本研究旨在建立一套以國小五年級數學領域「長方體與正方體」能力 指標為評量內容的電腦適性診斷測驗系統,同時輔以學生知識結構結合貝 氏網路之電腦補救教學模組,除了診斷出學生能力指標的學習成效外,還 可以讓學生進行自我學習,以此模式反覆練習達臻嫻熟之境地。 本研究首先分析指標內容,建立能力指標專家知識結構並建立結合知 識結構與貝氏網路之架構,並依此結構命題,進行紙筆診斷測驗,測驗完 成後,再依據試題順序理論和貝氏網路理論,建立電腦適性診斷測驗施測 流程,以進行測驗及評估補救教學之成效。 本研究發現: 1. 電腦適性診斷測驗施測的平均施測題數是 21.25 題,與紙筆測驗相 比,平均可以節省 7.75 題。 2. 經過電腦適性補救教學後,學生的平均分數有進步,達到顯著差 異。 所以,本研究所提出之電腦適性診斷測驗系統和補救教學模組,確實 可以「因材施測」與「因材施教」,並且有良好的學習成效。
The aim of this study is to design a Computerized Adaptive Diagnostic System and Computerized Adaptive Remedial Instruction based on the student`s knowledge structure and Bayesian network theory which using the competence indicators with " The Volume and Content of Cuboid and Cube" of grade 5 in Nine-Year Compulsory Curriculum as an example . The student`s learning performance of the target competence indicators not only can be diagnosed but also provided the chance for student to learn by himself and get into the way with this model . Firstly , we analyzed the contents of textbooks, the structures of competence indicators were constructed . Then, the network with the theory of knowledge structure and Bayesian network were constructed ,too. After that , the items of test basing the framework were designed . The Ordering theory , semantic structure analysis, and Bayesian network analysis are used to decide the students' knowledge structure,model assessment data, and identify bugs and sub-skills,so those parameters can be used in the system . Some conclusions are finded and briefly outlined as follows: 1. The number of items tested by students in the Computerized Adaptive Diagnostic Test System is 21.25 averagely. This system can save 7.75 items averagely, and the test-taking time is also saved simultaneously. 2. The average score of students is remarkably improved after taking the Computerized Adaptive Remedial Instruction. Therefore, the two systems ─ Computerized Adaptive Diagnostic Test System and Computerized Adaptive Remedial Instruction, proposed in this study can factually test and remedy students’ abilities individually.