Students' stated pre-interaction intentions only weakly correspond to their enacted help-seeking with an LLM¶

Claim · All claims
Evidence · 1 study · associational r? · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? Only 32.9% of intention-behavior mass fell on matching categories, close to the 32.4% expected from marginal distributions. → Abdelghani 2026

Evidence¶

Abdelghani 2026¶

Abdelghani, R., Kaiser, P., Murayama, K. (2026). Changes in Help-Seeking Strategies Predict unaided Performance during AI-based Mathematical Learning. https://arxiv.org/abs/2606.28472 doi:10.48550/arxiv.2606.28472

q2 · i? · associational · r?

Comparison of students' pre-chat multi-select intent selections with their subsequently coded request types in the LLM chat. Correspondence was near chance: "32.9% of the intention–behavior mass fell on matching categories" versus 32.4% expected from marginals.

"Overall correspondence between intended and enacted categories was weak: 32.9% of the intention–behavior mass fell on matching categories, close to the 32.4% expected from their marginal distributions."

Discussion¶

Related Claims¶

• Grade-9 students' LLM-supported mathematical interactions are strongly request-dominated, with little explicit planning, monitoring, or evaluation — related
• Students' LLM use in CS2 fell into three themes: understanding, assisting code writing, and writing code, with only a few using LLMs to write code — related
• Student-agent interaction produced far more help-seeking but a non-uniform cognitive distribution dominated by lower-order engagement (54.52%), unlike the balanced distribution of student-instructor interaction — related
• Adolescent students' stated pre-chat intentions with an AI math tutor tilt strongly toward scaffolded support rather than answer offloading — reports the opposite
• Learning-oriented intentions did not significantly predict enacted in-chat behavior; only selecting final solutions predicted executive and answer-seeking requests — a narrower finding that bears on this claim
• Behavioral intention translates only weakly into self-reported use behavior, and the extended UTAUT model explains intention formation better than enactment — a narrower finding that bears on this claim
• Students' help requests to the AI tutor were predominantly instrumental rather than executive, with procedural content and low verification — related
• Metacognitive knowledge and intentions to use effective strategies do not reliably translate into actual strategy use (intention-behavior gap) — a narrower finding that bears on this claim
• Adolescent digital self-regulation is enacted through a typology of four strategy types differing in executive effort versus external scaffolding — related