台灣正常人修訂版時限聽覺序列加法測驗之常模研究

Abstract

背景：時限聽覺系列加法測驗（PASAT）由於同時需要工作記憶及處理速度的涉入而對腦功能變化有高敏感度、比起其他相關測驗較為困難而適用於高功能的受試者，且能反映部分的壓力容忍度，具生態效度，因此為廣泛使用於注意力、工作記憶及處理速度評估的工具之一。然而現有之PASAT常模樣本代表性不足、鮮少將可能影響測驗表現之智力及數學能力納入考量、且缺乏提供心理計量資料，不利於臨床使用。此外，測驗版本選擇使用修訂版時限聽覺系列加法測驗（PASAT-R），因考量其花費時間短、可避免造成疲勞、並可減低練習效果及天花板效應之優點。目的：本研究主要目的為建立臺灣地區正常人的PASAT-R常模，主要目標為：一、探討發展人口學變項校正常模之必要性及相關變項對測驗表現的影響；二、評估各分數之心理計量特性；三、本常模之適用性；四、短版PASAT-R的適用性。方法：本研究收集臺灣306名16至92歲、教育程度0至18年之正常人於PASAT-R之常模資料，同時收集魏氏成人智力量表第三版、尼爾森修訂版卡片分類測驗、彩色路徑描繪測驗、路徑描繪測驗、史楚普叫色測驗、語意流暢度測驗等資料來檢驗PASAT-R之效度。其中30名受試者於四個月後進行再測以考驗信度。結果：年齡及教育程度為主要預測測驗表現之人口學變項，故有發展人口學變項校正常模之必要，且智力與測驗表現有顯著相關。本測驗具有良好之信度及效度，而本研究提供之常模資料亦具有良好之代表性、新近性及適切性。另外，不同刺激呈現時間間隔的分測驗可反映不同的臨床應用性。結論：本研究以統計方法排除人口學變項對測驗表現的影響，再以校正分數建立整體測驗分數以及單一分測驗分數之百分等級對照表，供臨床工作者於偵測腦功能改變、評估治療及復健效果、並作為預後評估指標。未來應擴大常模樣本數，並收集臨床病人組之測驗表現及神經影像學資料，以累積驗証的效度資料，並訂定適當切截點；另外，亦建議建立正常人於重複施測情境下的練習曲線資料，以利協助判斷認知功能的變化或治療之成效。

Background: The Paced Auditory Serial Addition Test is one of the tests most frequently used to assess attentional processing, working memory, and information processing speed since it is sensitive to brain dysfunctions due to its demands on both working memory and processing speed, is well suited to higher functioning individuals due to its level of difficulty, and has good ecological validity reflected by tolerance to stress. However, the current available norms with questionable representativeness of samples, lack of examination of relations between the test and other related variables, and lack of examination of psychometric properties limit the clinical utility. Moreover, the Paced Auditory Serial Addition Test-Revised (PASAT-R) is the selected version that has advantages over other counterparts, including less time consuming, patient discomfort, and practice and ceiling effects. Objective: The study was aimed to provide a norm for the PASAT-R. The specific objectives were to explore: (1) the necessity of demographic-corrected norms and effects of IQ, (2) the issue of psychometric properties, (3) the issue of norm appropriacy, and (4) the feasibility of short forms. Methods: Three hundreds and six participants were collected through stratified sampling by current age (ranging from 16 to 92 years old), education (ranging from 0 to 18 years), and area of residence. The WAIS-III (Chinese version), the Nelson’s Modified Card Sorting Test, the Color Trials Test, the Trail Making Test, the Stroop Color and Word Test, and the Semantic Association of Verbal Fluency were also administered to examine the validity. The re-test for reliability was derived from 30 participants of the norm groups in four months later. Results: That age and education had influence on each PASAT-R scores highlighted the necessity of demographic-adjusted norms. IQ was significantly correlated with all PASAT-R performances. Relatively high reliability and validity were further confirmed from the study. The normative data had good representativeness, recency, and relevance. In addition, every single individual block trial can reflect distinct clinical utilities. Conclusion: The present study established an appropriate normative data based on multiple regressions and provided the referential table for percentile ranks for the overall PASAT-R score and each single block trial score. It is useful for differential diagnosis, treatment evaluation, and prognosis. Nonetheless, the further work on a large scale to boost the test’s normative collection and validity in various brain-lesioned patients is necessary. Likewise, further investigation on the issue of the test’s practice effect that is often associated with repeated testing for monitoring medical treatment repercussions on mental function is also mandatory.