結合知識結構之Q矩陣設計於DINA模式之估計成效探究

The Research in DINA Model with Different Q Matrix Designs Incorporating Knowledge Structures

本研究提出結合專家知識結構與學生知識結構之Q矩陣設計，透過實證資料探討結合知識結構之Q矩陣在認知診斷模式—DINA模式的估計成效；並探究DINA模式和IRT模式試題參數的相關性。研究結果顯示：（1）DINA模式猜測參數g與IRT模式猜測參數之間沒有顯著相關，DINA模式1-s-g參數與IRT模式鑑別度參數之間有顯著相關。（2）結合專家知識結構與結合學生知識結構的Q矩陣，都能提升認知屬性之診斷辨識率。（3）在Q矩陣中加入專家知識結構辨識率較學生知識結構佳。

The purpose of this study was to explore the performance of cognitive diagnostic model with different Q matrix designs, the expert knowledge based structure and the student knowledge based structure. Otherwise, the relations of item parameters between DINA model and three-parameter IRT model were explored. The results showed in the following: 1) There is significant relationship between 1-s-g parameter in DINA model and the discrimination parameter in IRT model;2) The diagnosis accuracy of DINA model increased when the Q matrixes weredesigned with the expert knowledge structure and with the student knowledgestructure; 3) With two types of the Q matrixes, the performance of the expert knowledge structure was better than that of the student knowledge structure.