以魏氏成人智力量表第三版四因素為基礎的簡短式測驗組型：臨床應用指引

Selection of Factor-Based WAIS-III Tetrads in the Taiwan Standardization Sample: A Guide to Clinical Practice

本研究使用台灣版魏氏成人智力量表第三版之標準化樣本，共888位16至84歲之正常成年人，來檢視以魏氏成人智力量表第三版四因素為基礎的簡短式測驗組型。以各種心理計量特質、時間限制、以及比例與線性方程式的質性估計程序來比較54組測驗組型，結果支持使用線性方程式的程序。「類同－矩陣推理－算數－數字符號」組型在效度、內容代表性、以及節省時間的考量上，都比其他組型為優。Kaufman（1990）的「類同－圖畫補充－算數－數字符號」組型，以及「常識－圖畫補充－算數－數字符號」組型兩者也都是整體而言相對較佳的估計組型。對於認為圖形設計分測驗可提供高度臨床資訊的臨床工作者，另外兩個組型，「類同－圖形設計－算數－數字符號」與「常識－圖形設計－算數－數字符號」，亦可以提供很好的估計品質。然而，目前的發現顯示即使是最好的組型還是有相當程度的錯誤分類比率，而且這些以四因素為基礎的測驗組型傾向會低估能力很高的正常成年人。使用者必須注意到這些簡短版的測驗組型，主要目的僅作為篩檢的用途。本研究結果提供相對的心理計量特質，來幫助我們選擇不同的魏氏成人智力量表第三版的簡短式測驗組型。

Factor-based WAIS-III tetrads were investigated with the Taiwan WAIS-III standardization sample of 888 normal adults, ages 16 to 84 years old. Various psychometric characteristics, time constraints, as well as qualities of estimation among prorating and linear equating procedures, were compared among 54 tetrads. Our results supported the use of the linear equating procedure. The Similarities- Matrix Reasoning- Arithmetic- Digit Symbol form exceeded others with respect to validity, content representation, and time saving. Kaufman's (1990) Similarities-Picture Completion- Arithmetic- Digit Symbol combination, and the Information- Picture Completion- Arithmetic- Digit Symbol short forms were both among the most efficient estimations. For clinicians who highly value the clinical information of Block Design, two other forms, Similarities- Block Design- Arithmetic- Digit Symbol and Information- Block Design- Arithmetic-Digit Symbol, were found providing better estimation quality. Nonetheless, current findings revealed that even preferred tetrads had substantial misclassification rates, and these factor- based tetrads tend to show under-estimation for people with top abilities. Users are cautioned to use these short forms for screening purposes only. Results of this study are consistent with literature and assist in choosing among WAIS- III factor- based tetrads by providing their relative psychometric quality.