Speaker
Description
The growing use of Generative AI has changed how students approach academic work, making conversational AI a common source of support during learning. While existing research has examined AI adoption and its impact on academic performance, less is known about what motivates students to seek AI assistance during a learning task. This exploratory study examines how elite scholarship students engage in AI-enabled cognitive offloading during an authentic academic assignment. It introduces student friction, a proposed concept describing the cognitive and emotional resistance students experience before shifting from independent problem-solving to AI-assisted support. Guided by Cognitive Offloading Theory and metacognitive regulation, data will be collected during a single task-based session with a purposive sample of government-funded scholarship recipients, including AI interaction logs, prompting behaviour, task completion records, and post-task questionnaire responses. Rather than focusing only on usage frequency, the study explores how students decide to use AI, refine their prompts, and evaluate AI-generated responses, offering initial evidence for the proposed framework and practical insights for designing responsible AI use in higher education.