From prompting to proof: an in-class ChatGPT group chat activity for MEMS research literacy
Year: 2026
Type: conference-paper
Abstract: This work-in-progress innovative practice paper addresses the increasing prevalence of generative AI in engineering education, a shift that introduces substantial risks of error, misinformation, fabricated references, and misplaced trust into students’ study practices. These challenges are especially acute, where students have uneven prior knowledge and conventional teaching may struggle to provide equal structured support. This paper reports an innovative practice implemented in a postgraduate MEMS module, using ChatGPT Group Chat as a shared, synchronous inquiry space for SOI technology. Unlike typical private prompting, the intervention utilised a projected central AI conversation, making prompts, responses, and verification visible to peers. This design encouraged a multimodal learning ecosystem, accommodating direct group-chat prompting, parallel private research, and partner-based discussion. Evaluation drew on an anonymous post-class survey and the classroom activity record. Results indicate that students perceived the activity as effective for learning SOI content (89%) and valuable for increasing awareness of the risks of uncritical reliance (M = 4.89/5), prompt sensitivity (M = 4.78/5), and critical comparison of AI tools (M = 4.78/5). Ultimately, the paper contributes a transferable, low-overhead model for integrating generative AI into engineering education while preserving teamwork, verification, and disciplinary judgement.
Authors: Toh, Bee Yen, Buchanan, Neil
Institution: Queen's University Belfast
Language: English
Subfield: Education
Field: Social Sciences
Domain: Social Sciences
SDG: Quality education
Open Access status: green
OpenAlex ID: W7163602878
