105年國中教育會考數學科試題分析- 以新北市立某國民中學為例
No Thumbnail Available
Date
2017
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
本研究旨在以新北市立某國民中學九年級學生參加105年國中教育會考數學科作答反應資料情形，作為研究者關注學生學習表現與分類的依據，以獲得教學改進資訊。本研究方法依據古典測驗理論、試題反應理論與學生問題表進行試題分析，對於九年級學生學習數學科成果的了解是有正面的幫助。研究結果如下：
一、在試題檢核上，「問題解決」的試題命題佔大多數；在九年一貫能力指標上，所有試題都在能力範圍內；「幾何(S)」的命題數最多。
二、整份測驗的信度值良好，其Cronbach-α係數值約為0.83，表示測驗具有可靠性，試題有良好的測驗品質。
三、在CTT試題分析上，試題平均難易度指數約為0.44，屬於中間偏難；試題平均鑑別度指數約為0.52，具有良好鑑別度；本試題有87%的誘答選項符合誘答辯識原則。
四、在IRT試題分析上，試題Rasch平均難易度指數約為－0.32，屬於中間難度；學生能力平均值約為－0.55；最大訊息量約為12.48出現在能力值為0.00處，顯示本份試題對於中等能力的學生，能提供較多有用的訊息。
五、在學生問題表分析上，差異係數平均值為0.52；試題係數在試題B型& B'型區佔總體的60%以上，本份試題是困難的；學生注意係數在C& C'型區約佔總體的69%，學生學習能力以「中後程度者」佔多數。
六、新北市立某國中試題平均通過率最低的數學內容主題是「幾何(S)」，試題通過率與全國、新北市全體比較差異最大的是「代數(A)」。
依據試題分析的研究結果，本研究對測驗命題者及未來後續的研究提出相關建議。
The aim of this research is to examine one of junior middle school ninth grade student’s responses to test questions in mathematical examination of Comprehensive Assessment Program for Junior High School Students in New Taipei City. The researcher pays attention to students’ study results and performance groupings in order to get information about improving teaching. According to the item analysis of various approaches including classical test theory (CTT), item response theory (IRT) and student-problem chart, the researcher can acquire more useful information about students’mathematical test results.The results were as follows： 1.The analysis of questions on the checklist, the proposition emphasis on “problem solving”；in nine consistent indicators of ability, all questions inside the capacity range；the proposition of questions emphasis on “Geometry”. 2. The whole test reliability is about 0.83, with good reliability, which means that reliability questions, that there is a good test quality. 3.In CTT item analysis, the item difficulty index averages about 0.44 for somewhat hard difficulty；item discrimination index averages about 0.52 for good discrimination；87% options of this test are very good. 4. In IRT item analysis, the item difficulty index on Rasch logic model averages about －0.32 for moderate difficulty；the ability value of students averages about －0.55；the largest testing quantity of the information formula is about 12.48 with an ability value of 0.00, which means that the whole test supplies helpful information for moderate student’s ability. 5. In S-P chart analysis, the disparity index averages about 0.52 inside the acceptable range；the types of B and B' based on the caution index of problem are above 60%,which means that the whole test is difficiult；the types of C and C' based on the caution index of student are about 69% ,which means that most of the students are low-achieved. 6. The lowest average correst rate of mathmatici topic is on “Geometry” at a junior iii high school in New Taipei City；the largest difference of correst rate in the mathmatici topic is on “Algebra” relatively. Based on the results of item analysis, this research will propose some instrumental commentaries and upcoming research.
The aim of this research is to examine one of junior middle school ninth grade student’s responses to test questions in mathematical examination of Comprehensive Assessment Program for Junior High School Students in New Taipei City. The researcher pays attention to students’ study results and performance groupings in order to get information about improving teaching. According to the item analysis of various approaches including classical test theory (CTT), item response theory (IRT) and student-problem chart, the researcher can acquire more useful information about students’mathematical test results.The results were as follows： 1.The analysis of questions on the checklist, the proposition emphasis on “problem solving”；in nine consistent indicators of ability, all questions inside the capacity range；the proposition of questions emphasis on “Geometry”. 2. The whole test reliability is about 0.83, with good reliability, which means that reliability questions, that there is a good test quality. 3.In CTT item analysis, the item difficulty index averages about 0.44 for somewhat hard difficulty；item discrimination index averages about 0.52 for good discrimination；87% options of this test are very good. 4. In IRT item analysis, the item difficulty index on Rasch logic model averages about －0.32 for moderate difficulty；the ability value of students averages about －0.55；the largest testing quantity of the information formula is about 12.48 with an ability value of 0.00, which means that the whole test supplies helpful information for moderate student’s ability. 5. In S-P chart analysis, the disparity index averages about 0.52 inside the acceptable range；the types of B and B' based on the caution index of problem are above 60%,which means that the whole test is difficiult；the types of C and C' based on the caution index of student are about 69% ,which means that most of the students are low-achieved. 6. The lowest average correst rate of mathmatici topic is on “Geometry” at a junior iii high school in New Taipei City；the largest difference of correst rate in the mathmatici topic is on “Algebra” relatively. Based on the results of item analysis, this research will propose some instrumental commentaries and upcoming research.
Description
Keywords
古典測驗理論, 國中教育會考, 試題分析, 試題反應理論, 學生問題表, classical test theory (CTT), Comprehensive Assessment Program for Junior High School Students, item analysis, item response theory(IRT), student-problem chart (S-P chart)