AI in Mathematics and Statistics Education: Student and Lecturer Perspectives

Key Findings

• 59% of surveyed students have used AI for help with assignments.
• 56% of students believe that using AI for assignments constitutes cheating.
• 72% think that AI often provides incorrect answers to math and statistics questions.
• 60% are concerned that using AI for assignments could devalue their degree.
• 33% of students report that their use of AI tools for assignments has increased over the last academic year.
• Lecturers express concerns about maintaining academic integrity while recognising the potential benefits of AI.

Study Overview

This study aims to understand the impact of AI, particularly Generative AI tools like ChatGPT, on student learning and assessment practices in mathematics and statistics education. We used a mixed-methods approach to gather insights from both students and lecturers.

• Survey: An online survey conducted in June 2024 of 145 mathematics and statistics students to assess their use of AI tools, ethical concerns, attitudes towards AI-proofing strategies, and perceptions of AI's impact on the value of their degrees. The sample included AI users (those who have used AI tools like ChatGPT for assignments) and non-AI users to enable comparative analysis.
• Focus Groups: Two focus groups with 6 students each, conducted in June 2024, divided into AI and non-AI users, were undertaken to capture diverse perspectives and ensure a balanced discussion without perceived conflict.
• Lecturer Insights: Comments and feedback were collected through survey responses and discussions with lecturers to understand their views on AI's role in education and academic integrity.

Student Perspectives

Students reported diverse experiences with AI:

• "AI is hopeless at answering any assignment questions. It's useful for coding if you get an error message, you can just copy and paste it into ChatGPT and it'll give you an idea of how to correct it."
• "I've used AI to analyse lecture notes, so we got lecture notes for one module that's 200 pages long, then it's quite useful at filtering out the information I need, or it's also good for revision plans."
• "ChatGPT is hilariously bad at maths. It's very rarely provided anything more useful than just, you know, guessing and checking."

Lecturer Insights on AI in Assignments

Lecturers highlighted several key points based on their survey comments and discussions:

• "I think most Warwick students are probably more horrified than we are at the idea of cheating. I have the impression they are quite reluctant to use AI, even legitimately."
• "In thinking about AI I believe it is important to distinguish where we are now, in 2024, with where we will be in a few years. Today, ChatGPT, and I imagine other free AI tools, seem to be terrible at handling mathematics assessments, and I would advise a student to avoid using it."
• "We need to make it clearer to students that solving problems on assignments is their best learning opportunity, and that getting help (from AI or elsewhere) deprives them of the opportunity to solve the problem on their own."
• "I think it's super useful, but it can't replace the skill of problem-solving in every context, so I think we need to discuss this properly and openly with students."

Detailed Analyses

Survey Results Analysis

Examine the full quantitative data from our student survey on AI use.

Focus Group Insights

Explore in-depth qualitative data from student discussions on AI in academic work.

Implications, Limitations, and Conclusion

Our research shows that while AI use is common among students, their attitudes toward its role in education are complex and often conflicted. Despite widespread and growing use, concerns about academic integrity persist. Lecturers acknowledge both the limitations and potential of AI tools in solving complex problems.

These findings suggest a need for:

• Clear guidelines and education on AI's appropriate use in academic settings
• Reassessment of current assignment structures, particularly take-home assignments
• Further research on AI's long-term impact on mathematical education

While our findings align with broader discussions on AI use across departments, they specifically reflect mathematics and statistics contexts and may not represent all educational disciplines. To address the challenges uncovered, ongoing dialogue among students, educators, and administrators is essential to strike a balance between AI integration and maintaining rigorous academic standards in STEM education.