Befriending ChatGPT and other superchatbots: An AI-integrated take-home assessment preserving integrity
Project Overview
The document examines the role of generative AI, particularly ChatGPT, in transforming educational assessments, with a focus on take-home exams. It addresses the challenges of academic integrity arising from the potential misuse of AI tools, while simultaneously proposing an innovative assessment method that leverages ChatGPT as a resource rather than a shortcut. This approach encourages students to engage critically with AI-generated responses, fostering skills in critical thinking and problem-solving. By guiding students to use ChatGPT's insights in a structured fashion, the method aims to maintain the integrity of assessments and enhance overall learning outcomes, demonstrating how generative AI can be effectively integrated into educational practices to support deeper learning while mitigating risks associated with its misuse.
Key Applications
AI-integrated take-home assessment technique using ChatGPT
Context: Undergraduate students in a Metallurgical and Materials Engineering program at National Institute of Technology Tiruchirappalli, India.
Implementation: The assessment tasked groups of students to use ChatGPT as a primary source for answering questions, requiring them to include prompts and responses, and to evaluate the AI's answers critically.
Outcomes: The assessment maintained integrity and resulted in diverse responses from students, indicating effective engagement with ChatGPT.
Challenges: Challenges include ensuring students do not overly rely on AI, and the potential for factual inaccuracies in AI responses.
Implementation Barriers
Integrity Threat
Concerns over the misuse of ChatGPT leading to academic dishonesty in assessments.
Proposed Solutions: Integrate ChatGPT in a structured way that requires students to critically assess its responses and report their process.
Technical Limitations
ChatGPT may produce inaccurate or unreliable information.
Proposed Solutions: Develop AI-based language analysis tools and employ existing plagiarism detection tools to identify AI-generated content.
Project Team
P G Kubendran Amos
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: P G Kubendran Amos
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