過去研究指出焦慮症及憂鬱症的自陳式或臨床評分的測量方式雖有幅合效度,但區別效度卻不高。Clark和 Watson (1991)提出焦慮症和憂鬱症的三角模式(tripartite model of anxiety and depression),企圖釐清焦慮症及憂鬱症間的關係。本研究以憂鬱症、焦慮症的門診病人及社區正常人組共233名為研究對象,依據三角模式編寫正向情感、負向情感、身體激起三個向度的量表,以改善既存量表的缺點,期望能籍由兩種心理疾症的清楚區別與早期發現,以便為病患提供臨床上適切的療育。同時,對常人而言亦能提供其有關情感及生理表現的資料,增進自我的了解。我們以Rasch模式及其延伸模式來分析資料,篩選題目,將原始分數的順序資料轉化為具有等距特性的量尺分數。結果顯示出正向情感、負向情感、身體激起量表具有頗高的受試者信度(.92~.94)及再測信度(.85~.89),而各量表間的相關亦與過去研究發現一致。在三角模式的檢驗方面,比較三個組別在三個分量表的平均得分上之差異,所得結果與三角模式符合。此外,區辨分析的檢驗亦說明了正向情感及身體激起兩個變項能提升區別兩組病患的區辨力。
Previous studies have indicated that traditional self-report measures of anxiety and depression have good convergent validity but poor discriminant validity. Clark and Watson (1991) proposed a tripartite model that helps explain the relationship between anxiety and depression. The primary goal of this study was to develop a new instrument based on the tripartite model, which consists of three scales: positive affect (PA), negative affect (NA), and physiological hyperarousal (PH). The self-report instrument can be used to help distinguish anxiety and depression patients, and help evaluate affection and physiological states of general persons. The final version of the scales was administered to 45 anxiety patients, 46 depression patients, and 142 normal adults. Rasch analysis (Rasch, 1960) was used to analyze test data, screen poor items, and establish interval scales. The resulting PA, NA and PH scales demonstrated good person reliability (.92~.94) and test-retest reliability (.85~.89). The relationship among the three scales was consistent with previous findings. In the test of the tripartite model, we compared the mean scores on the three scales across groups. The results were consistent with the model. Discriminant analysis also suggested that strengthening the PA and PH components in the scales may enhance their discriminative power.