BACKGROUND Qualitatively atypical language development characterized by non-sequential skill acquisition within a developmental domain, which has been called developmental deviance or difference, is a common characteristic of autism spectrum disorder (ASD). We developed the Response Dispersion Index (RDI), a measure of this phenomenon based on intra-subtest scatter of item responses on standardized psychometric assessments, to assess the within-task variability among individuals with language impairment (LI) and/or ASD. METHODS Standard clinical assessments of language were administered to 502 individuals from the New Jersey Language and Autism Genetics Study (NJLAGS) cohort. Participants were divided into four diagnostic groups: unaffected, ASD-only, LI-only, and ASD + LI. For each language measure, RDI was defined as the product of the total number of test items and the sum of the weight (based on item difficulty) of test items missed. Group differences in RDI were assessed, and the relationship between RDI and ASD diagnosis among individuals with LI was investigated for each language assessment. RESULTS Although standard scores were unable to distinguish the LI-only and ASD/ASD + LI groups, the ASD/ASD + LI groups had higher RDI scores compared to LI-only group across all measures of expressive, pragmatic, and metalinguistic language. RDI was positively correlated with quantitative ASD traits across all subgroups and was an effective predictor of ASD diagnosis among individuals with LI. CONCLUSIONS The RDI is an effective quantitative metric of developmental deviance/difference that correlates with ASD traits, supporting previous associations between ASD and non-sequential skill acquisition. The RDI can be adapted to other clinical measures to investigate the degree of difference that is not captured by standard performance summary scores.

中文翻译：

---

标准化语言测试的任务内变异性可预测自闭症谱系障碍：反应分散指数的初步研究。

背景技术以发展领域内的非顺序技能获得为特征的定性非典型语言发展，被称为发育偏差或差异，是自闭症谱系障碍（ASD）的共同特征。我们开发了反应分散指数（RDI），它是根据标准心理测验评估上项目反应的子测试内散布来衡量此现象的一种方法，以评估语言障碍（LI）和/或ASD个体的任务内变异性。方法对来自新泽西州语言和自闭症遗传学研究（NJLAGS）队列的502个人进行了标准的语言临床评估。参加者分为四个诊断组：不受影响，仅ASD，仅LI和ASD + LI。对于每种语言量度，RDI定义为测试项目总数与错过的测试项目的权重之和（基于项目难度）的乘积。评估了RDI的组间差异，并针对每种语言评估调查了LI个体中RDI与ASD诊断之间的关系。结果尽管标准成绩无法区分仅使用LI的人群和ASD / ASD + LI的人群，但在所有表达，语用和元语言语言指标上，与仅使用LI的人群相比，ASD / ASD + LI的人群的RDI评分更高。在所有亚组中，RDI与定量ASD特征呈正相关，并且是LI个体中ASD诊断的有效预测指标。结论RDI是与ASD性状相关的发育偏差/差异的有效量化指标，支持ASD和非顺序技能习得之间的先前关联。RDI可以适应其他临床指标，以调查标准性能摘要评分未捕获的差异程度。