許庭嘉Hsu, Ting-Chia薛惠文Hsueh, Hui-Wen2024-12-172024-08-032024https://etds.lib.ntnu.edu.tw/thesis/detail/c7babe6b463c5fb6db052a462387aba6/http://rportal.lib.ntnu.edu.tw/handle/20.500.12235/123058運算思維是21世紀的人類不可或缺的能力，為了找出能有效提升運算思維的方式，本研究發展一套人工智慧影像辨識採購機器人教材，結合教育機器人、混成式學習和問題導向等有利於提升運算思維的要素，期望透過此課程提升學習者之運算思維。本研究採用準實驗研究法，旨在探討問題導向法對人工智慧影像辨識機器人混成式課程學習者之人工智慧學習成就、程式設計導向運算思維、機器人自我效能及學習行為的影響，課程皆實施於混成式學習環境中，透過實體講述對學習者進行課程重點摘要、補充及檢討，並於實作單元中以教學影片的方式讓學生進行學習。控制組使用傳統講述教學法，教師以實體方式與學生問答互動；實驗組則使用IGGIA問題導向學習法，搭配問答機器人進行系統性的問答，期望透過問題引導方式給予學習者更明確的學習方向，以解決混成式學習中常見數位分心的問題。研究結果顯示，透過問題導向學習法確實能有效減緩學習者數位分心的問題並有效提升其學習成就，且能夠使學習者展現出更多主動學習的行為，然而，在程式設計導向運算思維中控制組有較好的學習表現，自我效能中則沒有顯著差異。Computational thinking is an indispensable skill for individuals in the 21st century. In order to identify effective ways to enhance computational thinking, this research has developed a set of artificial intelligence (AI) image recognition robot materials that combine elements favorable for improving computational thinking, such as educational robots, blended learning, and problem-based learning. It is hoped that through this course, learners' computational thinking can be enhanced. This research employs a quasi-experimental method to investigate the impact of using a problem-based learning in blended learning with AI image recognition robots on learners' outcomes in AI learning, programming-oriented computational thinking, robot self-efficacy , and learning behaviors of learners. Both the control group and the experimental group are in a blended learning environment. The main points of the curriculum are summarized, supplemented, and reviewed for students through physical lectures, and learning is facilitated through instructional videos during practical units. The control group uses traditional lecture teaching, with teachers engaging in physical question-and-answer interactions with students. The experimental group utilizes the IGGIA problem-based learning, coupled with systematic question-and-answer sessions with a chatbot. It is expected that the problem-based learning provides learners with a clearer learning direction, addressing common digital distractions in blended learning. The research results indicate that the problem-based learning method can effectively alleviate learners' digital distraction and significantly improve their learning achievements. It also fosters more proactive learning behaviors among learners. However, in programming-oriented computational thinking, the control group showed better learning performance, and there was no significant difference in self-efficacy.運算思維人工智慧教育機器人混成式學習問題導向學習法Computational thinkingArtificial IntelligenceEducational RobotsBlended LearningProblem-Based Learning學術論文