| 摘要: | 本研究旨在探討人工智慧（AI）應用於輔助國小學生進行說書影片製作與分享，透過「AI說書同伴」系統，提升學生的文本理解能力、表達能力與創造力，並降低學生的學習焦慮及增強其自我效能。本研究以趣創者理論（IDC Theory）與鷹架理論（Scaffolding Theory）為理論基礎，透過趣創者理論中的「創造環」（Creation Loop）三個階段─模仿、結合與登台的循環歷程，協助學生逐步從模仿內容、整合自身經驗與情感，到最終公開展示影片。此外，結合鷹架理論所強調的動態調整、漸進撤除及責任移轉等核心概念，提供學生適切的學習支持，以降低其於影片製作與公開演講過程中的認知負荷及焦慮感，進而提升整體學習效能與自我效能。
本研究採用前實驗研究設計（Pre-experimental design）之單組後測設計（One-shot case study），以桃園市某國小三年級18名學生為研究對象。透過問卷調查、訪談及系統使用記錄等方式收集量化與質性資料。量化資料透過描述性統計、信度分析、相關分析及多元迴歸分析處理。質性資料則採主題分析法進行歸納與討論。
研究結果顯示，「AI說書同伴」系統透過即時互動回饋與影片分析建議，有效降低學生說書歷程中的學習焦慮，增強表達技巧與自信心。訪談結果亦指出，多數學生表示能從AI提供的即時回饋中獲得明確且實用的改善建議，有助於促進主動參與及內容創作之動機。此外，學生經歷由模仿與結合到最終登台分享的歷程，亦驗證IDC理論中創造環的有效性。同時，系統提供適時的支持與回饋，引導學生逐步從依賴走向自主學習的過程，呼應鷹架理論核心原則。
本研究建議未來可持續優化系統功能，尤其在提升互動即時性及回饋精準性方面，並建議進一步探討於不同學科領域的應用成效，以擴展AI輔助學習系統之實務價值與理論貢獻。
;This study explores the use of artificial intelligence (AI) to assist elementary school students in creating and presenting book recommendation videos. The AI-assisted system, AI Storytelling Companion, is designed to improve students’ reading comprehension, oral expression, and creativity, while reducing learning anxiety and enhancing self-efficacy. Grounded in Interest-Driven Creator (IDC) Theory and Scaffolding Theory, the system adopts the IDC Creation Loop—Imitation, Combination, and Stage Performance—to guide students through a structured creative process. Students imitate storytelling formats, integrate personal experiences and emotions, and ultimately present their videos publicly. Following Scaffolding Theory principles—dynamic adjustment, gradual fading, and responsibility transfer—the system provides timely support to reduce cognitive load and anxiety, improving learning effectiveness and self-efficacy.
This study employed a pre-experimental design using a one-shot case study involving 18 third-grade students from an elementary school in Taoyuan, Taiwan. Data were collected through questionnaires, interviews, and system usage logs. Quantitative data were analyzed using descriptive statistics, reliability analysis, correlation, and multiple regression. Qualitative data were processed via thematic analysis to extract key learning experiences and feedback.
Findings indicate that the AI Storytelling Companion system effectively reduced learning anxiety and improved students’ expression and confidence through real-time feedback and video analysis. Interviews revealed that students found the feedback clear, helpful, and motivating for participation and content creation. The learning process from imitation to public sharing validated the applicability of the IDC Creation Loop. The system also supported students’ progression from dependence to autonomy, aligning with Scaffolding Theory.
This study recommends continuing to refine the system, especially by improving real time interactivity and the precision of feedback. Future work may explore the effectiveness of applying the system in various subject areas to broaden the practical value and theoretical contribution of AI assisted learning tools. |
