Efficiency of task switching and effects of prediction errors as a function of the number of action alternatives

当不同的任务在时间上连续执行时，任务切换与切换成本相关，即处理速度的减速和错误率的增加。这些转换成本可以通过准备来降低，但目前还不清楚，因为有效的准备依赖于外部刺激的可用性(外部准备)，或者可以仅基于内部表征进行，如计划(内源准备)。我们可以表明，当人们只在两个任务之间切换时，内源准备的程度较小(就像大多数任务切换研究中一样)，而四个任务之间的切换更容易受到内源准备的影响。此外，我们还获得了在两个任务之间转换时建立对抗约束机制的证据，这种机制被认为促进了基于外部刺激的任务转换。此外，我们还提供了证据表明，当某个任务与预期任务不同时，对该任务的显性期望会促进认知控制。我们计划通过对比期待任务的过程和在任务中选择任务的过程来跟进这些研究。任务期望和任务选择在人们体验到对即将到来的任务的控制程度方面有所不同。我们提出了一个范式，其中目标控制度对于任务预测和任务选择是相同的。我们的目标是调查当任务数量从四个减少到两个时，这两种情况是否都会导致建立对抗限制。此外，我们计划比较不同于预期或选择的任务对认知控制过程的招募，并检验不同数量的任务以及不同数量的运动动作对这些过程的调节作用。此外，我们计划调查任务预测和任务选择的经验控制程度的差异是否伴随着代理感的变化，以及这种变化是否反过来调节认知控制过程的招募。

When different tasks are performed in temporal succession, task switches are associated with switch costs, that is, a deceleration of processing speed and an increase in error rates. These switch costs can be reduced by preparation, but at present, it is unclear inasmuch efficient preparation depends on the availability of external stimuli (exogenous preparation) or can proceed solely on the basis of internal representations like plans (endogenous preparation).We could show that the extent of endogenous preparation is smaller when people switch among only two tasks (as in most studies of task switching), whereas switches among four tasks are more susceptible to effects of endogenous preparation. Furthermore, we obtained evidence for a mechanism of establishing antagonistic constraints when switching among two tasks which is thought to facilitate task switching based on external stimuli. In addition, we provided evidence that explicit expectations of a certain task result in a boost of cognitive control when the relevant task differs from the expected one. We plan to follow up on these studies by contrasting processes of expecting a task with processes of choosing among tasks. Task expectation and task choice differ with respect to the degree that people experience control over an upcoming task. We propose a paradigm in which the degree of objective control is the same for task predictions and task choices. Our goal is to investigate whether both situations result in the establishment of antagonistic constraints when the number of tasks is reduced from four to two. Furthermore, we plan to compare the recruitment of cognitive control processes induced by a task that differs from the expected or chosen one, and to examine the modulating effects of different numbers of tasks and, in comparison to that, different numbers of motor actions on these processes. In addition, we plan to investigate whether differences in the degree of experienced control for task predictions and task choices go along with variations in the sense of agency, and whether such variations in turn modulate the recruitment of cognitive control processes.