Abstract: Objective To develop a screening questionnaire of autism suitable for children with Chinese cultural background, and to test its reliability and validity. Methods Screening Questionnaire of Autism in Toddlers (SQAT) was compiled based on the relevant research results and DSM-5 diagnostic criteria for autism spectrum disorder (ASD).Totally 238 children in outpatients or physical examination and 392 children in rehabilitation training institutions were selected for the SQAT assessment, 28 of them were retested 2 to 4 weeks later, and some children were tested by criteria scale.The item selection, reliability and validity of the questionnaire were made by expert evaluation, item analysis, correlation analysis, factor analysis and ROC analysis. Results Thirteen items with good responsiveness (0.394 - 0.626), discrimination (0.465 - 0.697) and common factor loading (0.317 - 0.605) were selected to form the SQAT, which included social and behavioral components.The S-CVI/UA and S-CVI/Ave of expert evaluation was 0.99 and 0.89, respectively.The Cronbach'α coefficients, split-half reliability and retest reliability of the total scale and two components were 0.770 - 0.851, 0.838 - 0.890 and 0.601 - 0.734, respectively.Exploratory factor analysis extracted two factors with eigenvalue >1 and accounted for 50.2% of variance, and confirmatory factor analysis showed that the two factor model of SQAT had a good fit (RMSEA=0.055, NFI=0.872, IFI=0.926, CFI=0.924).The correlation coefficients of total and component scores with ABC and M-CHAT-R/F were 0.518 - 0.656 and 0.305 - 0.833, respectively (P<0.01), and total score of ASD group (23.26 ± 4.73) was higher than that of GDD group (17.32 ± 3.39) and TD group (15.85 ± 2.01) (P<0.01).The sensitivity, specificity and positive predictive value of the SQAT in screening ASD were 82.00%, 80.60% and 96.46%, respectively. Conclusion SQAT has good reliability and validity, which can be used as a preliminary screening tool for children with ASD.

Key words: autism spectrum disorder, screening questionnaire, reliability, validity, infants and toddlers

摘要： 目的 编制适合中国文化背景的婴幼儿孤独症筛查问卷,并检验其信效度。方法 借鉴国内外相关研究成果和参考DSM-5孤独症谱系障碍(ASD)症状标准,编制婴幼儿孤独症筛查问卷(SQAT),选取门诊或体检儿童238名和康复训练机构患儿392名进行初测,2~4周后对其中28名被试进行重测,部分儿童接受校标检测。通过专家评价、条目分析、相关分析、因子分析和ROC分析等筛选条目和检验问卷的信度和效度。结果 筛选出13个具有较好反应度(0.394~0.626)、区分度(0.465~0.697)和公因子负荷(0.317~0.605)的条目组成问卷(SQAT),包含社交和行为两个因子,专家评价内容效度S-CVI/UA和S-CVI/Ave 分别为0.99和0.89。全量表和两因子的Cronbach'α系数为0.770~0.851、分半信度为0.838~0.890、重测信度为0.601~0.734。探索性因子分析提取两个特征根>1的因子、累积方差贡献率50.2%,验证性因子分析显示两因素模型具有良好的拟合度(RMSEA=0.055, NFI=0.872, IFI=0.926, CFI=0.924)。量表总分和各因子与ABC量表和M-CHAT-R/F量表得分的相关系数分别为0.518~0.656、0.305~0.833(P＜0.01),ASD组量表总分(23.26±4.73)高于GDD组(17.32±3.39)和TD组(15.85±2.01)(P＜0.01)。SQAT筛查ASD的灵敏度为82.00%、特异度为80.60%、阳性预测值为96.46%。结论 婴幼儿孤独症筛查问卷具有良好的信效度,可作为ASD儿童的初步筛查工具。

关键词: 孤独症谱系障碍, 筛查问卷, 信度, 效度, 婴幼儿