反應性 (responsiveness) 是評估工具偵測受測者功能改變之能力，亦為成效指 標必備的心理計量特性。中風病患姿勢控制評估量表 (Postural Assessment Scale for Stroke Patients, PASS) 及其短版 (Short Form-PASS, SFPASS) 是為中風病人特 製且具良好心理計量特性之平衡功能測驗，具臨床常態使用之潛能。昔日研究曾 比較PASS 與SFPASS之團體與個別層級反應性於中風後14 天且間隔二週的病患。 然而，單一研究無法斷定評估工具之反應性，且PASS 與SFPASS 應用於復健住 院病患的個別層級反應性尚屬未知。因此，本研究目的為：交叉驗證PASS 與 SFPASS 應用於中風復健住院時期之團體與個別層級反應性。 共379 位中風病患參與本研究，其分別於復健住院後及出院前一周評估PASS 共二次；SFPASS 則擷取自PASS。團體層級反應性指標為t 檢定驗證團體層、凱 濟斯效應值 (Kazis’ effect size) 及標準化平均反應值 (standardized response mean, SRM)；個別層級反應性指標則為進步超過最小可偵測差異值 (minimal detectable change, MDC) 之人數比例。 PASS 與SFPASS 於前、後測之平均分數差異顯著 (p < .001)，且效應值相似 （Kazis’ effect size：0.74 與0.70；SRM：1.09 與1.00）。PASS 可呈現較多進步超 過MDC (63.0 %比53.0 %) 之病患，且此差異顯著 (p < .001)。 本研究結果顯示PASS 與SFPASS 皆具良好的團體層級反應性，但SFPASS 之個別層級反應性較PASS 差。我們建議使用者宜採用PASS 做為復健住院病患之 療效評估工具；而SFPASS 則可用以追蹤病患群體之平衡功能以提升施測效率。

Responsiveness is the ability of measures to detect people’s change, which is necessary for outcome measures. The Postural Assessment Scale for Stroke Patients (PASS) and Short-Form PASS (SFPASS) were developed for stroke patients with sound psychometric properties, which have a potential for routine clinical assessments. A previous study has examined the group-level and individual-level responsiveness of the PASS/SFPASS in patients at 14 days after stroke. However, the individual-level responsiveness of the PASS/SFPASS was unclear. Thus, the purpose was cross-validation of the group- and individual-level responsiveness of the PASS and SFPASS in inpatients receiving rehabilitation. A total of 397 participants were recruited. The PASS was assessed within one week after admission and before discharge from the rehabilitation ward, and the SFPASS was derived from PASS. Group-level responsiveness was examined using paired t-test , Kazis’ effect size, and standardized response mean (SRM). Individual-level responsiveness was examined through the proportion of people whose change in balance function was larger than minimal detectable change. Mean differences of the PASS and SFPASS between pre- and post-test were significant (p < .001). Both measures had similar effect size (Kazis’ effect size: 0.74 and 0.70; SRM: 1.09 and 1.00). The PASS is more sensitive at detecting people’s change (63.0 %; 53%), and the difference was significant (p < .001). PASS had better individual-level responsiveness than SFPASS while the group-level responsiveness was similar. We suggest that the PASS is better for detecting the change of balance function for individual, while SFPASS is more efficiency for group change.