AI 영어회화 앱 의인화 유형에 따른 사용자 경험 차이 분석 = Analyzing User Experience Differences by Anthropomorphism Type in AI English Conversation Apps

다국어 초록 (Multilingual Abstract)

The advancement of artificial intelligence (AI) technology is reshaping the landscape of English education, with AI-based anthropomorphic conversation applications offering learners highly immersive and personalized learning experiences. This study examines the impact of conversational AI anthropomorphism types on user experience (Immersion, psychological burden, and learning satisfaction) and explores differences according to learners’ proficiency in English conversation. To this end, eight AI conversation applications were analyzed to identify four anthropomorphism types: text/voice-based, animal character- based, human character-based, and human-like interfaces. One hundred adult learners were then divided into high- and low-proficiency groups and asked to perform identical tasks with each interface. Quantitative and qualitative data were collected through surveys and in-depth interviews. The analysis revealed that human-like interfaces produced the highest overall Immersion and learning satisfaction but tended to increase psychological burden for lower-proficiency learners. In contrast, animal character-based and text/ voice-based interfaces were more effective in maintaining Immersion and reducing burden among lower- proficiency learners. These findings demonstrate that the effects of AI anthropomorphism vary depending on learners’ proficiency levels. The study highlights the necessity of tailoring interface types to learner proficiency and is expected to provide practical insights for the design of personalized user experience (UX) and the development of AI-based English education platforms.