二種視知覺測驗應用於腦性麻痺兒童之信度及收斂效度

Abstract

背景與目的：視知覺障礙是腦性麻痺(cerebral palsy, CP)兒童除動作之外，常見的問題。雖然研究或臨床上，視知覺測驗(Motor-Free Visual Perception Test-Revised, MVPT-R)和視知覺技巧測驗(Test of Visual-Perceptual Skills-Revised, TVPS-R)常用於評估CP兒童的視知覺功能，然而目前尚無研究證實MVPT-R和TVPS-R運用在CP兒童時之心理計量特性，因此無法確知這二種測驗是否適用於CP兒童視知覺功能之評量。故本研究欲探討MVPT-R和TVPS-R運用在CP兒童時之信度（包括施測者內在信度、施測者間信度和內在一致性）及收斂效度。 方法：從醫院復健科及學校系統篩選出符合研究標準之CP兒童，結果共收集到52位，年齡介於五歲五個月至八歲九個月。在信度驗證部分，初評由一位 職能治療師執行，約7天後將個案隨機分配至兩組，一組由同一治療師(施測者內在信度)，另一組則由另一位治療師(施測者間信度)進行再評。收斂效度驗證方面，部分個案侷限於上肢之功能，只有47位受試者完成測驗，評量時間約在信度驗證初評的前7天進行。選用兩種視覺-動作整合測驗和四種視覺-動作整合測驗的補充測驗(視知覺及動作協調補充測驗)，共包含六種測驗、三種不同建構，作為檢驗MVPT-R和TVPS-R的收斂效度之校標。再以MVPT-R和TVPS-R的全量表及分量表間之關連性，驗證兩視知覺測驗之收斂效度。MVPT-R和TVPS-R的全量表和分量表之再測信度驗證，以六種統計指標，包括組內相關係數（intraclass correlation coefficient, ICC）、Bland & Altman圖形分析、測量標準誤（standard error of measurement, SEM & SEM%）和個案最小真實改善程度值(smallest real difference, SRD & SRD%)分析。 結果：(1) 信度方面: MVPT-R和TVPS-R總分的ICC (95% CI)值為0.92-0.97 (.82 - .99)，表示同一施測者和不同施測者使用全量表的MVPT-R和TVPS-R時，皆具有高度的一致性。全量表的MVPT-R和TVPS-R之SEM%值低(5.3%-11.3%)，顯示測量誤差大致數值小。全量表的MVPT-R和TVPS-R之SRD%則介於14.7%至31.3%。部分TVPS-R的分量表、多數TVPS-R和MVPT-R的項目，施測者內在信度及施測者間信度呈現一致性不高之情形。TVPS-R和MVPT-R所有項目的alpha值介於0.86至0.98，顯示項目之間具備高度內在一致性。(2) 收斂效度方面：MVPT-R和TVPS-R和視知覺補充測驗具有高度至中度的相關性(Spearman rho= .60 to .91)，和視覺-動作整合測驗及動作協調補充測驗則是中度 相關(rho= .40 to .67)。全量表和分量表，以及不同分量表間之關連性分析，也顯示相同測量概念之分量表間具有中度相關(rho= .73 to .78)。 結論：本研究發現全量表的MVPT-R和TVPS-R具有良好的信度，但此結果不適用在TVPS-R的分量表，以及MVPT-R和TVPS-R的個別項目。從關連性驗證結果，MVPT-R和TVPS-R具備收斂效度。本研究結果顯示：全量表的MVPT-R和TVPS-R適用於CP兒童視知覺功能之評量，但不適合使用分量表及個別項目評量CP兒童。 關鍵字：腦性麻痺，視知覺，信度，效度

Background and Purposes: The Motor-Free Visual Perception Test-Revised (MVPT-R) and the Test of Visual-Perceptual Skills-Revised (TVPS-R) have been widely administered to assess the visual-perceptual function of children with cerebral palsy (CP). However, the psychometric properties of both measures remain unknown in children with CP, limiting the application and interpretation of the MVPT-R and TVPS-R in both clinical and research settings. The purposes of this study were to examine the inter-rater reliability, intra-rater reliability, internal consistency, and convergent validity of the MVPT-R/TVPS-R in children with CP. Methods: A convenience sample of 52 children with CP was recruited from rehabilitation units or schools. The MVPT-R and the TVPS-R were administered on 52 participants by an occupational therapist (OT). Seven days later, 52 participants were randomly assigned to the same OT or another OT. About seven days before the reliability investigation, brief visual perception tests, visual-motor tests, and motor tests were administered on 47 participants to determine the convergent validity. In addition, the correlations between the total scores and the subscale scores of the MVPT-R/TVPS-R were used in a convergent validity investigation. Relative and absolute reliability indices were used to analyze the rater reliability, including the intraclass correlation coefficient, Bland-Altman method, standard error of measurement (SEM, SEM%), and smallest real difference (SRD, SRD%). Results: The intra-rater and inter-rater agreements of total scores of the MVPT-R/TVPS-R were excellent, with ICCs ranging from .92 to .97. The SEM% was small (<10%) for the intra-rater reliability of the MVPT-R and total scores of the TVPS-R. The SEM% was also small for the inter-rater reliability of the MVPT-R. The SRD% ranged from 16% to 31% for total scores of the MVPT-R/TVPS-R. However, the rater reliability were not consistent or satisfactory for some subscales and most items. The internal consistency was good for the MVPT-R/TVPS-R (alpha= .87 to .98). The total scores of the MVPT-R/TVPS-R had high to moderate correlation with other visual perceptual tests or subscale scores of the MVPT-R/TVPS-R (rho= .60 to .91) and moderate correlation with visual-motor and motor tests (r=.40 to .67), roughly supporting the convergent validity of both measures. Conclusion: The total scores of the MVPT-R and the total scores of the TVPS-R were reliable and internally consistent in children with CP. However, the intra-rater and inter-rater reliability were not consistent and satisfactory for some subscales and most items. The convergent validity of the MVPT-R/TVPS-R was roughly supported. The use of total scores of the MVPT-R/TVPS-R in research and clinics was recommended, but not the use of the subscale scores or item scores in children with CP. Key Words: Cerebral Palsy, Visual Perception, Reliability, Validity