本研究試著以國中自然生活科技八年級之「聲音」單元作為探討,結合了貝 氏網路、OT順序理論(Ordering Theory)為基礎的教學內容,尤以透過電腦呈現 建構適性題庫,這是傳統紙筆測驗所無法表達出來的迷思概念,就可以透過這一 連串的設計,教師可以透過這一連串的實驗過程,分析學生的迷思概念以及需要 補強的部分,於是針對缺點進行補強的工作。研究結果顯示,使用建構反應試題 得以有效評估學生之技能。 研究結果發現: 1.試題信度高達.82,本研究單元試題之適性測驗的省題率,當閾值為.05 時, 精準度達.95,此時省題率為14.2%。 2.整個測試能立即診斷出受試者所犯的錯誤類型和所學會的技能有哪些,平均 正確率達到98.05%。 由以上結果可知,本研究所建置的電腦化建構反應題與診斷模式開發,可達 到因材施教目的,對於學生學習能力可顯著提升教學成效的。
This study tried to state the ”voice” element in eighth-grade science and technology of natural life is used as a combination of Bayesian network based learning content and OT Theory(Ordering Theory). Especially through the construction of computer adaptive question bank show, which is expressed in the traditional pencil and paper tests are not misconceptions. Teachers through the process of this series of experiments to analysis of student misconceptions and the need to reinforce some of . The results showed that the use of constructed response questions to effectively assess the skills students. The results showed: 1.The reliability of the test was .82. 2.The test subjects immediately diagnosed type mistakes and learn the skills by which the average accuracy rate to 98.05 percent. These results show that, in this study build computerized diagnostic model constructed response questions and the development can be individualized to achieve the purpose for learning ability can significantly enhance teaching effectiveness.