Nisonger Center and Department of Psychology, The Ohio State University, Nisonger Center, 305 McCampbell Hall, 1581 Dodd Drive, Columbus, OH 43210, USA.
J Autism Dev Disord. 2012 Jun;42(6):1075-86. doi: 10.1007/s10803-011-1348-0.
The current study tested several competing models of the autism phenotype using data from modules 1 and 3 of the ADOS. Participants included individuals with ASDs aged 3-18 years (N = 1,409) from the AGRE database. Confirmatory factor analyses were performed on total samples and subsamples based on age and level of functioning. Three primary models were tested, including a one-factor model, the DSM-IV model, and the anticipated DSM-V model. Results indicated all models fit similarly. Module 1 ratings yielded better indices of fit across all models and higher inter-factor correlations than Model 3. Model fits were impacted by age and level of functioning. The lack of differentiation between models suggests that the structure of ASD symptoms is complex to measure statistically.
本研究使用 ADOS 的模块 1 和 3 的数据，测试了几种自闭症表型的竞争模型。参与者包括来自 AGRE 数据库的年龄在 3-18 岁的自闭症谱系障碍个体（N=1409）。对总样本和基于年龄和功能水平的子样本进行了验证性因子分析。测试了三个主要模型，包括一个单因素模型、DSM-IV 模型和预期的 DSM-V 模型。结果表明，所有模型的拟合度相似。在所有模型中，模块 1 的评分都产生了更好的拟合指数，且因子间的相关性也高于模型 3。模型拟合受到年龄和功能水平的影响。模型之间缺乏差异表明，自闭症症状的结构很难在统计学上进行测量。
