| dc.description.abstract | Geometry learning is important. However, usually, using textbooks in class makes students lose motivation because of too much emphasis on memorizing and understanding geometry formulas without applying them to solve daily life problems in their surroundings.
Based on the aforementioned issues, in this study, we proposed a geometry learning application focusing on 3D objects, namely Smart-3D UG with the Smart QAC Mechanism. This application was developed on a mobile device with two proposed main functions: (1) Augmented Reality (AR) feature for enabling 3D visualization of real geometry objects in authentic context (i.e., surrounding area); (2) Smart Questions and Clarifications (QAC) mechanism with Artificial Intelligence (AI) to generate meaningful questions and clarifications related to student’s geometry learning practice. In addition, we enabled multiple representations of the proposed Smart QAC, i.e., visual representation, logical representation, and linguistic representation.
We designed three stages of learning activities (i.e., contextualization, personalization, and socialization). In each stage, students need to answer the geometry-related questions before they receive the clarifications from AI. To investigate the influence of the Smart QAC Mechanism on student learning achievements, such as geometry ability (GA), spatial ability (SA) and estimation ability (EA), we conducted one quasi-experiment with 76 fifth-grade students. They were divided into three groups, i.e., the experimental group (EG) and two control groups (CG-A and CG-B). There were two tasks in the experiment. I.e., teacher design and free Exploration. The experiment lasted for nine weeks, and mixed-method analysis was used.
The results revealed some interesting findings. First, the EG significantly outperformed the CG-A and CG-B on learning achievement related to GA, SA, and EA. Second, The Smart QAC mechanism was strongly correlated with learning achievement. Furthermore, EG demonstrated higher measurement precision than CG-A, particularly in teacher design activities and free exploration. This result indicates that the smart QAC helps EG students be more successful at measuring precise objects. Moreover, the error rate of EG is less than CG-A and reaches a significant difference in free exploration activity. It implied that our proposed smart QAC mechanism enhances geometry understanding and ability in authentic contexts. More importantly, students showed a positive attitude and high intention toward Smart 3D-UG with the smart QAC. Therefore, enhanced smart QAC mechanisms can potentially be applied in class to support more complex geometry learning in authentic contexts. | en_US |