Learning, Teaching and Assessment with AI Series: Ethical Use of AI [Exercise 04] – Refining Student Study Notes Using AI Feedback

Learning, Teaching and Assessment with AI Series: Ethical Use of AI [Exercise 04] – Refining Student Study Notes Using AI Feedback

Tuesday, November 3, 2026

Image from Magnific.com

The Learning, Teaching and Assessment with AI series continues. Here follows Exercise 04 by Professor Derek Yu (Department of Economics) on how different approaches to using AI tools such as ChatGPT can meaningfully affect student learning outcomes in Economics. This example demonstrates how AI can be used not only to generate content, but also to critically review and refine student-generated study notes, thereby strengthening conceptual understanding in Microeconomics.

Context: Using AI to Refine Understanding of Perfect Competition

Returning to the two first-year Microeconomics students from Exercise 03, Isabella and Jabulani, both have now acquired the prescribed textbook after initially relying on AI tools due to limited access to learning resources. They have been introduced to one of the most conceptually challenging topics in the course: Perfect Competition. Faced with the complexity of the chapter, Jabulani avoids engaging deeply with the textbook and instead asks ChatGPT to summarise the key characteristics of a perfectly competitive market. In contrast, Isabella first engages with the prescribed textbook and develops her own set of study notes before using AI to obtain feedback and improve her understanding. This contrast illustrates two different learning approaches: AI substitution versus AI-assisted refinement. 

Structured Prompt and Methodologically Sound AI Output

Jabulani asked ChatGPT: “Can you please summarize the 5 key characteristics of perfect competition market structure?”

ChatGPT responded with a basic list:

• Many buyers and sellers
• Homogeneous products
• Free entry and exit
• Perfect information
• Long-run economic profit: zero (normal profit only)

While factually correct, the response remains basic and underdeveloped, lacking explanatory depth and economic precision.

Human-Led Refinement and External Validation

Isabella first developed her own study notes based on the prescribed textbook:

• Many buyers and many sellers
• Homogeneous products (standardised goods such as milk)
• Completely free market entry and exit
• Perfect information about prices, products, and production methods
• In the short run, firms may earn economic profit, normal profit, or incur losses, but in the long run only normal profit is possible (zero economic profit)

Isabella then asked ChatGPT to review her notes for improvement.

ChatGPT provided targeted feedback:

• Add that no single buyer or seller can influence market price (price taker condition)
• Clarify “perfect information” as including prices, products, and production methods
• Minor refinements for precision and completeness

Isabella then improved her notes accordingly by incorporating these refinements, strengthening both clarity and academic alignment.

Final Outcome: Strengthening Learning Through AI-Assisted Refinement

The final learning outcome demonstrates that:

• AI-generated summaries alone may be correct but incomplete
• Student-generated notes based on the textbook are more accurate and conceptually grounded
• AI is most effective when used as a feedback and refinement tool rather than a replacement for learning
• Learning is strengthened when students actively revise their own understanding using both textbook knowledge and AI feedback

This process results in improved conceptual clarity of perfect competition and stronger academic ownership of learning material.

Reflection: Critical Engagement with AI in Conceptual Learning

This example highlights several key principles:

• AI can generate basic summaries, but may omit important economic refinements
• Engaging with the prescribed textbook first leads to deeper conceptual understanding
• AI is most useful when used for feedback, correction, and refinement rather than substitution
• Student-generated notes promote active learning and knowledge construction
• Critical thinking is strengthened when students compare, revise, and validate information

As emphasised in the learning process:

• Always begin with disciplinary sources (textbook engagement)
• Use AI to refine and improve understanding, not replace it
• Actively revise notes based on feedback and verification

Implications for Teaching and Research Practice

This example demonstrates how AI can be effectively integrated into Economics teaching:

• AI can function as a formative feedback tool for student-generated work
• Students develop stronger critical evaluation and self-correction skills
• Lecturers can encourage iterative learning through drafting, feedback, and refinement cycles
•  Teaching shifts from passive content acquisition to active knowledge construction and improvement

This reinforces AI as a learning partner that supports reflection, refinement, and deeper engagement with disciplinary content, rather than a shortcut to prepared answers.