建模實驗促進學生的讀圖素養與曲線擬合：來自測驗、眼動、姿勢，與口語分析的證據

This study conducted two main experiments. The results of the first experiment has already been translated into English and submitted to Physical Review Physics Education Research. This study’s purpose was to explore 52 non-science major students’ scientific descriptions and model-building behaviors with the aid of a technology-based laboratory. Results showed that while 80% of the participants used propositional statements to describe the given phenomenon before using the technology-based laboratory, they utilized more image-based and mathematics-based representations after using the laboratory. Regarding model-building behaviors, most participants illustrated their difficulties in connecting mathematical representations with actual phenomena. The second experiment investigated 92 college students’ processes of scientific argumentation with kinematic phenomena. The participants were assigned to the high- and low-ability groups based on their scores in the Test of Understanding Graphs in Kinematics. We then explored the participants’ oral presentation and eye-movement behaviors as they used various data types (i.e., table, graph plot, and equation) in three argumentation evaluation tasks and one argumentation generation task. Please see below for the three sections of data analyses that have been completed or are on-going. Experiment 2–1 also explored high/low- ability college students’ eye-movement behaviors in their processes of argumentation in argumentation evaluation task and argumentation generation task. Results showed that the high-ability group demonstrated more transitions between representations, greater deliberative thinking, longer total fixation duration, and used more equations than the low-ability group. It revealed that experts could not only connect and integrate the multiple representations, but also be more able to adopt a gaze aversion strategy to think longer on tasks. The sequential analysis of high-/low- ability students’ eye-movement behaviors in the four tasks have executed, and the result section is still on-going. Experiment 2–2 has been accepted by the 2019 International Conference of Science Education in Taiwan. This study investigated high/low- ability college students’ eye-movement behaviors as they proved their points through arguments with multiple representations. Results indicated that the high-ability group contained stronger simple interaction effect than the low-ability one in terms of the percentage of single representation’s fixation duration in the total fixation duration and the FD/P ratio (i.e., the total fixation duration for each paragraph was divided by the area of each text paragraph), indicating that the high-ability students are more capable of transforming one multiple representation into another. Meanwhile, we also found that the high-ability group paid more attention to mathematical equations than the low-ability one, and they used representations in accordance with the tasks. Experiment 2–3 analyzed the college students’ processes of argumentation and the strategies they adopted. Furthermore, one subject was selected to investigate to what extent the eye-mind hypothesis applies in scientific argumentation. Results indicated that the low-ability students’ argumentation level were lower, including only claim, data and/or warrant. In contrast, the high-ability students’ argumentation level was much higher. However, no significant difference was found in the accuracy of the argumentations between the two groups. The examination of the participant’s argumentation strategy and the subject’s eye-movement pattern are still on-going.