Date
Publisher
arXiv
The rapid integration of Large Language Models (LLMs) into educational technology threatens to reduce mathematical learning to mere answer generation. This paper presents a generative study, usability study, and 12-participant field deployment of AITutor, an interactive system that translates theoretical pedagogical mechanisms into concrete user interface features. We explore how junior-high students preparing for high-stakes exams (Zhongkao) interact with AI tutoring. Through mixed-methods triangulation (7,379 telemetry events, 8 contextual observations, 10 interviews), we reveal that students actively resist traditional Socratic dialogue under time pressure, repurposing "answer-first" shortcuts as vital diagnostic checkpoints. We demonstrate how features like layered worked examples, step-linked visual grounding, and metacognitive scaffolding lower the interaction cost of reasoning repair. We contribute a "Reasoning-Centered Product Loop," offering actionable implications for designing AI that structurally supports the inspection, local repair, curriculum verification, and delayed retrieval of mathematical reasoning in the wild.
Study at a glance
- Question: How can AI tutor interfaces balance students' demand for speed with the cognitive requirements of mathematical reasoning?
- Setting and data: 12 junior-high students (grades 7–9) in China preparing for Zhongkao exams, May 15–26, 2026; 7,379 telemetry events, 8 observations, 10 interviews.
- Method: Mixed-methods field deployment with backend telemetry, remote screen-share observations, and semi-structured interviews.
Why it matters
AI tutors that instantly deliver polished solutions risk bypassing essential learning mechanisms. Students facing exam pressure need interfaces balancing fast answer verification with the slower cognitive requirements of inspection and reasoning repair.
Key findings
- Students actively resisted unskippable Socratic dialogue, using final answers as diagnostic checkpoints to decide whether and where to inspect step-by-step reasoning.
- Synchronizing text explanations with dynamic geometry diagrams reduced cognitive split-attention, whereas missing visual labels caused reasoning to break down.
- Students quickly rejected mathematically correct AI solutions when methods exceeded the junior-high syllabus scope, showing trust depended heavily on curriculum fit.
Summary drafted with AI from the full paper and may contain errors. See the full paper for details.
What is the application?
Who is the user?
Who age?
Why use AI?
Study design
