本研究為了了解國小高年級生對於SDGs陸域保育議題的認知、探究推理能力與決策面向， 因此結合AI與探究式學習， 設計以台灣石虎保育為核心的教學活動， 活動皆為個人活動 每位學習者在AI探究推理助手的引導下 根據情境故事中線索推理石虎發生意外的鄉鎮及傷亡原因， 過程中， 學習者利用真實數據進行驗證及解讀， 步 推理並揭開石虎的生存危機， 學習者在得知這些資訊後 需提出相應的保育專利計畫， 隨後， 學習者參與競價活動， 模擬保育資源的爭奪， 會保保育資源的稀缺與重要性 透過本次探究推理活動以及競價模擬 培養學習者問題解決能力 並加深學習者們對SDGs議題的認知， 提對對陸域保育及續發發的的意 ， 本研究以質性研究為主 量化問卷為輔， 首先 為探討背景不同的學習者在與AI進行探究推理歷程中其探究推理能力 透過背景調查單與對話內容來評分分析探究推理各歷程與整會的能力表現， 其次 為探討學習者執行保育專利計畫決策的考慮面向， 透過保育學習單的決策進行統整與歸納 最後 透過問卷調查蒐集學習者在遊戲活動中的會驗感受與回饋建議 以作為學習成效與活動設計改進的參考 研究結果說明， 學習者在探究推理歷程中以統整與表達表現最佳 而在延伸與主動階段則最弱， 進一 統計分析指出， 喜愛科目與AI使用頻率具中至大效果量， 保育類動物認知層級與科學的覽/競賽經驗具小至中效果量 而性別、推理遊戲與保育活動經驗則僅有有限的效果量影響 且多數學習者整會探究推理能力表現多集中於基礎層級， 其次為熟層層級 最後為加加層層級， 根據保育學習單分析， 大多數學習者執行保育專利計畫的決策面向為環境保育面向， 其次為經發發的面向 最後則是問題導向面向 顯示出學習者對環境保育議題的觀念非常正向， 最後， 本研究活動獲得學習者的普遍好評 認為融入AI探究引導助手的活動流程更有趣更好玩 有助於推理探究過程 並提對對保育相關議題的認知與責任。;To understand upper-grade elementary students’ awareness of SDG terrestrial conservation issues， as well as their inquiry-based reasoning skills and decision-making abilities， this study combined AI with inquiry-based learning to design instructional activities centered on the conservation of Taiwan’s leopard cat. All activities were conducted individually; guided by an AI inquiry-based reasoning assistant， each learner used clues from a scenario-based story to deduce the township where the leopard cat accident occurred and the cause of the injury or death. Throughout the process， learners used real-world data to verify and interpret information， gradually deducing and uncovering the survival crisis facing the leopard cat. After gathering this information， learners must propose corresponding conservation action plans. Subsequently， learners participate in a bidding activity that simulates competition for conservation resources， allowing them to experience the scarcity and importance of such resources firsthand. Through this inquiry-based reasoning activity and the bidding simulation， learners develop problem-solving skills， deepen their understanding of SDG issues， and enhance their awareness of terrestrial conservation and sustainable development. This study primarily employs qualitative research methods， supplemented by quantitative questionnaires. First， to examine the inquiry and reasoning abilities of learners from diverse backgrounds during the process of inquiry and reasoning with AI， we used background survey forms and dialogue transcripts to score and analyze their performance across each stage of the inquiry and reasoning process as well as their overall ability. Second， to explore the factors learners considered when making decisions regarding the conservation patent project， we synthesized and categorized their decision-making processes using conservation worksheets. Finally， a questionnaire survey was conducted to collect learners’ experiences and feedback regarding the game-based activities， which will serve as a reference for evaluating learning outcomes and improving activity design. The research findings indicate that learners perform best in the integration and expression stages of the inquiry-based reasoning process， while they perform weakest in the extension and initiative stages. Further statistical analysis indicates that subject preference and AI usage frequency have moderate to large effect sizes， while knowledge of endangered species and experience with science fairs/competitions have small to moderate effect sizes. Gender， reasoning games， and experience with conservation activities， however， have only limited effect sizes. Furthermore， the overall performance of most learners in inquiry-based reasoning is concentrated at the foundational level， followed by the proficient level， and finally the needs-improvement level. Analysis of the conservation worksheets revealed that most learners prioritized environmental conservation when making decisions regarding the conservation project， followed by economic development， and lastly， problem-oriented considerations. This indicates that learners hold very positive views on environmental conservation issues. Finally， this research activity received widespread positive feedback from learners， who found the activity process—integrated with the AI inquiry guide—more interesting and engaging. They reported that it facilitated the reasoning and inquiry process and enhanced their awareness of and sense of responsibility toward conservation-related issues.
