- 學位論文
台大數字奇偶序列測驗:台灣北部區域常模與信度效度研究
The National Taiwan University Odd-Even Number Sequencing Test (TOENS): North Taiwan Normative Data and Reliability and Validity Study.
摘要
數－字序列測驗是WAIS-III中一項主要的分測驗，它是用來評估工作記憶。然而，由於語文上的障礙，即台灣人對英文字母不熟悉，導致WAIS-III中文版該分測驗的實用性有限。因此，本研究企圖發展出一項可比擬數－字序列分測驗的新測驗，不但具備心理計量特性，對不熟悉英文字母者又能夠去除語言隔閡，並將之運用在臨床情境中。『台大數字奇偶序列測驗』的設計就是取代掉英文字母，利用數字的奇數和偶數序列來組成每個項目。本研究目的主要是檢驗台大數字奇偶序列測驗的心理計量特性，並進一步探討台灣北部地區正常成人之人口學變項與測驗的關係。 本研究參考WAIS-III中文版建構常模之程序，共收集台灣北部地區300位16-84歲健康成人的測驗資料，並將之區分成11個年齡組。所有人均接受台大數字奇偶序列測驗、認知功能及情緒狀態篩檢，其中30位受試者於三個月後進行重測，其中又42位受試者同時接受WAIS-III的記憶廣度測驗和數－字序列測驗、WMS-III的空間廣度測驗、修改式威斯康辛卡片分類測驗、及語意流暢度測驗來檢驗台大數字奇偶序列測驗之信效度。 信度結果顯示各年齡組之折半信度介於.69~.95，相隔三個月之重測信度值為.75；校標關聯效度方面，以記憶廣度測驗、空間廣度測驗、及數－字序列測驗等傳統被視為工作記憶評估工具作為效標，各測驗均與本測驗總分有顯著相關；建構效度方面，本研究採用的方法是探討工作記憶的神經基礎與工作記憶應呈現年齡相關的減退趨勢，前者是由與修改式威斯康辛卡片分類測驗的相關來間接證實台大數字奇偶序列測驗其神經機制與背外側前額葉有關，後者則是從老年人的測驗表現顯著低於年輕人得到驗證。本研究提供不同年齡組的量尺分數常模，目前資料顯示本測驗較不適用於低教育程度之老年個案。人口學變項中，有顯著的年齡與教育程度效果，而性別則沒有顯著差異。 台大數字奇偶序列測驗已證實是一項具有良好信效度的聽覺工作記憶評估工具；然而，正常老年人的常模資料，特別是低教育程度者，未來須進行更多研究來加以修正。另外，台大數字奇偶序列測驗呈現顯著的年齡和教育程度效果，接續的常模研究應同時建立年齡和教育程度常模。
並列摘要
The Letter Number Sequencing Test (LNS) has been included in the WAIS-III to be one of the main subtests measuring working memory function. However, the practical utility of the subtest in the Taiwan version of the WAIS-III has been limited because of its language barrier, the unfamiliarity of letters, for Taiwanese. The present study was thus to make an attempt to develop a test compatible with the LNS in terms of psychometric properties and free of the language boundary for people who are unfamiliar with letters but may take it in clinical-related settings. The National Taiwan University Odd-Even Number Sequencing Test (TOENS) whose items are composed of odd and even number serial instead of letters was subsequently formed. The specific aims of the study were to examine its psychometric properties and further to explore its relations to demographic variables of normal adult in north Taiwan. Based on the standardized procedure of the Taiwan version of the WAIS-III, a total of 300 normal participants with age range from 20 to 84 years old was recruited and further divided into 11 age groups. All participants received the TOENS and screening tests measuring general mental function and emotional status. Thirty of these 300 subjects were retested 3 months later, and 42 also received the Digit Span and LNS subtests of the WAIS-III, the Spatial Span subtest of the WMS-III, the modified WCST, and the Semantic Association of Verbal Fluency for examining reliability and validity features of the TOENS respectively. The results revealed the split-half reliability coefficients of the TOENS ranging from .69 to .95 for the 11 age groups respectively; and the test-retest reliability coefficient is .75. The concurrent validity was demonstrated with significant correlations with the conventional working memory measures, including the Digit and Spatial Span and the LNS subtests. The construct validity was explored and confirmed in terms of the neural underpinnings of working memory, the dorsolateral prefrontal cortices, in which their functioning could be reflected by the level of performance on the neuropsychological tests, such as the modified WCST, as well as of a remarkable aging-related decrement of working memory test performance level. The results indicated a significant relationship between the TOENS and the modified WCST, and a notable decline of performance on the test in the elderly compared with their younger counterparts. Normative data were primarily stratified by age groups, with the exception of the data for the elderly with a low educational level. Our participants’ performance level on the TOENS test was significantly associated with age and education variables while there was no gender effect. Based on the present results, it appears that the TOENS bears adequate psychometric properties and clinical utility in measuring auditory working memory. However, the applicability of the normative data in normal elderly individuals, particularly those with a low education level, does evidence confines and awaits further investigation. In addition, in light of effects of age and education on our normal adult individuals’ performance on the TOENS, more study on the normative data with both age and education correction rather than with the age only is merited.