Mankovich Amanda, Blume Jessica, Wittke Kacie, Mastergeorge Ann M, Paxton Alexandra, Naigles Letitia R
Department of Psychological Sciences, UConn Child Language Lab, University of Connecticut, Storrs, CT, United States.
Human Development and Family Sciences, Research in Early Developmental Studies Lab, Texas Tech University, Lubbock, TX, United States.
Front Psychol. 2022 Oct 21;13:999396. doi: 10.3389/fpsyg.2022.999396. eCollection 2022.
The current research study characterized syntactic productivity across a range of 5-year-old children with autism and explored the degree to which this productivity was associated with standardized measures of language and autism symptomatology. Natural language samples were transcribed from play-based interactions between a clinician and participants with an autism diagnosis. Speech samples were parsed for grammatical morphemes and were used to generate measures of MLU and total number of utterances. We applied categorical recurrence quantification analysis, a technique used to quantify patterns of repetition in behaviors, to the children's noun-related and verb-related speech. Recurrence metrics captured the degree to which children repeated specific lexical/grammatical units (i.e., recurrence rate) and the degree to which children repeated combinations of lexical/grammatical units (i.e., percent determinism). Findings indicated that beyond capturing patterns shown in traditional linguistic analysis, recurrence can reveal differences in the speech productions of children with autism spectrum disorder at the lexical and grammatical levels. We also found that the degree of repeating noun-related units and grammatical units was related to MLU and ADOS Severity Score, while the degree of repeating unit combinations (e.g., saying "the big fluffy dog" or the determiner-adjective-adjective-noun construction multiple times), in general, was only related to MLU.
当前的研究对一系列5岁自闭症儿童的句法生成能力进行了特征描述，并探究了这种生成能力与语言标准化测量及自闭症症状学之间的关联程度。自然语言样本转录自临床医生与被诊断为自闭症的参与者之间基于游戏的互动。对语音样本进行语法语素分析，并用于生成平均语句长度（MLU）和话语总数的测量值。我们将分类递归量化分析（一种用于量化行为重复模式的技术）应用于儿童与名词相关和与动词相关的言语中。递归指标捕捉了儿童重复特定词汇/语法单位的程度（即重复率）以及儿童重复词汇/语法单位组合的程度（即确定性百分比）。研究结果表明，除了捕捉传统语言分析中显示的模式外，递归还可以揭示自闭症谱系障碍儿童在词汇和语法层面言语产出的差异。我们还发现，重复与名词相关单位和语法单位的程度与平均语句长度和自闭症诊断观察量表（ADOS）严重程度得分相关，而重复单位组合的程度（例如多次说出“那只又大又毛茸茸的狗”或限定词 - 形容词 - 形容词 - 名词结构）一般仅与平均语句长度相关。
