Department of Psychology, Sapienza University of Rome, Rome, Italy.
ISTC Institute for Cognitive Sciences and Technologies, CNR, Rome, Italy.
Psychon Bull Rev. 2018 Jun;25(3):1161-1169. doi: 10.3758/s13423-017-1335-3.
We reanalyzed previous experiments based on lexical-decision and reading-aloud tasks in children with dyslexia and control children and tested the prediction of the difference engine model (DEM) that mean condition reaction times (RTs) and standard deviations (SDs) would be linearly related (Myerson et al., 2003). Then we evaluated the slope and the intercept with the x-axis of these linear functions in comparison with previously reported values (i.e., slope of about 0.30 and intercept of about 300 ms). In the case of lexical decision, the parameters were close to these values; by contrast, in the case of reading aloud, a much steeper slope (0.66) and a greater intercept (482.6 ms) were found. Therefore, interindividual variability grows at a much faster rate as a function of condition difficulty for reading than for lexical-decision tasks (or for other tasks reported in the literature). According to the DEM, the slope of the regression that relates means and SDs indicates the degree of correlation among the durations of the stages of processing. We propose that the need for a close coupling between orthographic and phonological processing in reading is what drives the particularly strong relationship between performance and interindividual variability that we observed in reading tasks.
我们重新分析了基于词汇判断和朗读任务的阅读障碍儿童和正常儿童的先前实验,并测试了差异引擎模型(DEM)的预测,即条件反应时(RT)和标准偏差(SD)的平均值将呈线性关系(Myerson 等人,2003)。然后,我们比较了这些线性函数的斜率和与 x 轴的截距与先前报告的值(即斜率约为 0.30,截距约为 300ms)。在词汇判断的情况下,参数接近这些值;相比之下,在朗读的情况下,发现斜率(0.66)更大,截距(482.6ms)更大。因此,个体间的变异性随着阅读任务条件难度的增加而呈更快的速度增长,而不是词汇判断任务(或文献中报告的其他任务)。根据 DEM,将平均值和 SD 相关联的回归的斜率表明了加工阶段持续时间之间的相关程度。我们提出,阅读中对拼字和语音处理的紧密结合的需求是导致我们在阅读任务中观察到的表现和个体间变异性之间特别强的关系的原因。