|Abstract: ||傳統計算再測信度的方法有許多缺失，在理論與方法上也較不嚴謹。本文即在介 紹二次式分配準則（ quadratic assignment paradigm, 簡寫為 QAP ）的技術， 並將此技 術應用在再測信度的計算方法上，同時亦舉例討論說明此技術比傳統技術為優之處。結論認 為：二次式分配準則的技術，不僅可以提供一個比傳統方法更客觀、嚴謹、和精確的再測信 度指標，同時也可以驗證該指標的穩定性是否受施測時機、以及隨機性的影響。本文最後建 議在未來的應用上，應該偏向使用此技術，以取代傳統的舊技術。|
The traditional method for computing test-retest reliability suffers from a number of defects. Its underlying methodology is not rigorous either. The purpose of this paper is to introduce the technique of quadratic assignment paradigm (QAP) and to apply this methodology to the computation of test-retest reliability. It also provides an example to show the advantages of the QAP over the traditional method. It is concluded that the QAP can be applied to the computation of test-retest reliability, provide a more objective, rigorous and precise estimate of test-retest reliaility than the traditional method does, and be used to test for significance of stacility and randomness. For future applications, the QAP technique is recommended as a substitute for the traditional method in computing test-retest reliability.