Improving mathematical questioning in teacher training
Project Overview
The document explores the integration of generative AI in education, specifically focusing on a text-based interactive conversational agent aimed at enhancing teachers' mathematical questioning skills through AI-driven simulated classroom environments. By employing human-centered design, deep learning, and uncertainty quantification, this system seeks to provide effective teacher training while recognizing inherent limitations such as data scarcity and challenges related to user expectations. The findings underscore the significance of human-AI collaboration in educational settings, highlighting that while AI can offer innovative tools for educators, successful implementation requires careful consideration of these challenges. Overall, the document illustrates the potential of generative AI to transform educational practices by fostering improved pedagogical skills and adapting to teachers' needs in a supportive manner.
Key Applications
Text-based interactive conversational agent for practicing mathematical questioning
Context: Teacher training for effective mathematical questioning strategies
Implementation: Developed a dialogue system that allows teachers to practice questioning techniques with a simulated student using AI
Outcomes: High user satisfaction and improved conversation success rate; teachers felt they were able to practice and enhance their questioning skills effectively
Challenges: Limitations of conversational agents in modeling complex dialogues, data scarcity, and user expectation misalignment
Implementation Barriers
Model and Data Limitations
Deep learning models used in conversational agents can fail due to reliance on spurious relationships in datasets, leading to challenges in educational contexts where data is limited.
Proposed Solutions: Incorporate human-AI collaboration to mitigate data scarcity and model fragility.
User Expectation Limitations
Difficulty in setting accurate user expectations for task-specific conversational agents which can lead to a mismatch in perceived capabilities.
Proposed Solutions: Improve user feedback mechanisms and clarify the limitations and capabilities of the conversational agents.
Project Team
Debajyoti Datta
Researcher
Maria Phillips
Researcher
James P Bywater
Researcher
Jennifer Chiu
Researcher
Ginger S. Watson
Researcher
Laura E. Barnes
Researcher
Donald E Brown
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Debajyoti Datta, Maria Phillips, James P Bywater, Jennifer Chiu, Ginger S. Watson, Laura E. Barnes, Donald E Brown
Source Publication: View Original PaperLink opens in a new window
Project Contact: Dr. Jianhua Yang
LLM Model Version: gpt-4o-mini-2024-07-18
Analysis Provider: Openai