Equity in the Use of ChatGPT for the Classroom: A Comparison of the Accuracy and Precision of ChatGPT 3.5 vs. ChatGPT4 with Respect to Statistics and Data Science Exams
Project Overview
The document examines the role of generative AI, particularly ChatGPT3.5 and ChatGPT4, in education, with a focus on statistics and data science. It reveals performance disparities between users of the free and paid versions of these AI tools, suggesting that students with access to the paid version may achieve better learning outcomes, thereby creating potential inequities in educational opportunities. While generative AI holds promise for democratizing education by providing innovative learning aids, the research warns that unequal access to these resources may exacerbate existing knowledge gaps among students. Consequently, the document underscores the critical need to prioritize equity in the deployment of AI technologies within educational settings to ensure all learners benefit from advancements in AI.
Key Applications
Comparison of ChatGPT3.5 and ChatGPT4 in answering statistics exam questions
Context: Educational context of higher education statistics and data science courses; target audience includes undergraduate and graduate students.
Implementation: Exam questions were entered into ChatGPT3.5 and ChatGPT4, and the responses were scored and compared based on accuracy.
Outcomes: ChatGPT4 outperformed ChatGPT3.5 significantly, especially on questions involving images where ChatGPT3.5 could not provide accurate answers.
Challenges: The main challenge is the equity issue regarding access to the paid version of ChatGPT4, which may disadvantage students who cannot afford it.
Implementation Barriers
Financial Barrier
The cost of access to ChatGPT4 ($20/month) may prevent lower-income students from using the more advanced and capable AI tool. Potential solutions include providing subsidized access to AI tools for low-income students or offering free versions with similar capabilities.
Accessibility Barrier
Some students may face challenges in using AI tools due to visual impairments or lack of familiarity with technology. Implementing text-to-speech capabilities and ensuring compatibility with screen readers can enhance accessibility.
Project Team
Monnie McGee
Researcher
Bivin Sadler
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Monnie McGee, Bivin Sadler
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