| 摘要: | 在實作導向的STEAM教育課程中，學生常因缺乏即時協助或理解的落差，操作錯誤而在程式設計與機器人組裝過程中遭遇學習瓶頸。本研究聚焦於國小探究式STEAM機器人教學活動，導入具備即時回饋功能的生成式人工智慧，協助學生理解概念、排除錯誤，以順利完成學習任務。課程設計融合6E教學模式，引導學生經歷動機引發、問題探索、概念解釋、工程實作、延伸挑戰與反思評估等階段，讓學習具有連貫性，逐步建立對知識的理解。研究目的在於探討生成式人工智慧於課堂中對學生學習動機、科技知識、課程參與行為及作品表現得影響，並透過比較實驗組與對照組的學習表現，進一步評估其教學應用價值。
本研究對象為台灣北部地區國小三年級至六年級共48位學生，分為實驗組26人與對照組22人，進行為期八週的STEAM棒球機器人課程。實驗組學生在課程中使用具對話功能的生成式人工智慧作為學習輔助工具，協助理解程式邏輯與結構設計，對照組則採傳統教學進行。課程以棒球為主題，融合程式設計、結構組裝、感應器運用與球場任務挑戰，並搭配動手實作，引導學生完成機器人製作。系統設計部分，結合生成式人工智慧的互動對話功能，讓學生在操作過程中能即時提問與獲得回應，有助於理解課程內容與思考可能的解決方式。透過前後測問卷、任務成果與學習過程記錄進行資料蒐集與分析，探討生成式人工智慧在實作課程中的應用成效。
本課程採用6E教學模式，生成式人工智慧在不同階段中扮演多重角色。在參與階段作為催化劑，引發學生的興趣與學習動機；在探索與解釋階段，提供學習資源與概念說明，引導學生深入理解內容；進入實作與延伸階段，協助調整設計與拓展想法；在評估階段中，提供回饋與建議，引導學生整理學習成果。這樣的教學安排讓學生在各階段都能獲得適時協助，讓學習過程能夠順利進行。
研究結果顯示，將生成式人工智慧融入探究式學習與6E教學模式，確實有助於提升小學生在機器人課程中的學習興趣、參與度以及心流體驗，並促進他們在解決問題與運算思維方面的表現。透過生成式人工智慧提供即時回饋與設計建議，學生能在學習歷程中持續修正與最佳化作品，降低挫折感，提升學習動機與課堂參與意願。生成式人工智慧輔助教學能有效強化學生在跨學科情境中的應用能力，並促進創意表現與合作能力，為國小STEAM課程帶來更多元的教學可能性。未來建議可擴大研究對象範圍，並嘗試不同學科與學習情境的應用，進一步探討生成式人工智慧在教學現場中的實用性與成效，期望為機器人教育與生成式人工智慧輔助教學模式提供更多具參考價值的實證資料。
;This study explores the integration of Generative Artificial Intelligence (GenAI) into robotics programming education to enrich inquiry-based science learning, particularly in the STEAM (Science, Technology, Engineering, Arts, and Mathematics) domains, with a focus on its impact on elementary science education. Through hands-on STEAM activities, students enhance problem-solving skills, collaboration, and develop a strong interest in science learning. Utilizing Scratch, a free and open programming language, students not only learn programming basics but also deepen their understanding and application of scientific concepts. The research targets elementary school students, incorporating technology, mathematics, and physical education into a series of STEAM education experiments. The importance of enhancing science learning, fostering creativity, and cultivating teamwork.
Inquiry-based science learning encourages questioning, investigation, and knowledge construction through exploration and experimentation. Recent advancements in Artificial Intelligence, especially Generative AI (GenAI), offer novel tools to enhance this educational approach. This paper examines how integrating GenAI can enrich the learning experience, focusing on a STEAM project involving the design and implementation of a baseball robot.
Utilizing the 6E experiential learning model, GenAI assumes multiple roles across the learning stages. Initially, in the Engage phase, GenAI acts as a catalyst, captivating student interest through Scratch, thereby igniting curiosity. In the subsequent Explore phase, GenAI transitions into a mentor, providing tailored learning pathways and resources, facilitating guided exploration. As the learning progresses into the Explain phase, GenAI transforms into an instructor, simplifying intricate concepts and theories through textual content. During the Engineer phase, GenAI serves as a design assistant, assisting students in utilizing tools like LEGO SPIKE for project development. Moving forward to the Enrich phase, GenAI becomes an inspiration, expanding students′ knowledge and fostering interdisciplinary integration and innovative thinking. Finally, in the Evaluate phase, GenAI transitions into an assessor, delivering real-time feedback and assessments to aid students and teachers in reviewing and reflecting on learning outcomes. GenAI plays a crucial role in scientific inquiry activities, offering expertise, guidance, and support throughout the project phases, thereby enriching students′ learning experiences and fostering knowledge exchange in STEAM fields.
The results of this study indicate that integrating generative artificial intelligence (GenAI) into inquiry-based learning and the 6E experiential learning model effectively enhances elementary school students’ learning interest, engagement, and flow experience in robotics courses. It also improves their performance in problem-solving and computational thinking. Through the real-time feedback and design suggestions provided by GenAI, students were able to continuously revise and optimize their projects during the learning process, thereby reducing frustration, increasing learning motivation, and improving their willingness to participate in classroom activities. The findings suggest that AI-assisted instruction can significantly strengthen students’ interdisciplinary application abilities while fostering creativity and collaborative skills, offering more diverse teaching possibilities for elementary school STEAM education. It is recommended that future research expand the sample size and explore applications across different subjects and learning contexts to further examine the practicality and effectiveness of generative AI in educational settings. Such efforts are expected to provide valuable empirical evidence for the development of robotics education and AI-assisted teaching models. |
