- 期刊
- OpenAccess
合作問題解決自我效能量表發展與科學學習成就關係之探討
Development of the Collaborative Problem Solving Self-Efficacy Scale With the Relationship on Science Learning Achievement
摘要
本研究旨在發展國中生合作問題解決自我效能（Collaborative Problem Solving Self-Efficacy, CPS-SE）量表作為合作問題解決（Collaborative Problem Solving, CPS）課堂評量工具，並應用以檢驗量表與學生科學學習成就的關係。研究對象是方便取樣自臺灣五個縣市的男女合班國中生，共1,045人。基於學生能力國際評量計畫（Programme for International Student Assessment, PISA）2015 CPS評量架構而建構量表的構面及題項。量表分析過程包括五個階段：一、項目分析；二、探索性因素分析；三、信度分析；四、驗證性因素分析；和五、複核效度分析。研究發現三個值得注意的結果：首先，建構效度的過程中，萃取三個因素，包括共享理解、行動解決和團隊組織。其次，每個分量表的Cronbach's α係數皆超過.95，透過結構方程模式（Structural Equation Modeling, SEM）分析競爭模型，選擇適配度最佳且最簡約的模型，顯示模型和資料的適配度良好，表示CPS-SE量表具備良好的信度和效度。第三，自然科學學習成就表現的高成就生與中成就生在共享理解、行動解決和團隊組織等三個分量表，皆顯著優於低成就生，高成就生與中成就生則無顯著差異。
並列摘要
The purpose of this study was to develop the Collaborative Problem Solving Self-Efficacy (CPS-SE) scale as a Collaborative Problem Solving (CPS) classroom assessment tool and to use it to examine the relationship between students' CPS-SE and their science learning achievement. Convenience sampling was used to select 1,045 students from coeducational secondary schools in five Taiwan counties. The dimensions and items of the scale were constructed based on the 2015 CPS assessment framework of the Programme for International Student Assessment (PISA). The scale analysis process comprised of five stages: (1) item analysis, (2) exploratory factor analysis, (3) reliability analysis, (4) confirmatory factor analysis, and (5) cross-validation. Three noteworthy results were found. First, there were three factors extracted through factor analysis during the establishment of construct validity: sharing understanding, taking action on problems, and team organization. Second, the Cronbach's α for each subscale exceeded .95. Applied Structural Equation Model (SEM) the best fit and most parsimonious model was chosen based on model comparisons. Statistical analysis of the model showed goodness-of-fit to the data, which indicated that CPS-SE scale had high consistency and reliability. Third, students who were high and medium academic achievers on science learning scored significantly higher than those who were low academic achievers on three subscales: sharing understanding, taking action on problems, and team organization. There was no significant difference between high-achieving and medium-achieving students on the three subscales.