孤独症谱系商儿童版-中文版的效度和信度  

Reliability and validity of the Chinese version of Autism Spectrum Quotient-Children's version

作　　者：李洪成 刘靖[1] 贾美香[1] 杨柳[1] 李雪[1] 张爱武[1] 庞瑾 LI Hongcheng;LIU Jing;JIA Meixiang;YANG Liu;LI Xue;ZHANG Aiwu;PANG Jin(Peking University Institute of Mental Health,Peking University Sixth Hospital,Key Laboratory of Mental Health,Ministry of Health(Peking University),National Clinical Research Center for Mental Disorders (Peking University Sixth Hospital),Beijing 100191,China;Ankang Hospital,Tianjin City Public Security Bureau,Tianjin 300240,China)

机构地区：[1]北京大学精神卫生研究所,北京大学第六医院,卫生部精神卫生重点实验室(北京大学),国家精神心理疾病临床医学研究中心(北京大学第六医院),北京100191 [2]天津市公安局安康医院,天津300240

出　　处：《中国心理卫生杂志》2018年第12期1000-1005,共6页Chinese Mental Health Journal

基　　金：科技部重大慢病项目(2017YFC1309900);国家自然科学基金(81571339);首都特色临床项目(Z171100001017088).

摘　　要：目的:检验孤独症谱系商(AQ)儿童版-中文版的效度信度。方法:对符合美国精神障碍诊断与统计手册第4版(DSM-IV)孤独症诊断标准的门诊101例4～11岁孤独症儿童和124例年龄相匹配的正常儿童进行AQ儿童版-中文版评定。以两组间AQ儿童版-中文版总分及因子分比较确定区分效度;以DSM-IV诊断为金标准,以ROC方法确定界限分及灵敏度、特异度; 10例孤独症儿童和10例正常儿童的两位家长同时单独填写AQ儿童版-中文版以检验评分者信度; 10例孤独症儿童和10例正常儿童家长于首次评定2周后再次进行AQ儿童版-中文版评定,以检测重测信度;以Cronbach a系数检验内部一致性。结果:孤独症组AQ儿童版-中文版总分及各分量表分均高于正常对照组[如总分:(94. 2±15. 6)vs.(54. 4±13. 2),P <0. 001]; ROC曲线下面积为0. 98 (P <0. 01),界限分取73时,量表的灵敏度为0. 96,特异度为0. 90;量表总分评分者信度ICC和重测信度ICC分别为0. 80 (P <0. 001)和0. 92 (P <0. 001);量表内部一致性检验Cronbach a系数为0. 94 (P <0. 01),5个分量表的Cronbach a系数为0. 69～0. 89 (均P <0. 01)。结论:AQ儿童版-中文版的效度、信度较好,灵敏度、特异度较高,可用于儿童孤独症的临床筛查。Objective: To test the validity and reliability of the Chinese version of Autism Spectrum Quotient-Children’ s Version( Chinese version of the AQ-child). Methods: Totally 101 children aged 4-11 years old meeting criterion of autistic disorder in Diagnostic and Statistical Manual of Mental Disorders,Fourth Edition( DSM-IV)and 124 normal children matched with autistic group in age were included and evaluated with the Chinese version of AQ-child. The comparison of total scale score and subscale scores were used to analyze the discriminant validity. With DSM-IV diagnosis as the gold standard,ROC was used to determine the cutoff point and sensitivity and specificity. Two caregivers of 10 children with autistim and 10 normal children evaluated the children with the Chinese version of AQ-child separately at the same time to test the inter-rater reliability. Ten children with autistic disorder and 10 normal children were evaluated again with the Chinese version of AQ-child two weeks later by the same evaluator to investigate the test-retest reliability. Results: The scores of total scale and all subscales were higher in autism group than in normal control group [e. g. total scale score,( 94. 2 ± 15. 6) vs.( 54. 4 ± 13. 2),P < 0. 001]. The area under the ROC curve was 0. 98( P < 0. 01). At cutoff of 73,the sensitivity was 0. 96,and the specificity was0. 90. For the inter-rater reliability and the test-retest reliability of Chinese version of AQ-child,the intraclass correlation coefficient( ICC) of total score was 0. 80( P < 0. 001) and 0. 92( P < 0. 001) respectively. The coefficient of internal consistence of the scale was 0. 94( P < 0. 01),the coefficient of internal consistence of the five subscales ranged 0. 69-0. 89( Ps < 0. 01). Conclusion: The Chinese version of AQ-child has good validity and reliability,and high sensitivity and specificity. It could be used as a clinical screening instrument for autism in children.

关 键 词：孤独症 孤独症谱系商儿童中文版 效度 信度 

分 类 号：R749.94[医药卫生—神经病学与精神病学]