Abstract:

The continuous advancements in large language models (LLMs) and agent technologies have made LLM-based agents a promising new tool for enhancing the quality and efficiency of teaching and learning in the educational domain. Based on the functional positioning of LLM-based agents, single-agent systems are capable of supporting various teaching and learning tasks, such as content generation, intelligent feedback, and assessment. However, due to the homogeneous nature of their interaction characteristics and functional attributes, single-agent systems face limitations in promoting deeper cognitive development. In contrast, multi-agent systems can simulate various educational roles, enhancing the diversity and depth of learning interactions and thereby providing more personalized and profound learning experiences. Considering that the application of agents in the learning process primarily relies on learners’ initiative, this study based on the ARCS model designed motivational learning activities for both single-agent and multi-agent systems to ensure the effective implementation of learning activities and conducted a quasi-experimental study in the context of English reading. The experimental results reveal that motivational learning activities supported by multi-agent systems significantly outperform single-agent systems in improving students’ performance in reasoning, evaluation, and application tasks. Furthermore, multi-agent systems exhibit stronger motivational effects and effectively promote deeper cognitive development, particularly in abstraction and generalization skills. The study underscores the value of multi-agent systems in supporting deeper learning and provides insights for further exploration of their applications in education.

Key words: large language model, multi-agent, ARCS model, motivational learning activities, deeper learning