利用全方位設計原則編製國小數學成就測驗之初探研究：以四年級整數四則運算為例

評量是了解學生學習成果的重要手段，但如何讓學生能公平參與考試一直是重要的議題。本研究依全方位設計原則編製國小整數四則運算成就測驗，並分析其信效度；接著以國小五年級學習障礙學生和一般學生各20名為對象，探討兩組學生在兩種形式之複本題目(紙本和電腦報讀)的作答表現是否有差異，以了解在遵循此原則下所編製之試題可否減少無關變異構念的干擾；此外，也探討受試者對兩種版本的作答感受。結果發現，自編成就測驗之古典測驗理論分析結果具有良好鑑別度和信效度，試題難度為.75、鑑別度為.34、Cronbach α係數值為.85、複本信度為.86。在應用研究中則發現，學生組別和測驗方式間無交互效果，自編成就測驗中作答表現會因學生能力不同而有所差異，但不會因測驗呈現方式不同而有所差異；不管是學習障礙學生或一般學生均偏好在電腦化語音報讀版本進行測驗。

關鍵字

全方位設計評量；整數四則運算測驗；學習障礙

並列摘要

Assessment is one of the major methods to evaluate students' learning outcomes. How to make students participate in assessment fairly has been an important issue. This study aimed to develop a mathematical achievement test for fourth graders based on the principles of universal design. An experiment was conducted to examine the effectiveness of the universally-designed test after its reliability and validity had been confirmed. Twenty fifth graders with learning disabilities and their 20 typical peers were recruited to take the test displayed in two forms, paper-based and computer-based. In addition, the participants were asked to fill out a questionnaire to show their preference for the form of the test they took. The reliability and validity of this achievement test were acceptable. The average difficulty index of the items was .75, the average discrimination index of the items was .34, the reliability coefficient of Cronbach' Alpha of the test was .85, and the reliability coefficient of parallel-forms (paper-based and computer-based) was .86. Although the performance difference between students with ∕ without learning disabilities was significant, there was no significant difference between two forms of the test. There is no significant interaction between the student groups and the test forms either. However, both groups preferred the computer-based test.