Sense and Sensibility: What makes a social robot convincing to high-school students?
Project Overview
The document examines the role of generative AI in education, particularly focusing on the use of social educational robots and their impact on high-school students' decision-making in subjects like electric circuits. The study demonstrates that these robots can effectively persuade students to conform to their answers on true-false questions, thereby significantly influencing student performance. Key findings indicate that the perceived certainty of the robot plays a crucial role in shaping students' responses, and that students' prior familiarity with AI technologies affects their vulnerability to misinformation. This highlights the imperative for educational robots to calibrate their expressions of certainty in order to foster critical thinking skills and reduce the potential for overtrust in AI systems. Overall, the document underscores the transformative potential of generative AI in educational settings, while also cautioning against the risks of misinformation and the need for thoughtful implementation to enhance learning outcomes.
Key Applications
Social educational robot used for interactive learning
Context: High school students in a practical electrical engineering program
Implementation: Students interacted one-on-one with a robot discussing true-false questions about electric circuits.
Outcomes: 75% of students performed beyond expected capacity; high alignment with robot's answers, especially when it displayed certainty.
Challenges: Risk of students being misled by incorrect information from the robot; critical thinking may be hindered by overtrust.
Implementation Barriers
Trust and Misinformation
Students may overly trust the robot, leading to acceptance of incorrect information.
Proposed Solutions: Educators should teach students to critically assess AI-generated information and adjust robot certainty displays based on reliability.
Project Team
Pablo Gonzalez-Oliveras
Researcher
Olov Engwall
Researcher
Ali Reza Majlesi
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Pablo Gonzalez-Oliveras, Olov Engwall, Ali Reza Majlesi
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