Callier Center for Communication Disorders, University of Texas at Dallas.
Reliant Rehabilitation, Plano, TX.
J Speech Lang Hear Res. 2019 Jun 19;62(6):1813-1822. doi: 10.1044/2019_JSLHR-L-18-0437. Epub 2019 May 20.
Purpose We investigated the finite verb morphology composite (FVMC), a measure associated with developmental language disorder (DLD), in a large community sample to provide evidence on its distribution, its association with other variables, and its sensitivity and specificity. Method We coded percent occurrence of FVMC morphemes in obligatory contexts in archival language samples from 676 six-year-olds conversing with an adult family member. Using multiple regression, we quantified the variance in FVMC scores accounted for by language sample characteristics (number of utterances and obligatory contexts) and child/family variables (performance IQ, family history, maternal education, and adult FVMC). We calculated sensitivity and specificity of low (< 85%) FVMC scores for predicting low (-1 SD) scores on measures of utterance length, receptive vocabulary, and nonword repetition. Results FVMC scores ranged from 33% to 100% ( M = 93%). Number of obligatory contexts ( R = 3%), performance IQ ( R = 11%), and adult FVMC ( R = 18%) were significant covariates. Sensitivity ranged from 25% to 35%; specificity was from 91% to 92%. Conclusion FVMC scores at the age of 6 years were generally high, but 12% of the participants had FVMCs below 85%, and scores were significantly associated with characteristics of the language samples, children, and adult interlocutors. Sensitivity of the FVMC was considerably lower in this community sample than in previous studies comparing groups of children who met criteria for DLD or typical language. Evidence from large representative samples is important when developing and validating potential clinical markers of DLD.
我们研究了与发育性语言障碍（DLD）相关的有限动词形态复合（FVMC），旨在为其分布、与其他变量的关联以及其敏感性和特异性提供大量社区样本的证据。
我们对 676 名与成年家庭成员交谈的 6 岁儿童的档案语言样本中的 FVMC 词素在强制性语境中的出现百分比进行了编码。使用多元回归，我们量化了语言样本特征（话语数量和强制性语境数量）和儿童/家庭变量（表现智商、家族史、母亲教育程度和成人 FVMC）对 FVMC 分数的方差的解释。我们计算了 FVMC 分数低于 85%的低分数（< -1 SD）预测话语长度、接受性词汇和非词重复等测量结果的低分数的敏感性和特异性。
FVMC 分数范围为 33%-100%（M = 93%）。强制性语境数量（ R = 3%）、表现智商（ R = 11%）和成人 FVMC（ R = 18%）是显著的协变量。敏感性范围为 25%-35%；特异性为 91%-92%。
6 岁时的 FVMC 分数通常较高，但 12%的参与者的 FVMC 分数低于 85%，且分数与语言样本、儿童和成人对话者的特征显著相关。与之前比较符合 DLD 标准或典型语言的儿童组的研究相比，该社区样本中 FVMC 的敏感性要低得多。在开发和验证 DLD 的潜在临床标志物时，来自大型代表性样本的证据非常重要。
