執行功能平板測驗模組的開發

Development of a Tablet-Based Task Battery for Executive Function Assessment

本研究的目的為開發一套以理論為基礎且免費的執行功能平板測驗模組。根據理論架構，執行功能可分為三個階層：初階包含工作記憶與抑制，中階為認知彈性，高階為推理、規劃及問題解決。其中，高階執行控制依賴初階與中階執行控制以有效運作，因此此模組主要測量初階與中階執行控制能力，並納入處理速度的評估。模組共有17個不同的作業，分別歸類於六個類別：處理速度、訊息維持、更新、抑制、轉換以及資源分配。處理速度以簡單與選擇反應時間表現評估。訊息維持的評估包含四類型作業：特徵屬性、特徵配對、延遲配對以及延遲反應。特徵屬性維持作業包含三類型表徵：空間、符號及物體，由後二者評估長期記憶涉入的效能。特徵配對的維持廣度則以符號或物體與空間位置的連結配對測量，亦可評估長期記憶涉入的效能。延遲配對樣本作業則測量延遲時間如何影響參與者維持無法命名顏色－－圖形的整合訊息，並可由表現錯誤釐清錯誤來自顏色或圖形的混淆。延遲反應作業要求參與者登錄一或四個色彩幾何圖形，並在保留時距後針對一或四個刺激作判斷。抑制功能則以色字字義判斷、逆向眼動以及停止訊號作業評估對解決顏色與字義衝突、抑制強勢眼動傾向以及抑制已預備好的反應的能力。更新功能以流動記憶作業測量參與者更新空間、符號或物體訊息的能力。轉換功能以路徑描繪作業、心花作業以及圖形轉換作業分別評估在顏色－－數字、作業規則以及顏色－－形狀轉換的能力。資源分配功能則以旋轉廣度作業測量參與者在判斷旋轉數字與箭頭方位一長短二者間有效分配認知資源的效能。模組裡有10個作業以正確率與反應時間評估能力，有7個作業以廣度評估。當以廣度為指標時，本模組以一上一下的方式調整作業難度。研究參與者共208位大學生，參與二至三次實驗，每次測量時間約50分鐘。結果均與文獻所得的現象相同，支持平板測量以及用一上一下調整法測量廣度的可行性。由於目前版本的限制，此平板模組有其適合使用的脈絡。未來研究可考量以限時內作業正確率為指標，並納入語言能力的評估與更多元的測量作業。更重要的是系統性的評估模組的信度與效度，並建立不同年齡的常模。

The aim of this study was to develop a theory-driven, tablet-based battery of executive function tasks that is free of charge. According to a theoretical framework, three levels of executive function can be identified. The basic level consists of working memory and inhibitory control, the second level focuses on cognitive flexibility, and the third level includes reasoning, planning, and problem solving. Given that the functions at the third level require effective processing at the first two levels, this battery measured individual abilities on executing the latter functions. Additional functions included executive controls that have been suggested to be important. The battery consisted of 17 tasks assorted into six categories: processing speed, information maintenance, inhibition, updating, switching and resource allocation. Processing speed was measured by the speed in simple and choice reaction times. Information maintenance included maintenance of single items and paired associates. Three types of representations were used for single items: spatial information, symbols that are difficult to name, and objects that can be interfaced with the representations in long-term memory. A delayed matching-to-sample task and a delayed response task were also adopted to assess the information maintenance ability. The former task evaluated maintenance over 5 and 10 seconds compared with perceptual matching; the latter examined the effects of encoding and retrieving one or four items. Inhibition functions were assessed by three tasks: the reverse Stroop task, stop signal task, and anti-saccade task. Updating functions were also evaluated by three tasks: running memory of spatial information, running memory of symbols, and running memory of objects. The trail-making test, a figure task, and a hearts-and-flowers task were used to measure the ability to switch for cognitive flexibility. These three tasks differ in the dimensions for switching: between color and digits, between task rules, and between color and shapes. A rotation span task was adopted to assess the ability to allocate cognitive resources. In the battery, 10 tasks adopted accuracy and reaction time as the measures, and 7 adopted a 1-up-1-down staircase method to measure the span of information maintenance, updating, and resource allocation. Two hundred and eight college students participated in the study across two or three sessions of approximately 50 minutes each, with a rest between tasks. The pattern of results was in agreement with the findings in the literature, supporting the use of a tablet-based battery that employed a staircase method for measuring working memory span. Given the limitations in the current version, a recommendation of the context for using this battery is provided in the discussion. Future studies could consider measuring the abilities of processing linguistic information, evaluating each executive function with a broader range of tasks, and using an accuracy-based battery. More importantly, future studies should examine the reliability and validity of this battery while developing norms for different age levels.