Schwarz Wolfgang, Ischebeck Anja
University of Nijmegen
J Math Psychol. 2001 Jun;45(3):452-479. doi: 10.1006/jmps.2000.1336.
The central bottleneck model of dual-task performance (H. Pashler and J. C. Johnston, 1998, Quarterly Journal of Experimental Psychology, 46A, 51-82) and the serial processing model of precue utilization (R. Gottsdanker, 1992, Acta Psychologica, 79, 21-43) are based on a common formal structure: They both represent response time as RT(tau)=max(X-tau, Y)+Z, where X, Y, Z denote the duration of certain processing stages specified by the models and tau denotes the onset asynchrony (SOA) between two stimuli. We consider this model within a stochastic framework in which the stage durations are random variables following an arbitrary joint distribution and derive properties of the function relating E[RT(tau)] to SOA. We present a distribution-free result which relates the slope of this function to the distribution of the random durations of the assumed processing stages. Our results allow for a direct, model-based interpretation of data from related experiments; specifically, they show how the slope of the SOA-function depends on experimental factors which selectively influence individual processing stages. We explain the implications of our results for models of dual-task performance and precue utilization and illustrate their application to data obtained by M. C. Smith (1969, Acta Psychologica, 30, 220-231) and R. Gottsdanker (1992, loc. cit.) Copyright 2001 Academic Press.
双重任务表现的中央瓶颈模型（H. 帕什勒和J. C. 约翰斯顿，1998年，《实验心理学季刊》，46A，51 - 82）和预线索利用的串行处理模型（R. 戈特丹克，1992年，《心理学报》，79，21 - 43）基于一个共同的形式结构：它们都将反应时间表示为RT(tau)=max(X - tau, Y)+Z，其中X、Y、Z表示模型指定的某些处理阶段的持续时间，tau表示两个刺激之间的起始异步（SOA）。我们在一个随机框架内考虑这个模型，其中阶段持续时间是遵循任意联合分布的随机变量，并推导将E[RT(tau)]与SOA相关联的函数的性质。我们给出了一个无分布结果，该结果将此函数的斜率与假定处理阶段的随机持续时间的分布相关联。我们的结果允许对来自相关实验的数据进行直接的、基于模型的解释；具体而言，它们展示了SOA函数的斜率如何取决于选择性影响各个处理阶段的实验因素。我们解释了我们的结果对双重任务表现和预线索利用模型的影响，并说明了它们在M. C. 史密斯（1969年，《心理学报》，30，220 - 231）和R. 戈特丹克（1992年，同刊）所获得数据中的应用。版权所有2001年学术出版社。
