[[alternative]]The effects of performance assessment on the mathematics achievement and attitude of elementary students

[[abstract]]本研究主要目的在探討實作評量對學生的數學成就與數學態度的影響。本研究採單因子獨立樣本實驗設計，實驗組一接受實作評量與討論教學、實驗組二接受實作評量與講述教學、實驗組三接受傳統評量與講述教學，各組均接受八次的實驗處理。實驗研究樣本來自台北縣某國小六年級3個普通班的學生，樣本總人數為102人，隨機分派為三個實驗組，每組34人。 使用的研究工具有「數理科實作評量工具」、「國民教育階段學生基本學習成就評量」、「柯氏國民小學數學科成就測驗」、「國民小學國語文成就測驗」、「國小系列學業性向測驗」、「數學態度量表」等六種。研究者將學生起始能力（包括數學能力、國語能力與智力），依據百分等級25以下、25-75、以及75以上，區分為低、中等、高三個能力層次。本研究以變異數分析、共變數分析等統計方法進行資料分析。本研究主要發現如下： 一、 在數學成就方面 1.控制數學、國語、智力與實作評量前測等共變因素後，在數學成就和傳統評量上，實驗組三高於實驗組一與實驗組二；在實作評量上，三個實驗組未達統計的顯著差異。2.控制國語、智力與實作評量前測等共變因素後，在數學成就、實作評量、傳統評量上，三個數學能力層次均未達顯著差異。3.控制數學、智力與實作評量前測等共變因素後，在數學成就、實作評量、傳統評量上，三個國語能力層次均未達顯著差異。4.控制數學、國語與實作評量前測等共變因素後，在數學成就和實作評量上，中、高智力組均高於低智力組；但在傳統評量上，只有中等智力組高於低智力組。5.不同的數學、國語、智力等起始能力與實驗處理，在數學成就、實作評量、傳統評量上，均無顯著的交互作用。 二、 在數學態度方面 1.在數學態度總量表，及在學習數學的信心、數學有用性、數學探究動機、對數學成功的態度、重要他人的數學態度、數學焦慮等分量表上，三個實驗組均未達統計上的顯著差異。2.在數學態度總量表，及在學習數學的信心、數學探究動機、數學焦慮等分量表上，高數學能力組高於中、低數學能力組，又中等數學能力組高於低數學能力組。在數學有用性上，高數學能力組高於中、低數學能力組。在對數學成功的態度、重要他人的數學態度上，三個數學能力層次均未達顯著差異。3.不同的數學起始能力與實驗處理，在數學態度總量表，及在學習數學的信心、數學有用性、數學探究動機、對數學成功的態度、重要他人的數學態度、數學焦慮等分量表上，均無顯著的交互作用。 本研究並根據研究結果，對教育上的應用及未來研究提出建議。[[abstract]]The purpose of this study was to investigate the influence of performance assessment on mathematics achievement and attitude. An independent experimental design was used. One hundred and two year-six students were selected from the normal classes of an elementary school in Taipei city. These students were randomly assigned into three groups. GroupⅠ received training with performance assessments and discussion teaching; GroupⅡ received training with performance assessments and narration teaching; GroupⅢ received training with traditional tests and narration teaching. The subjects were also assessed on their math ability, language ability, intelligence and performance assessment pretest in the initial stage of the experiment. All the students were finally tests with both a traditional test and a performance assessment. Six instruments were used in this study. They were ＂The Performance assessment for Science and Mathematics＂, ＂The Elementary Education Basic Abilities Test＂, ＂The Ko Elementary Math Achievement Test＂, ＂The Elementary School Language Achievement Test＂, ＂Math Aptitude Test＂, and ＂The Math Attitude Scale＂. ANCOVA and ANOVA were used to analyze the data. The results indicated that: 1. When the math ability, language ability, intelligence and performance assessment pretest variables were controlled, there were significant differences between groups in the traditional test score and average math achievement with Group Ⅲ better than the other two groups, but no significant difference between groups in performance assessment was obtained. 2. When the language ability, intelligence and performance assessment pretest variables were controlled, math abilities did not result in significant differences in average math achievement, performance assessment and traditional assessment. 3. When the math ability, intelligence and performance assessment pretest variables were controlled, language abilities did not result in significant differences in average math achievement, performance assessment and traditional assessment. 4. When the math ability, language ability and performance assessment pretest variables were controlled, significant intelligence differences in average math achievement and performance assessment were obtained with high-intelligence group and middle-intelligence group better than the low-intelligence group; for performance in the traditional test, the middle-intelligence group was better than the low-intelligence group. 5. Math ability, language ability, and intelligence did not interact with assessment methods on the average math achievement, performance assessment and traditional assessment. 6. There was no significant difference in math attitude between assessment methods, but significant math ability differences in math attitude were obtained in that the high-math ability group had the most positive attitude, followed by the middle-math ability group and then the low-math ability group. 7. There was no significant interaction in math attitude between math abilities and assessment methods. Some recommendations for the field of elementary education and future study are suggested