Roelofs Ardi
Radboud University Nijmegen, Centre for Cognition, Donders Institute for Brain, Cognition and Behaviour, Nijmegen, The Netherlands.
Acta Psychol (Amst). 2010 Feb;133(2):146-53. doi: 10.1016/j.actpsy.2009.11.003. Epub 2009 Dec 5.
Models of attention and context effects in naming performance should be able to account for the time course of color-word Stroop interference revealed by manipulations of the stimulus onset asynchrony (SOA) between color and word. Prominent models of Stroop task performance (Cohen, Dunbar, & McClelland, 1990; Cohen & Huston, 1994; Phaf, Van der Heijden, & Hudson, 1990) fail to account for the fact that response time (RT) and Stroop interference peak at zero SOA and diminish with word preexposure. The models may be saved by assuming that the time course of interference is determined by a strategic orienting of attention to color onsets when SOA is predictable. To test this temporal predictability hypothesis, SOA was blocked or randomly mixed in Experiment 1. In addition, the time interval between color onsets was randomly variable in Experiment 2. Although RTs were affected, none of the randomization manipulations influenced the typical shape of the time course of Stroop effects. These findings provide evidence against the temporal predictability hypothesis and thereby against prominent models of the Stroop task.
命名表现中注意力和语境效应的模型应能够解释通过操纵颜色和单词之间的刺激起始异步性（SOA）所揭示的颜色词斯特鲁普干扰的时间进程。著名的斯特鲁普任务表现模型（科恩、邓巴和麦克莱兰，1990年；科恩和赫斯顿，1994年；法夫、范德海登和哈德森，1990年）未能解释反应时间（RT）和斯特鲁普干扰在零SOA时达到峰值并随着单词预暴露而减弱这一事实。通过假设当SOA可预测时，干扰的时间进程由对颜色起始的注意力策略性定向决定，这些模型可能会得到挽救。为了检验这个时间可预测性假设，在实验1中SOA被分块或随机混合。此外，在实验2中颜色起始之间的时间间隔是随机变化的。尽管反应时间受到了影响，但没有一种随机化操作影响斯特鲁普效应时间进程的典型形状。这些发现提供了反对时间可预测性假设的证据，从而也反对了著名的斯特鲁普任务模型。
