Impact of generative AI chatbots and interaction modes on the speaking proficiency of adolescent EFL learners
Tzu‐Yu Tai, Howard Hao-Jan Chen
Taipei Medical University National Taiwan Normal University
阅读操作
确认中在文库中上传 PDF 后可生成中文音频讲解。
摘要与影响
Speaking proficiency in a second or foreign language (L2) is crucial in language learning; however, it poses challenges and induces anxiety in learners. Recent advancements in generative artificial intelligence (GAI) and natural language processing address these challenges using GAI chatbots. However, further empirical research is required to validate their efficacy in enhancing L2 speaking proficiency. This study investigated the impact of GAI chatbots on the speaking proficiency of 88 eighth-grade Taiwanese English as Foreign Language (EFL) learners participating in a summer program. The participants interacted with CoolE Bot, a GAI chatbot, for 45 min each day, and were divided into three groups: (1) the Bot-S group, who interacted with CoolE Bot via speech; (2) the Bot-SW group, with flexible usage of speech and text; and (3) the non-Bot group, who interacted with teachers and peers in conventional English classes. Quantitative (English-speaking) and qualitative (semi-structured interviews) data were collected and analyzed, revealing that CoolE Bot significantly enhanced the speaking proficiency of EFL learners. Based on participant interviews, the highly interactive and versatile conversational capabilities of CoolE Bot maintained a coherent dialogue flow and handled intricate inquiries. Generating contextually relevant responses based on extended interactions facilitated authentic communication, thus improving speaking proficiency. Incorporating multiple characters, human-like personalities, and synthesized voices sustained the motivation of the learners and engaged them in English-speaking practices. Both speech-only and speech-text interaction modes enhanced speaking proficiency in human-GAI chatbot interaction. Integrating GAI technology with the flexibility of EFL learners to choose interaction modes or utilize supportive functions facilitated cognitive offloading, thereby increasing their English speaking proficiency.
逐年被引趋势
关键指标
同类平均 = 1
同领域 · 同年份 · 同类型
Google Scholar 与 OpenAlex 的被引统计范围不同,数值存在差异属正常。
AI 辅助阅读
依据:摘要
可就本文提问;依据不足时会说明。
学术脉络
学科主题
社会科学EFL/ESL Teaching and Learning
English Language Learning and Teaching · Speech and dialogue systems
参考文献 68
此处列出前 3 条
引用本文 12
按被引量排序,此处列出前 3 条