De Luca Maria, Marinelli Chiara Valeria, Spinelli Donatella, Zoccolotti Pierluigi
Neuropsychology Unit, IRCCS Fondazione Santa Lucia, Rome, Italy.
Laboratory of Applied Psychology and Intervention "DREAM", Department of History Society and Human Studies, University of Salento, Lecce, Italy.
Exp Brain Res. 2017 Oct;235(10):3093-3109. doi: 10.1007/s00221-017-5041-1. Epub 2017 Jul 25.
We examined the slowing in vocal reaction times shown by dyslexic (compared to control) children with that of older (compared to younger) adults using an approach focusing on the detection of global, non-task-specific components. To address this aim, data were analyzed with reference to the difference engine (DEM) and rate and amount (RAM) models. In Experiment 1, typically developing children, children with dyslexia (both attending sixth grade), younger adults and older adults read words and non-words and named pictures. In Experiment 2, word and picture conditions were presented to dyslexic and control children attending eighth grade. In both experiments, dyslexic children were delayed in reading conditions, while they were unimpaired in naming pictures (a finding which indicates spared access to the phonological lexicon). The reading difficulty was well accounted for by a single multiplicative factor while only the residual effect of length (but not frequency and lexicality) was present after controlling for over-additivity using a linear mixed effects model with random slopes on critical variables. Older adults were slower than younger adults across reading and naming conditions. This deficit was well described by a single multiplicative factor. Thus, while slowing of information processing is limited to orthographic stimuli in dyslexic children, it cuts across verbal tasks in older adults. Overall, speed differences in groups such as dyslexic children and older adults can be effectively described with reference to deficits in domains encompassing a variety of experimental conditions rather than deficits in single specific task/conditions. The DEM and RAM prove effective in teasing out global vs. specific components of performance.
我们采用一种专注于检测全局、非任务特定成分的方法，将诵读困难儿童（与对照组相比）与年长成年人（与年轻成年人相比）在语音反应时间上的减慢情况进行了对比。为实现这一目标，我们依据差异引擎（DEM）以及速率和数量（RAM）模型对数据进行了分析。在实验1中，正常发育的儿童、患有诵读困难的儿童（均为六年级学生）、年轻成年人以及年长成年人阅读单词和非单词并为图片命名。在实验2中，向八年级的诵读困难儿童和对照组儿童呈现单词和图片条件。在两个实验中，诵读困难儿童在阅读条件下反应延迟，而在图片命名方面未受损害（这一发现表明其语音词汇的获取未受影响）。阅读困难可由一个单一的乘法因子很好地解释，而在使用具有关键变量随机斜率的线性混合效应模型控制了过度相加性之后，仅存在长度的残余效应（而非频率和词汇性）。在阅读和命名条件下，年长成年人比年轻成年人反应更慢。这一缺陷可由一个单一的乘法因子很好地描述。因此，虽然诵读困难儿童的信息处理减慢仅限于正字法刺激，但在年长成年人中则贯穿于言语任务。总体而言，诸如诵读困难儿童和年长成年人等群体中的速度差异，可参照涵盖多种实验条件的领域中的缺陷，而非单一特定任务/条件中的缺陷来有效描述。差异引擎（DEM）和速率与数量（RAM）模型在区分表现的全局与特定成分方面证明是有效的。
