「身體功能測試」(Physical Performance Test, PPT)是一項適合老人的身體功能測試工具。本研究除了將原始7項PPT翻成中文版外，也鑑於台灣長者教育程度偏低，而將其中易受影響的分項目「寫句子」修訂成「複製圖」，形成改良版PPT。故研究目的：(1) 比較中文版與改良版PPT之使用可行性，和(2) 驗證兩版本身體功能測試的內在一致性和同時、收斂與區辨效度。本研究共招募30名健康老人及80名失智者，每位均接受中文版和改良版PPT測試、蒙特利爾認知評估(Montreal Cognitive Assessment, MoCA)，和失智症功能障礙評估量表(Disability Assessment for Dementia, DAD)。各分項可行性，除寫句子僅達61%，在複製圖可達88%，其他分項也達79-98%；且經複迴歸分析發現，只有「寫句子」分項表現與教育年數有關(β= .461, p< .001)；兩版PPT整體具良好的內部一致性(Cronbach’s α= .90 及.91)；兩版PPT總分達高度顯著相關(r= .99, p< .001)，具良好的同時效度；兩版PPT總分，與MoCA、DAD總分和DAD分項分數呈現介於0.62至0.70之間(p< .001)的顯著相關，具好的收斂效度。且兩版PPT總分區辨正常人和失智者的接收者操作特徵曲線下方之面積為0.97和0.98 (p< .001)，具良好的區辨效度。故兩版本PPT應用於台灣社區老年人具良好之信效度，然而改良版PPT的可行性優於中文版PPT，特別適合台灣老年族群的需求。

The Physical Performance Test (PPT) is a tool for identifying physical change in elderly. In addition to translating the original PPT into Chinese, we further revised the "write a sentence" subtest into the "copy the pictures" to form the modified PPT (mPPT) version for use in Taiwan society with a generally low education level. The purpose of this study was to: (1) compare the feasibility of using Chinese PPT and mPPT, and (2) examine internal consistency, concurrent validity, convergent validity, and discriminant validity for the elderly in Taiwan. A total of 30 healthy elders and 80 patients with dementia were recruited. The Chinese PPT, mPPT, Montreal Cognitive Assessment (MoCA), and Disability Assessment for Dementia (DAD) were administered. The feasibility was above 80% for most sub-items (79%-98%) except for the sentence-writing sub-item (61%). The results of multiple regression analysis showed that only the sentence-writing sub-item was related to education years (β= .461, p< .001). The correlation between the Chinese PPT and mPPT scores was very high and significannt (r= .99; p< .001). The correlations between the two versions of PPT and MoCA, total DAD, or DAD subscores were moderate and significant, ranging from 0.62 to 0.70 (p< .001). The area under the relative operating characteristic curve of the total Chinese PPT and mPPT scores were 0.97 and 0.98 (p< .001) to differentiate between healthy controls and dementia patients. Both the Chinese PPT and mPPT have good internal consistency, concurrent, convergent, and discriminant validity. However, mPPT is more feasible than the Chinese version of PPT, suggesting that the mPPT is especially useful to the elderly with a low education level in Taiwan.