign to the tool. Metacognitive effects of GenAI are not uniform or inherent — they depend on whether students are intrinsically or extrinsically motivated and whether they frame GenAI as an instructor, enhancement tool, replacement tool, or collaborator.

Motivation and GenAI Role (RQ1)

Drawing on self-determination theory, the study adopts a binary intrinsic–extrinsic contrast. Intrinsically motivated students predominantly positioned GenAI as an instructor or collaborator; extrinsically motivated students overwhelmingly treated it as a replacement tool. This difference is not merely attitudinal but unfolds turn-by-turn in how students structure authentic interactions. Extrinsically motivated students were more likely to use GenAI for task completion and cognitive offloading.

Metacognitive Engagement by Motivation (RQ2)

Intrinsic motivation was associated with significantly higher engagement in higher-order metacognitive dimensions — evaluation and elaboration — forming reflective-iterative co-occurrence networks. Extrinsic motivation produced task-completion-oriented patterns concentrated in lower-order processing (orientation, planning, monitoring), where students submitted prompts and monitored outputs for usability but rarely interrogated, critiqued, or extended them. This suggests the cognitive offloading risk is more pronounced among extrinsically motivated students.

GenAI Roles Map to Metacognitive Profiles (RQ3)

Collaborator: the only role with a complete sequential chain spanning all six metacognitive categories (Orientation→Elaboration) plus an Elaboration self-loop, reflecting iterative knowledge construction. Present in only 7.5% of logs.
Enhancement Tool: a partial higher-order chain (M4→M5→M6); the student retains cognitive ownership while using GenAI as a supplement.
Instructor: a receptive comprehension-checking pattern (Orientation-Monitoring, Execution-Monitoring) with near-absent evaluation and elaboration.
Replacement Tool: only self-loops with no higher-order connections and entirely absent Evaluation/Elaboration — cognitive closure.

The augmentation–automation distinction is supported: retaining Agency over one's thinking yields substantially richer metacognitive engagement than full delegation.

Relevance to the wiki

This study contributes process-level empirical evidence to the wiki's treatment of Student AI Interaction, Metacognition, and Human AI Collaboration. It moves beyond the broad "cognitive offloading" framing (see Cognitive Offloading and related articles like Shaw Nave Cognitive Surrender 2026) to show that metacognitive outcomes depend on motivational stance and role framing. It supports Self Regulated Learning guidance that GenAI tasks be reframed as exploratory and dialogic, and that interfaces embed evaluative prompts to sustain Agency. It connects to Motivation, reducing misuse, and practical instructional design for higher education.

Connected Concepts

Connected Articles

Citation

Cui, Y., Yan, W., Chiu, T. K. F., & Nakajima, T. (2026). How motivation and roles influence metacognitive engagement in student-GenAI interaction. International Journal of Educational Technology in Higher Education.

View Source View Changes Edit