LLM-Enhanced Exploratory Learning with LaLaMoSuLe: Understanding the Interplay Between AI-Powered Educational Interfaces, Eudaimonic Experience, Autonomy, and Competence Perceptions
摘要
Individuals derive their well-being from various facets of life, one of which is the shape of eudaimonic well-being, which results from a striving for personal growth and meaningful activities. According to recent findings, AI may have an impact on eudaimonic experience in one’s activities, influencing the experience of autonomy, competence, and learning. We asked ourselves whether AI is fundamentally rejected in its form by eudaimonic-oriented individuals, or whether these perceptions can be influenced by the design of the AI interaction. Our developed Large Language Model Supported Learning tool (LaLaMoSuLe) supports explorative learning and vocab training with LLM-based functionalities. We conducted an online study with 228 participants who were clustered using a psychological measurement technique to distinguish between eudaimonic-oriented (seeking primarily for personal growth) and hedonic-oriented (seeking pleasure or comfort) individuals to evaluate the question of a possible aversion to AI use based on an overall attitude or just a question of interaction design. Using two further user experience scales, the results indicate that LaLaMoSuLe can lead to the experience of personal growth and these experiences were very significantly different ( \(p<\) .001***) for eudaimonic-oriented individuals regarding all analyzed dimensions - regardless of whether AI features were included or not. Since the presence of AI does not significantly affect the perception of technology for eudaimonic-oriented individuals, the perception of AI may be a matter of interaction design rather than fundamental orientation in one’s activities, opening up interdisciplinary research on the impact of AI, learning processes, and eudaimonic well-being.