年龄与发育进程问卷中国常模及心理测量学特性研究

Studies of the norm and psychometrical properties of the ages and stages questionnaires, third edition, with a Chinese national sample

目的探讨年龄与发育进程问卷一第3版（ASQ-3）中文版（ASQ-C）的全国常模及其心理测量学特性。方法通过正向翻译、逆向翻译等6个步骤将ASQ．3翻译和文化适应性改编成ASQ-C。采用定额样本，每套ASQ-C至少200人。根据2010年第六次全国人口普查资料，按照各行政区人口数、城镇与乡村人口比例等在中国大陆1～66月龄儿童中进行分层整群随机抽样，抽取有代表性的常模样本。剔除标准：平均海拔高于2000m、通行语言不是简体中文的社区和村；已经由权威机构明确诊断为发育迟缓的儿童。常模样本4452人，其中每套ASQ-C的样本量为218～227人，男2230人、女2222人，城镇2236人、乡村2216人。在常模样本中选择便N）L童作为测试者间信度及重测信度检验样本。6个行政区都参加了信度检验样本的收集，测试者间信度检验样本量为162人（每套ASQ-C为5～9人），重测信度检验样本量为168人（每套ASQ-C为6-10人）。北京盖泽尔作为效度标准，选用低出生体重儿随访门诊中的便利样本，样本量共198人。ASQ-C的常模及重测信度检验，由儿童的家长独立或在项目人员提供需要的帮助下完成；测试者间信度检验，由项目人员在儿童的父母完成常模ASQ-C后再填写一份相同的ASQ-C；效度检验在儿童的父母完成ASQ-C后由专业人员施测北京盖泽尔。采用SPSS13．0进行统计分析，建立离均差常模、进行信度及效度分析。结果常模样本在性别构成、城镇与乡村构成和家庭经济收入等都基本符合第六次全国人口普查资料；建立了20套ASQ-C及其21套常模。ASQ-C的内在信度Cronbach’S仅系数为0．8，测试者间信度r为0．8，重测信度r为0．8（P均〈0．0001）。ASQ-C识别发育迟缓的敏感度为87．50％、特异度为84．48％、准确度为84．74％。结论ASQ-C的常模具有代表性，有良好的信度及效度，可用于中国大陆适龄儿童的发育筛查及监测。

Objective To introduce the Ages and Stages Questionnaires,Third Edition （ASQ-3） ,to China,created ASQ-Chinese （ ASQ-C ） and carried out studies of its national norm and the psychometrical properties in the children aged 1 - 66 months in the mainland of China in collaboration with the author of theASQ System and under the authorizations from its publisher on translation, researches, publication and distribution of the ASQ-3. Method The ASQ-3 questionnaires were translated and adapted into a Simplified Chinese version,the ASQ-C,with six steps such as translation, back-translation and adaptation and so on to ensure consistency with the core of the original document and to have the cultural relevance in China. A stratified cluster sampling method was utilized to recruit children aged 1 -66 months with respect to demographic characteristics such as the proportion of populalion in each administrative region and in urban and rural areas and so on that are representative of 2010 China census data. A sample size of over 200 was collected for each ASQ-C age interval. Children were excluded from the normative sample who （ 1 ） are from communities or villages at an elevation of 2 000 m or above and （or） where simplified Chinese is not the official language, or （2） had been diagnosed as having a developmental delay by any authoritative organizations. The national normative sample for the ASQ-C had a total sample size of 4 452, sample size within each age interval ranged from 218 to 227 ,including 2 230 male cases and 2 222 female cases,2 236 urban cases and 2 216 rural cases. A convenience sample was recruited from the normative sample to examine inter-rater reliability and test-retest reliability in all six administrative regions. Researchers completed the ASQ-C on the same child with their parents for 162 children for inter-rater reliability（ the size of each ASQ- C age interval was 5 -9） ; parents of 168 children completed another age-appropriate ASQ-C for test-retest reliability during 10 - 15 days after they completed the normative ASQ-C （ The size of each ASQ-C age interval is 6 -10）. Another convenience sample was recruited from the follow-up of low birth weight infants for the concurrent validity of the ASQ-C in comparison with the Beijing Gesell. Parents of 198 children completed age-appropriate ASQ-C and professional administered to the children with the Beijing Gesell. In the ASQ-C norm and test-retest reliability, parents completed the age-appropriate ASQ-C, independently or with needed assistance. In inter-rater reliability, researchers completed the same ASQ-C after parents. In validity test, after parents completing age-appropriate ASQ-C, professional tested children with the Beijing Gesell. Data were analyzed using SPSS version 13.0 software. The mean and standard deviation of the national normative sample were calculated, reliability and validity of the ASQ-C was examined. Result The demographic characteristics of this Chinese sample match the 2010 China census data on gender, urban or rural location, and family income. All 20 intervals of the ASQ-C were standardized on 21 national normative samples. Cronbach＇s alpha coefficient for the whole measure was 0. 8. The Pearson correlation coefficient between the ASQ-C total scores of the two raters was 0. 8. The Pearson correlation coefficient between the ASQ-C total scores of the two times was 0. 8 （ all P 〈 0. 000 1 ）. The sensitivity of ASQ-C was 87.50% and the specificity of ASQ-C was 84. 48%. The percentage of the agreement between the ASQ-C and the Beijing Gesell was 84. 74%. Conclusion These findings indicate that the ASQ-C is a reliable and valid measure with a representative national sample aged 1 -66 months. It can be used to screen and monitor the development of children in the mainland of China.