平板電腦版符號數字轉換測驗應用於中風患者之心理計量特性

背景與研究目的：為針對國內正自行發展的電腦化認知評估工具(平板電腦版符號數字轉換測驗,簡稱T-SDMT)建立其心理計量特性，以使該評估工具能予臨床工作者更準確的資訊及提供更佳的療效依據，故本研究目的有五：(一)檢測T-SDMT單回施測中的再測信度及最佳施測次數。(二)檢測T-SDMT和魏氏成人智力量表第四版中文版的分測驗「符號替代測驗」(簡稱DSST) 於二週後的再測信度並比較其優劣。(三)檢測T-SDMT和DSST的同時效度。(四)檢測T-SDMT的收斂效度及區辨效度(和蒙特利爾認知評估MoCA做相關係數驗證)。 (五)檢測T-SDMT的生態效度(和日常生活活動電腦化適性測驗ADL CAT做相關係數驗證)。方法：徵求30位中風六個月以上之患者，施以前測及後測(間隔時間為二週)，每回都接受T-SDMT和DSST一次的練習測驗及四次的正式測驗，並於前測時加入ADL CAT，而後測時加入MoCA以進行相關效標的檢測。結果：T-SDMT單回多次測量皆呈現良好的再測信度且合宜的施測次數為二次。T-SDMT與DSST具良好的同時效度，及在間隔二週皆具良好的再測信度，且T-SDMT略優於DSST。T-SDMT與MoCA具良好的收斂效度及具有區辨效度。而在描述性統計上僅教育程度有統計上的顯著差異。另T-SDMT與ADL CAT呈不佳的生態效度。結論：T-SDMT單回施測以二次為合宜的施測次數，並於二週後的再測信度上略優於DSST，臨床效能上T-SDMT亦有潛力成為具篩檢出輕度認知障礙的評估工具。

Background and purpose: To establish psychometric characteristics for a computerized cognitive assessment tool (the Tablet-based Symbol Digit Modalities Test, T-SDMT) that is being developed in domestic so that the assessment tool can provide clinicians with more accurate information and better efficacy basis. The purposes of this study were five: (1) examine the test-retest reliability and best trial of tests for the T-SDMT in one session. (2) Comparison of test-retest agreement of the Wechsler Adult Intelligence Scale IV Chinese Version of the subtest “Digit Symbol Substitution Test”(DSST) and T-SDMT after two weeks. (3) examine the concurrent validity of T-SDMT and DSST. (4) examine the convergence validity and discriminative validity of T-SDMT (Correlation coefficient verification with Montreal Cognitive Assessment (MoCA)).(5) examine the ecological validity of T-SDMT (verification of correlation coefficient with Computerized Adaptive Test for Assessing Activities of Daily Living ,ADL CAT).Method: Thirty patients with stroke of six months or more were recruited for pre-test and post-test (interval time was two weeks). Each time they received T-SDMT and DSST training tests and four formal tests. ADL CAT were added during the pre-test, and MoCA was added for post-test testing.Results: Appropriate testing times in one session for T-SDMT is twice. Multiple measurements in one session for T-SDMT show good test-retest reliability. T-SDMT and DSST have good concurrent validity. T-SDMT and DSST have good test-retest reliability in two weeks apart, and T-SDMT is slightly better than DSST. T-SDMT and MoCA have good convergence validity and discriminative validity. In terms of descriptive statistics, only statistically significant differences in the education level were observed. Another T-SDMT and ADL CAT showed poor ecological validity.Conclusion: Appropriate testing times in one session for T-SDMT is twice.T-SDMT was slightly better than DSST in test-retest reliability after two weeks. T-SDMT has the potential to be an assessment tool for the screening of mild cognitive impairment on clinical efficacy.