作為認知測驗的手機圖形鎖： 彩色路徑解鎖測驗的開發

Abstract

輕度認知障礙（Mild Cognitive Impairment, MCI）的核心特徵是認知功能下降，若能及早發現，便能提供干預機會。目前主流評估方式為臨床診斷時使用紙本認知測驗，但紙本測驗存在諸多限制，如須主動進行測驗、受限於時空。為突破限制，數位表現型曾被應用於檢測MCI，能於個體未察覺時被動蒐集評估資訊，但仍有其研究面與實務面上的限制，如因果關係未釐清、侵犯隱私。此外，認知功能下降還可能隨情境短暫發生，具有潛在風險，如疲勞駕駛易導致交通事故。然而，無論是紙本測驗或數位表現型都無法有效捕捉此情形。因此，本研究開發了一種新的認知測驗：「彩色路徑解鎖測驗」（Color Trails Lock Test, CTLT），除了能突破過去方法的限制，亦能用經驗取樣法捕捉短暫的認知功能下降，並能透過測驗的設計方式降低練習效果。CTLT為彩色路徑描繪測驗（Color Trails Test, CTT）的變體，結合手機的圖形解鎖功能並數位化，具主動回報卻為被動偵測的特性。本研究亦對CTLT進行信效度檢驗，對認知功能正常的個體施測。在效度上，CTLT與CTT呈現顯著正相關（ρ = 0.63），二者與其他紙本測驗的相關表現也相似；在信度上，透過經驗取樣法獲得複本再測信度（ρ = .53~.61），信效度都符合研究預期。本研究不僅豐富了認知測驗的工具箱，更提供了認知評估與科技應用融合的思考方式，為認知測驗在實際應用領域帶來新的可能性。

The core characteristic of Mild Cognitive Impairment (MCI) is cognitive decline, which, if detected early, allows for timely intervention. Currently, the predominant assessment method involves paper-based cognitive tests administered during clinical diagnoses. However, these tests have several limitations, such as the need for active participation and restrictions related to time and space. To overcome these limitations, digital phenotyping has been utilized to detect MCI by passively collecting assessment information without the individual's awareness. Nevertheless, it still faces research and practical limitations, such as unclear causality and privacy concerns. Additionally, cognitive decline can occur transiently in certain situations, posing potential risks; for example, fatigue driving can easily lead to traffic accidents. Neither paper-based tests nor digital phenotyping can effectively capture this phenomenon. Therefore, this study developed a new cognitive test: the "Color Trails Lock Test" (CTLT). This test not only overcomes the limitations of previous methods but also uses the Experience Sampling Method to capture transient cognitive decline, and its design reduces practice effects. The CTLT is a variant of the Color Trails Test (CTT), incorporating the graphical unlocking feature of smartphones and digitizing the test. It has the characteristics of active reporting yet passive detection. This study also examined the reliability and validity of the CTLT by administering it to individuals with normal cognitive function. In terms of validity, the CTLT showed a significant positive correlation (ρ = 0.63) with the CTT, and both exhibited similar performance correlations with other paper-based tests. In terms of reliability, the alternate test-retest reliability (ρ = .53~.61), with both reliability and validity meeting research expectations. This study not only enriches the cognitive assessment toolkit but also provides a new perspective on the integration of cognitive assessment and technological applications, bringing new possibilities to the practical application field of cognitive testing.