Adolescent students' stated pre-chat intentions with an AI math tutor tilt strongly toward scaffolded support rather than answer offloading¶

Claim · All claims
Evidence · 1 study · associational r2 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? Before chatting with the AI tutor, students selected on average 4.66 of 8 goals, predominantly learning-oriented options such as step-by-step examples, tips and strategies, and concept explanations, with only 11.8% selecting final solutions. → Rania Abdelghani 2026

Evidence¶

Rania Abdelghani 2026¶

Rania Abdelghani, Peter Kaiser, and Kou Murayama. (2026). Regulating the AI Tutor: Intentions, Help-Seeking, and Self-Regulated Learning in Adolescent GenAI Use. Accepted at the NextGen Learning Interfaces Workshop, AIED 2026. https://arxiv.org/abs/2606.08568 doi:10.48550/arxiv.2606.08568

q2 · i? · associational · r2

Preliminary AI-reported descriptive results from the forethought phase of the classroom study, in which 98 Grade-9 students selected learning goals before chatting. The article reports these as descriptive percentages with no inferential test, quoting that choices were "strongly learning-oriented".

"students selected on average4.66of 8 goals, suggesting broad rather than focused intentions. Choices were strongly learning-oriented: step-by- step examples (82.9%), tips and strategies (80.3%), practice problems (71.1%), concept explanations (69.7%), and checking understanding (69.7%). Only11.8% selected “just give me the final solutions”"

Discussion¶

Related Claims¶

• Students' stated pre-interaction intentions only weakly correspond to their enacted help-seeking with an LLM — reports the opposite
• Students' help requests to the AI tutor were predominantly instrumental rather than executive, with procedural content and low verification — related
• Learning-oriented intentions did not significantly predict enacted in-chat behavior; only selecting final solutions predicted executive and answer-seeking requests — related
• Within a single LLM session, students' help-seeking shifts toward answer- and verification-seeking while explicit self-regulation modestly increases — related
• Limiting verbose Math Agent guidance sharply reduced answer giveaway but decreased cognitive engagement — related