Static, Dynamic, or Adaptive? Comparing Interface Scaffolding Methods in an Intelligent Tutoring System

Steering AI Tutors Through System Prompts: A Crossover Study on Self-Regulated Learning and Cognitive Engagement Scaffolds in CS1

The results suggest that the pedagogical behavior of AI tutors may not be easily steered through system prompts alone: embedding established SRL and CE frameworks did not produce detectable improvements on any preregistered outcome in a large, ecologically valid deployment.

Ask Better, Learn Better: An Adaptive Metacognitive Support System for Enhancing Student Prompting with LLMs

REAT: A Reflective Experience-Augmented Tutoring Framework for Multi-turn Mathematical Instruction

The proposed REAT framework couples experience distillation from historical dialogues with real-time adaptive retrieval, which significantly outperforms both prompt-only and supervised fine-tuning (SFT) baselines, particularly in improving complex, low-scoring tutoring scenarios.

LLM Pedagogical Behavior in AI Tutoring Interactions

A five-level scale is developed, validated against human annotations, that characterizes responses according to the degree of direct assistance LLMs provide when students use them as tutors in authentic learning interactions and provides an empirical baseline for LLM assistance in tutoring interactions and a measuremen...

Development and Implementation of an AI Tutor To Scaffold Problem-Solving in Introductory Chemistry

Developing problem-solving competency is a central goal of undergraduate chemistry education, yet many students struggle to engage productively with complex, multistep problems. At the same time, the rapid adoption of large language model (LLM)–based tools by students in STEM courses raises concerns about over-relian...

Adaptive Scaffolding Needs Contingency: An AI Tutor That Escalates and Fades on What the Learner Does

CoMeT matched the question-only tutor's demand, delivered artifacts twice as often as the assistant, and frustrated learners less than the question-only tutor, with delegation and load unchanged.