科學建模已被視為科學教育與實踐不可分割的基本學習目標。將計算工具融入科學實踐提供學生與科學模型互動、探索並建構科學現象的機會，使學習者能有效參與科學建模並對科學現象有更深的理解。然而，由於計算建模過程的複雜性涉及對特定領域知識的理解以及應用計算工具實踐的能力，鷹架引導與教學法的介入不可或缺。因此，本研究提出了一個基於系統思維理論之教學法框架，藉由結構化地分析與辨識複雜系統中的重要組成元素以及元素間的交互動態與靜態關係，引導學生運用計算工具建立複雜場景的科學模型。本研究並結合演示法與提問法兩種教學鷹架引導學生藉由系統思維建立模型，並觀察不同鷹架輔助是否影響學生在建模活動中的計算建模表現和行為模式。受試者為兩班10 年級學生參加為期九週的計算建模課程，兩班學生被隨機分配到觀察提示組(N=29)和演示組(N=36)。本研究蒐集學生紙本測驗的回答和於計算建模平台實作建模測驗之成品與行為，並進一步評估計算建模表現、行為策略。結果顯示計算建模框架是引導學生學習如何建構科學模型並進一步轉化為計算形式的有效方法。然而，研究結果也發現學生在兩種教學法下展現不同程度之模型品質和建模策略。此外，我們發現先備知識對學生的建模策略和建模表現也有不同程度的影響。;Scientific modeling has been advocated as an essential learning objective of science education. Integrating computational tools with science practice opens up a chance for learners to engage in effective scientific modeling as it allows students to interactively manipulate the scientific models and explore the obscure mechanisms behind the scientific phenomena. However, due to the complexity of the model construction process, scaffolding and effective intervention are needed to provide students with guidance in computational modeling activities. This study thus proposed a systems thinking pedagogical framework based on the theory to provide a guideline for educators and learners to learn how to construct a computational model. The framework was composed of five modules aiming to raise students’ awareness of identifying critical elements in the system and the identification of the interconnections of the elements in the systems during modeling. Demonstration and question prompt scaffolding were applied to guide students to construct computational models under the computational modeling framework to reveal whether the pedagogical scaffolding impacted their computational modeling performance and action during modeling activities. Two intact classes of 10th grade students participating in the 9-week computational modeling program were randomly assigned to the question-prompt group (n = 29) and demonstration group (n = 36). The computational modeling performance was collected through paper-based tests and a hands-on modeling test which asked students to construct a computational model on the platform. Students’ answers, product of computational modeling, and their actions were then evaluated and used for further analysis. The results showed that implementing such a framework is an effective approach which guides students to learn how to construct scientific models and further transfer into computational forms. However, this study also found that the students who learnt under the two types of scaffolding demonstrated different learning outcomes in the hands-on test and modeling strategies. Besides, we found that prior knowledge also had a different level of impact on the students’ modeling strategies and modeling performance.
