The effect of metacognitive prompts using generative AI on cognitive load, task performance, and self-efficacy in online self-regulated learning

Self-regulated learning is essential in online environments, where learners must manage large amounts of information, resulting in increased cognitive load and reduced task performance. While metacognitive prompts can support self-reflection and self-management, their classroom integration remains challenging. Recent advancements in generative AI, such as customised ChatGPT, provide new opportunities for more practical integration. To explore this potential, we conducted a study with 40 South Korean university students randomly assigned to either an experimental or a comparison group. All participants watched Python programming video lectures and completed problem-solving tasks using ChatGPT, but only the experimental group received metacognitive prompts. Cognitive load and self-efficacy were assessed through self-reported surveys, and task performance was evaluated based on problem-solving processes, outcomes, and retention tests. The results revealed significant group differences in problem-solving processes, with the experimental group showing a tendency towards lower cognitive load, higher self-efficacy, and improved task performance.