Applying the traffic light assessment framework to an AI-integrated coursework [Abstract]

Vaidhiyanathan, Ramalakshmi 2026. Applying the traffic light assessment framework to an AI-integrated coursework [Abstract]. Presented at: ITiCSE 2026: ACM Conference on Innovation and Technology in Computer Science Education, Madrid, Spain, 10 - 15 July 2026. ITiCSE 2026: Proceedings of the 31st ACM Conference on Innovation and Technology in Computer Science Education V. 2. New York, NY: ACM, pp. 799-800. 10.1145/3803401.3811964

PDF - Published Version Available under License Creative Commons Attribution. Download (865kB)

Abstract

This article discusses the observations on the deployment of Cotterell's Traffic Light assessment framework, specifically the Green Light approach, within a Year 2 Object orientation, algorithms, and data structures module at Cardiff University. Under a Green Light policy, students were explicitly permitted and expected to use generative AI (Gen AI) tools as part of their object-oriented design and implementation in a coursework. Contrary to expectations, several unintended outcomes were observed: reluctance to engage with AI, over-reliance on AI for reflective writing, generation of architecturally complex code beyond task scope using AI, and persistent fear of academic integrity consequences despite explicit permissions. We analyze these outcomes against the strengths and limitations of the Green Light approach and offer practical recommendations for computing educators designing AI-inclusive assessments.

Item Type:	Conference or Workshop Item - published (Paper)
Date Type:	Published Online
Status:	Published
Schools:	Schools > Computer Science & Informatics
Publisher:	ACM
ISBN:	9798400726330
Date of First Compliant Deposit:	15 July 2026
Last Modified:	15 Jul 2026 11:00
URI:	https://orca.cardiff.ac.uk/id/eprint/188233

Actions (repository staff only)

Edit Item