Abstract: Objective To study the construct validity and effective validity of the revised 〈China Developmental Scale for Children〉 and provide support for national norm and diagnostic development scale. Methods The revised 〈China Developmental Scale for Children〉 was used among sampled Beijing population.Item analyse,exploratory factor analysis,confirmatory factor analysis and structural equation modelling analysis were used to determine the scale structure and structure effectiveness.Gesell Development Scale was performed to external effective validity analyse. Results Among 2 779 from 0 to 84 months old children,1 648 were boy and 1 311 were girl.77 329 tests were performed.99.2% (259/261) of the test items had discriminability (critical ratio,P<0.05).Regroup the samples for every three months age.9 subsets KMO were 0.938,0.954,0.933,0.932,0.940,0.942,0.948,0.976,0.956;5 dimension of the scale and extract factor loading analyze generally matched.After performing compulsory testing verify analyses for 9 subsets,idea of five dimension and actual data were fitted well.Considering the characteristic of development scale and the sample request of verify factor analyses,adjacent month aging was regrouped and verify analyses,the analyses parameter CFI and NNFI were both greater than 0.9,79.2% (19/24) of the sample had RMSEA lower than 0.08,the highest value was 0.98 which still matched the five dimension idea.The Pearson factor between Gasell scale was 0.637. Conclusion The revised 〈China Developmental Scale for Children〉 has good structure effectiveness and discrimination.Five dimension structure is reasonable and matches the diagnose development scale parameters.The result can be adapted as the foundation of national norm scale.

Key words: child, development, scale, construct validity, factor analysis, confirmatory factor analysis

摘要： 目的 研究“儿心量表”修订后《中国儿童发育量表》的结构效度及校标效度,为制定全国常模以及诊断性发育量表提供可行性依据。方法 在北京市地区抽样人群中,用《中国儿童发育量表》初始修订表进行测试;运用项目分析、探索性因素分析、验证性因素分析及结构方程模型分析方法确定量表的结构及结构效度;以Gesell发展量表为外在校标进行相关分析。结果 共测试0~84月龄儿童2 779名,男童1 468名,女童1 311名。男:女比为1.1∶1,按项目实际受试人次计,共测试77 329人次。具有鉴别力的(CR值,P<0.05)项目占99.2%(259/261)。将量表每3个月的项目为一个整体(子集)进行因素分析, 9个子集的KMO值分别为0.938、0.954、0.933、0.932、0.940、0.942、0.948、0.976、0.956;旋转后提取因子特征值大于1的累积方差贡献率分别为66.9%、 55.2%、57.6%、59.4%、50.5%、59.7%、61.0%、59.7%、59.2%。量表的5个维度与提取因子载荷匹配基本吻合。通过对以因素分析的9个子集进行强制性验证性因素分析,5个维度构想与实际数据的拟合度良好。考虑发展量表特点及验证性因素分析样本量要求,再将前后连续、相邻月龄项目进行组合即1月龄与2月龄,2月龄与3月龄之后以此类推,再进行验证分析,分析参数CFI及NNFI均大于0.9,RMSEA在0.08以下占79.2%(19/24),最大为0.98,仍提示五维度的构想合理。与Gesell量表的关系(Pearson)系数为0.637。结论 修订后《中国儿童发育量表》结构效度较高,测验项目鉴别力强,五维度的量表结构合理,符合诊断性发育评估量表参数,可作为制定全国常模的基础量表。

关键词: 儿童, 发育, 量表, 结构效度, 因素分析, 验证性因素分析