- 期刊
- OpenAccess
聊天機器人結合擴增實境對大專院校學生科學學習自我效能、學習投入之影響：以科學模擬實驗為例
Effects of Integrating Chatbots and Augmented Reality on University Students' Science Learning Self-Efficacy and Engagement: The Case of a Scientific Simulation Experiment
摘要
本研究探討聊天機器人整合擴增實境提供之學習回饋，對大專院校學習者運用科學模擬實驗學習科學實驗操作之科學學習自我效能與學習投入之影響。研究參與者為89位大專院校學習者，皆無修習過氣體定律相關課程與使用聊天機器人進行學習的經驗，經由隨機分派至傳統學習組、聊天機器人組、擴增實境組、聊天機器人結合擴增實境組，四組學習者在學習前需填寫科學學習自我效能量表前測，在學習後需要填寫科學學習自我效能量表後測、學習投入量表與開放式問卷。根據詹森內曼法分析結果顯示，低科學學習自我效能學習者適合使用傳統學習、擴增實境與聊天機器人結合擴增實境提升科學學習自我效能。科學學習自我效能的高層次認知技能子構面較低的學習者使用傳統學習和擴增實境能促進高層次認知技能子構面。此外，在高層次認知技能子構面中，聊天機器人組與聊天機器人結合擴增實境組之間存在專家反轉效應。學習投入分析結果顯示擴增實境組的學習投入顯著高於聊天機器人組。本研究建議教學者於科學模擬實驗教學前可考量學習者的科學學習自我效能程度予以提供較適合的學習回饋方式。最後，擴增實境教材有助於提升學習者的學習投入。
並列摘要
This study aimed to examine the effect of incorporating chatbots and augmented reality on university students' science learning self-efficacy and engagement in a scientific simulation experiment. A total of 89 students, who did not have the law of gases knowledge and the experience of using chatbots to learn, were randomly divided into four groups: traditional learning group, chatbot group, augmented reality group, and chatbot with augmented reality group. Before the experiment, students completed a pre-test of the science learning self-efficacy scale. After completing the learning task, the students completed a post-test of the science learning self-efficacy scale, an engagement scale, and an open-ended questionnaire. The results of the Johnson-Neyman technique indicated that students with low science learning self-efficacy were suitable for traditional learning, augmented reality, and the chatbot with augmented reality for increasing science learning self-efficacy. For students with low higher-order cognitive skills sub-dimension of science learning self-efficacy, the use of traditional learning and augmented reality helped improve the higher-order cognitive skills sub-dimension. In addition, the expertise reversal effect was shown between the chatbot with augmented reality group and the chatbot group in the higher-order cognitive skills sub-dimension. The engagement of the augmented reality group was significantly higher than that of the chatbot group. It is suggested that instructors can consider students' science learning self-efficacy levels before student learning and choose suitable feedback. Lastly, the format of augmented reality material helps improve students' engagement.