- 學位論文
感覺處理能力剖析量表第二版幼兒版之中文版發展暨心理計量特性驗證
Development and Psychometric Properties of the Chinese Version of the Sensory Profile Second Edition (SP-2-C) Toddler Version
摘要
研究背景 感覺處理功能為大腦接收聽覺、視覺、觸覺、嗅覺，本體覺以及前庭覺等感覺訊息後進行調節、解讀及反應之能力，當個體因異常之感覺處理模式而未能適切接受或處理環境訊息時則為感覺處理障礙，會對孩童之生活參與、情緒行為有負面影響，亟需孩童家長及臨床人員重視。因此，良好之兒童感覺處理能力評估工具可以協助治療師及家長掌握孩童的感覺處理問題進而擬定後續介入計畫。兒童感覺處理能力剖析量表(Sensory Profile, SP)為臺灣常用之感覺處理的評估工具，其適用對象廣泛、全面評估多種感覺處理功能且心理計量特性驗證完整。相較其他工具，SP可更全面且準確評量孩童之感覺處理功能。然而，SP存在一些限制影響其適用性。因此，兒童感覺處理能力剖析量表(Sensory Profile-2, SP-2)改良SP之限制以提升適用性，包括統一SP三個年齡層之內容格式、擴大適用對象之年齡、大幅降低問卷題數以及更新題目內容，新增更多不同情境的題目。然而，目前尚無SP-2中文版，因此本研究將跨文化翻譯SP-2，並驗證其心理計量特性，以確認SP-2-C於臺灣孩童之適用性。 研究目的 考量研究人力及資源限制，本計畫目的有二：(1) 跨文化翻譯SP-2幼兒版，(2) 驗證SP-2-C幼兒版之心理計量特性，包括信度（內在一致性、再測信度）、效度（區辦效度、建構效度）。 方法 本研究方法分為跨文化翻譯及心理計量特性驗證兩大部分。（一）跨文化翻譯：研究小組取得SP-2出版社之授權，依照Beaton跨文化翻譯步驟， 進行前向和反向翻譯SP-2幼兒版，並審視修訂中文版，以符合語義等效性和文化相關性。最後完成修訂的初步中文版，進行預試。預試採方便取樣，透過問卷及半結構訪談評估SP-2-C幼兒版之可讀性，了解家長對問卷的理解程度及收集有關問卷的建議回饋。（二）心理計量特性驗證有三步驟：(1) 招募受試者：於網路發送招募單與研究訊息在家長群組。有興趣的家長掃描QR Code進入Google表單填寫。(2)資料收集：家長填寫SP-2-C幼兒版問卷一至兩週進行再次評估，以email方式再次寄問卷填寫連結，請家長再次填寫SP-2-C幼兒版問卷。 (3)資料分析：信度方面使用Intra-class correlation coefficient (ICC)分析再測信度、Cronbachα驗證內在一致性。效度方面以探索性因素分析(EFA, Exploratory Factor Analysis)及驗證性因素分析(CFA, Confirmatory Factor Analysis)驗證其建構效度。區辨效度則以Multivariate Analysis of Variance(MANOVA)或Multivariate Analysis of Covariance(MANCOVA)驗證。 結果 本研究完成跨文化翻譯及初步中文版預試後，我們增加了具體例子、調整較難理解字眼，以提升問卷的題目可讀性。心理計量特性驗證方面，本研究共收集182位孩童樣本，其中包含「典型發展組」152位和「發展障礙組」27位，並有54位孩童填寫再測問卷。結果顯示，臺灣孩童在「感覺尋求」、「動度」向度上分數顯著較低，而在「感覺迴避」、「感覺敏感」、「整體」、「聽覺」、及「行為」皆得分明顯高於美國孩童。 。信度方面，內在一致性之Cronbach’s α值範圍為0.681至0.845，再測信度的ICC範圍為0.567至0.717，而整體、動作及口腔感覺向度再測信度較低，ICC值分別為0.567、569及0.584。效度方面，發展障礙組在感覺尋求、視覺和動作向度中分數比典型發展組更高，於低登入量及聽覺向度則比典型發展組分數更高。在建構效度部分，顯示與原版SP-2一樣有四個向度： 感覺敏感、感覺迴避、低登入量及感覺尋求，依照EFA的結果，我們經討論後增減每一個因素題目，修改後之模型契合度良好，四個向度的Comparative Fit Index(CFI)範圍為0.937至1.000， Root Mean Square Error of Approximation(RMSEA)範圍為0.010至0.085， Standardized Root Mean Square Residual(SRMR)範圍為0.0252至0.0578。 結論 本研究進行了SP-2-C幼兒版跨文化翻譯以及心理計量特性驗證。信度方面，SP-2-C具有高內在一致性，但在動作、口腔感覺的再測信度較低，可能與幼兒快速發展的特性有關，建議未來研究縮短評估間隔。效度方面，SP-2-C幼兒版能有效區分典型發展兒童和發展障礙兒童的感覺處理功能。此外，SP-2-C幼兒版之建構效度良好，但題目架構與原版略有不同，考量台美文化、情境差異，建議臨床人員使用本研究之架構進行分數解釋。此外，臺灣的典型發展孩童在感覺處理模式量表的感覺尋求、感覺迴避、低登入量和感覺敏感向度和感覺系統量表的整體、聽覺、動作、行為四個向度上與美國孩童存在顯著差異。因此，本研究建議建立臺灣常模以確保評估結果的準確性，以及治療師在臨床應用時可進一步向家長確認題意以及了解孩童情緒行為的情境及原因。儘管存在樣本數不足和孩童診斷標準較主觀等限制，SP-2-C幼兒版可穩定且有效的評估臺灣幼兒之感覺處理功能。未來研究可擴展至更多診斷類型及年齡層，並建立本土化常模，以擴大SP-2-C之適用性。
並列摘要
Abstract Background: Sensory processing refers to the body's ability to receive and transmit environmental information to the brain for response and interaction. Sensory processing disorders may affect a child's emotions and behaviors, which further affect a child development and participation in daily life. To assist therapists and parents in identifying sensory processing issues and developing intervention plans, there is a need for an assessment tool to evaluate a child's sensory processing abilities. Sensory Profile (SP) is a currently available assessment tool in Taiwan for evaluating sensory processing. However, the SP has limitations affecting its applicability. To address these issues, an updated version of the SP (SP-2) was developed. SP-2 integrated the same scale design and score interpretation for each age group, reduced the number of questionnaire items for most age groups, and added more items in different contexts to enhance the applicability of the questionnaire. Therefore, the SP-2 can comprehensively and accurately assess children's sensory processing abilities and changes across various developmental stages. Consequently, it is crucial to validate the readability and psychometric properties of the Chinese version of SP-2. Aims to provide a comprehensive and accurate assessment tool for sensory processing functions to clinicians and parents. Objectives: Considering constraints on manpower and resources, this study aims to (1) pretesting for readability of the Chinese version of Toddler SP-2 (SP-2-C toddler version) and (2) validate the psychometric properties of the SP-2-C toddler version , including reliability(internal consistency and test-retest reliability) and validity(discriminant validity and construct validity). Methods: This study included two phases. Phase 1 is the cross-cultural adaptation of the SP-2. SP-s was cross-culturally adapted into Mandarin Chinese following the procedure outlined by Beaton et al., (2000). Then, using convenience sampling, the readability of SP-2-C Toddler Version was assessed through questionnaires and semi-structured interviews. This aimed to assess parent's comprehension of the questionnaire and collect their feedback of the questionnaire. Phase 2 is the validation of psychometric properties of the SP-2 Chinese version, consisting of three steps: (a) recruitment of participants: Recruitment flyers were posted in the social media groups of parents on the internet. The flyer contains study information and a QR code. Parents scan the QR code to access and complete the questionnaire on a Google form. (b) Data collection: Parents who completed the SP-2 questionnaire were invited to complete a second assessment via email one week later. (c) Data analysis: Reliability were examined using the Intra-class correlation coefficient (ICC) for test-retest reliability, and Cronbach's alpha for internal consistency. Construct validity were examined using confirmatory factor analysis (CFA), and discriminative validity were examined through MANOVA or MANCOVA. Result: After cross-cultural adaptation, we enhanced the overall readability of the questionnaire by adding specific examples and adjusting some wording and terms. However, we identified that some vocabulary and emotion-related items were still difficult for parents to assess. In psychometric validation, we recruited 182 children, including 152 typically developing children and 27 with developmental disabilities. Fifty-four children completed the retest questionnaire. Results showed significant differences between Taiwanese children in the typical development group and their American norms across three dimensions of the Sensory Processing Patterns Scale and four dimensions of the Sensory Systems Scale. Regarding reliability, internal consistency (n=182) ramged from acceptable to good (Cronbach's α = 0.681 to 0.845), and test-retest reliability (n=54) ranged from moderate to good (ICC = 0.558 to 0.716). Regarding discriminative validity, significant differences were observed between two group in average scores across dimensions of sensory seeking, low registration, auditory, visual, and movement. For construct validity, the SP-2-C toddler version showed a good fit where the hypothesized models were adjusted by adding and deleting items: CFI ranged from 0.937 to 1.000, RMSEA ranged from 0.010 to 0.085, and SRMR ranged from 0.0252 to 0.0578. Conclusion: In this study, we verified the reliability and validity of the SP-2-C Toddler Version. In terms of reliability, the SP-2-C toddler version demonstrated acceptable to good internal consistency and moderate to good test-retest reliability, Alhough test-retest reliability for movement and oral sensory areas was lower, it is possibly due to rapid developmental changes in toddlers; hence, future studies might consider shortening re-assessment intervals. Regarding validity, SP-2-C Toddler Version effectively discriminates sensory processing functions between two groups. Construct validity is good, yet considering differences between Taiwanese and American children, clinicians are advised to interpret scores by using our framework. Additionally, compared to American children, Taiwanese children show significant differences across multiple sensory processing dimensions, suggesting the establishment of Taiwanese norms. Despite limitations such as insufficient sample size and subjective diagnostic standards for children, SP-2-C Toddler Version proves to be a stable and effective sensory processing assessment tool for Taiwanese toddlers. Future research should expand to include more diagnostic types and age groups, establish local norms.