|Abstract: ||本研究旨在透過完整的量表建構系統，建構海洋科學學習成效量表，以應用各種海洋科學活動學習之評量。本研究選取「2017全民科學週-海洋Fun學趣」活動資料，並以逐字編碼與文獻進行關案要素的分析；再透過專家德懷術確認結構及發展題項，於「2018全民科學週-海洋Fun學趣」進行量表預試，資料分別以探索性因素分析及驗證性因素分析建構效度，完成海洋科學學習量表。 本研究發放372份問卷，回收率100%，分別以SPSS及AMOS兩種系統進行二次信、效度分析，研究結果如下： 一、 海洋科學學習成效量表分為：「海洋科學知識」、「海洋影響力」、「海洋生態」、「海洋資源與保育」、「海洋管家精神」、「海洋意識」、「教育推廣能力」、「問題解決能力」等8個分量表，共29題。 二、 量表信度考驗：整體Cronbach’s α信度為0.962，各分量表Cronbach’s α信度介於0.794~0.878之間，具有良好的內部一致性。 三、 量表效度考驗：本量表以探索性因素分析歸類題目構面，驗證性因素分析估算理論模式，並以模式適配性評估模式適配情形，結果顯示適配度良好。 未來可將本量表應用於各項海洋科學活動評量當中，以作為活動成效之評量標準，並累積海洋科學教育成果。|
In this study, through the comprehensive inventory construction system, this research built a marine science learning achievements inventory to apply the assessment of various marine scientific activities. A study collected data from "2017 Civil Science Week-Ocean Fun Learning" activities, and analysis dada, for verbatim coding, and literature analysis of the key elements. And then through the expert Delphi check the scope of the development of structural and project. Finally, in the 2018 Civil Science Week-Ocean Fun learning activity carried out the inventory test to survey the students’ marine science learning achievement.The data used exploratory factor analysis and confirmatory factor analysis to construct the validity, complete the marine science learning scale. In this study, 372 questionnaires were issued with a recovery rate of 100%. The two systems of SPSS and AMOS were used for the second-level reliability and validity analysis. The results are as follows: 1. This marine science learning achievementscale is divided into eight factors: "Marine Science Knowledge," "Marine Influence," "Marine Ecology," "Marine Resources and Conservation," "Marine Stewardship," "Marine Consciousness," "Education Promotion," and "Problems. "Solve ability" with a total of 29 questions. 2. The reliability test of the scale: The overall Cronbach’s alpha reliability is0.962, and the Cronbach’s alpha of each subscale is between 0.794 and0.878, with good internal consistency. 3. The validity of the scale test: This scale isprimarily sorted the factors by exploratory factor analysis,and used confirmatory factor analysis to estimate the theoretical model. The results show a good model fit with 29 questions in eight factors which include four factors related knowledge, two factors for affection, and two factors for skills. In the future, this scale can be applied to the assessment of various marine science activities as a measure of the effectiveness of activities and to accumulate marine science education results.