Domenger Dorothée, Schwarting Rainer K W
Experimental and Physiological Psychology, Philipps-University of Marburg, Gutenbergstr. 18, 35032 Marburg, Germany.
Exp Brain Res. 2007 Sep;182(2):223-31. doi: 10.1007/s00221-007-0987-z. Epub 2007 Jun 19.
The serial reaction time task (SRTT) is a well-established experimental tool to study cognitive and neural mechanisms of sequential performance in humans. We have recently developed a rodent version of the human serial reaction time task, in which rats have to respond to visual stimuli by nose-poking into one of four spatial locations in order to obtain food reward. In this task, rats display superior performance under sequential as compared to random conditions of stimulus presentation. Specifically, the subjects are able to profit from sequential regularities in terms of faster reaction times and higher response accuracy. Here, we studied the effects of violating a single stimulus in rats, which had been intensively trained under sequential conditions, and we asked whether these subjects, when confronted with sequence violations, still attend to the actual stimulus order (that is, show correct responses), or whether their behavior has become fully automated (leading to specific incorrect responses to violated stimulus positions). In two independent experiments using partly differing instrumental set-ups, we found that the responses to non-cued violations of single stimulus positions were mostly correct, that is, the animals were apparently attending to the stimuli. Nevertheless, these reaction times were slowed, which probably reflects cognitive resources necessary to respond correctly to the unexpected irregularities. When quantifying the minority of responses, which were incorrect, we found that most of them were directed to the position, where the stimulus would have appeared if the sequence had not been violated. These responses were faster than the correct ones (to the violated stimulus), which indicates that sequential responding had become partly automated. Together, our data show that both, attention and skill play a role for sequential performance in our SRT task, and that they can be dissected by quantification of specific response types. In future work, the neural correlates underlying these functional mechanisms will have to be addressed.
序列反应时任务(SRTT)是一种成熟的实验工具,用于研究人类序列行为的认知和神经机制。我们最近开发了一种啮齿动物版的人类序列反应时任务,在该任务中,大鼠必须通过将鼻子戳到四个空间位置之一来对视觉刺激做出反应,以获得食物奖励。在这个任务中,与随机呈现刺激的条件相比,大鼠在序列条件下表现出更优的性能。具体而言,受试者能够从序列规律中受益,反应时间更快,反应准确性更高。在此,我们研究了在序列条件下经过强化训练的大鼠中违反单个刺激的影响,并询问这些受试者在面对序列违反时,是仍然关注实际的刺激顺序(即做出正确反应),还是其行为已完全自动化(导致对违反刺激位置的特定错误反应)。在两个使用部分不同仪器设置的独立实验中,我们发现对单个刺激位置的非提示性违反的反应大多是正确的,也就是说,动物显然在关注刺激。然而,这些反应时间减慢了,这可能反映了正确应对意外不规则情况所需的认知资源。在量化少数不正确的反应时,我们发现其中大多数指向如果序列未被违反刺激会出现的位置。这些反应比正确反应(对被违反的刺激)更快,这表明序列反应已部分自动化。总之,我们的数据表明,注意力和技能在我们的SRT任务的序列行为中都起作用,并且可以通过对特定反应类型的量化来剖析它们。在未来的工作中,必须解决这些功能机制背后的神经相关性问题。