兒童社交勝任能力量表之發展及其於自閉症兒童之心理計量特性驗證

Abstract

社交勝任能力 (social competence) 是指個體能以符合社會情境的方式理解並回應社交線索的能力。自閉症兒童在日常生活中的社交勝任能力經常面臨挑戰，但現有由照顧者填寫之問卷量表普遍存在所測構念與其理論基礎一致性不足、建構涵蓋範圍不足，以及心理計量證據有限等問題。為回應上述限制，本研究發展由照顧者填寫之兒童社交勝任能力問卷 (Children’s Social Competence Questionnaire, CSCQ)，並驗證其於自閉症兒童之心理計量特性。CSCQ以多面向社交勝任能力架構為基礎，涵蓋社交技巧元素 (social skill elements) 、社交交互性 (social reciprocity) 、社交調適 (social adjustment) 與社交有效性 (social effectiveness) 等四個面向，以評估兒童在日常生活中的社交勝任能力。 本研究分為兩個階段。第一階段為發展CSCQ，程序包括參考量表的蒐集與分析、題項建構、專家內容效度評估及照顧者預試；第二階段為檢驗CSCQ之心理計量特性，包含內部一致性、再測信度、收斂效度、區辨效度及反應性。參與第二階段之兒童接受研究人員施測之評估，照顧者完成問卷量表。 第一階段結果支持CSCQ之內容效度，並根據專家建議修訂題項。第二階段樣本包含110位3至12歲自閉症兒童（平均年齡為76.9個月，標準差為27.2；男童占79.1%）及其照顧者，另納入15位一般發展兒童（平均年齡為82.9個月，標準差為20.6；男童占53.3%）作為比較樣本。研究結果顯示，CSCQ總分與各面向分數均呈現良好的分布特性。CSCQ總量表及各面向具有優異的內部一致性 (Cronbach’s α = .91–.98) ，並在2至3週的追蹤呈現良好的再測信度，組內相關係數 (intraclass correlation coefficients) 為 .81–.92。收斂效度結果顯示，CSCQ與心智理論 (Pearson相關係數r為 .74) 、適應功能 (rs = .41–.76) 、臨床觀察之社交勝任能力 (r = .54) 及自閉症特質 (rs = −.66至 −.43) 均呈現符合理論預期之關聯性 (ps < .01) 。區辨效度結果顯示，一般發展兒童CSCQ總分顯著高於自閉症兒童，t(14) = 4.31，p < .001，Cohen’s dz = 1.11；四個面向均達顯著差異，t = 2.50–5.54，p < .001–.026，Cohen’s dz = 0.64–1.43。此外，CSCQ總分於介入前後呈正向但未顯著變化，Z = 1.69，p = .091，r = .49；社交調適與社交有效性面向分數呈顯著增加，Z = 2.05–2.24，p = .025–.041，r = .59–.65，初步支持其變化反應性。 整體而言，本研究結果支持CSCQ作為一份具理論基礎、由照顧者填寫之日常生活社交勝任能力測量工具，可協助建立結構化的兒童社交勝任能力概況，且應用於自閉症兒童具有良好的信度、收斂效度與區辨效度，並具有追蹤變化的初步應用潛力。未來仍需持續進行驗證研究，以確認其建構效度、建立常模資料與分數解釋指引，並在較大型樣本中確認測量誤差與最小可偵測變化值之估計，同時建立最小臨床重要差異等解釋性指標。

Social competence refers to the ability to understand and respond to social cues in socially adaptive and contextually appropriate ways. Autistic children often experience challenges in social competence across daily contexts. However, existing informant-reported questionnaires commonly have limitations, including insufficient theory–construct alignment, incomplete construct coverage, and limited psychometric evidence. To address these limitations, this study developed the Children’s Social Competence Questionnaire (CSCQ), a caregiver-reported questionnaire, and examined its psychometric properties in autistic children. Based on a multidimensional framework of social competence, the CSCQ covers four domains of daily-life social competence: social skill elements, social reciprocity, social adjustment, and social effectiveness. The study was conducted in two phases. In Phase 1, the CSCQ was developed through identification of referenced measures, item generation, expert content validation, and caregiver-based pretesting. In Phase 2, the psychometric properties of the CSCQ were examined, including internal consistency, test–retest reliability, convergent validity, discriminative validity, and preliminary responsiveness. Children completed researcher-administered assessments, and caregivers completed questionnaire measures. The validation results of Phase 1 supported the content validity of the CSCQ, and revisions were made. The sample in Phase 2 included 110 autistic children aged 3–12 years (mean age = 76.9 months, SD = 27.2; 79.1% boys) and their caregivers, and 15 typically developing (TD) children (mean age = 82.9 months, SD = 20.6; 53.3% boys) were included as a comparison sample. The results showed that the CSCQ total and dimension scores demonstrated acceptable distributional properties. Internal consistency was excellent for the total scale and four dimensions (Cronbach’s alpha coefficients = .91–.98), and test–retest reliability after 2–3 weeks was good (intraclass correlation coefficients = .82–.92). Convergent validity findings showed theoretically expected associations with theory of mind (Pearson’s correlation coefficient, r = .74), adaptive functioning (rs = .41–.76), clinician-rated observational social competence (r = .54), and autistic characteristics (rs = -.66 to -.43), all ps < .01. Discriminative validity findings indicated that TD children had significantly higher CSCQ total scores than autistic children , t(14) = 4.31, p < .001, Cohen’s dz = 1.11; significant group differences were also observed across all four dimensions, t = 2.50–5.54, p < .001–.026, Cohen’s dz = 0.64–1.43. The CSCQ total score showed a positive but nonsignificant pretest-posttest change, Z = 1.69, p = .091, r = .49, whereas the Social Adjustment and Social Effectiveness dimension scores increased significantly, Z = 2.05–2.24, p = .025–.041, r = .59–.65, providing preliminary support for their responsiveness to change. These findings support the CSCQ as a theoretically grounded caregiver-reported measure of daily-life social competence in autistic children. The CSCQ may help organize caregiver-reported information into a structured profile of social competence and may provide a basis for examining social competence profiles and further investigating its potential utility for monitoring change. Further validation is needed to confirm its structural validity, establish normative data and score interpretation guidelines, and examine its measurement error, minimal detectable change, and minimal clinically important difference.