Modality-specific crosstalk in multitasking: Evidence from modality compatibility in task switching

实验研究显示，在人类同时处理多项任务时，性能会产生巨大的代价。在本项目的优先计划“人类在多重认知任务需求下的表现：从基本机制到优化的任务调度”的背景下，我们关注使用不同的任务切换范式的连续多任务处理。在这个范式中，基本的发现是，任务切换的表现比重复差。这些转换成本被用来标示认知控制过程在高阶任务表征(“任务集”)水平上的参与程度。本研究旨在通过考察与串联任务转换相关的任务表征的结构、灵活性和可塑性等问题，特别强调特定于通道的效应(“通道兼容性”，MC)，为PP的目标做出贡献。我们将MC定义为刺激通道与预期反应效果的通道相对应的程度(例如，听觉-声音任务映射将是通道兼容的，因为语音反应通常旨在产生听觉效果)。在第一个资助阶段，我们发现在两个不相容的通道映射(如听觉-手动和视觉-声音)之间切换时的转换成本比在两个兼容映射之间切换时的切换成本更大。我们的理论解释认为，MC在多任务中的这种效应来源于“表意运动”学习(即学习反应效应的偶然性和对反应效应的预期)。将MC的影响根植于人类行动控制的功能约束中，引发了一种广泛的经验方法，这引发了关于人类行动控制和任务集的特定背景下的灵活性和学习依赖的可塑性的研究问题。具体地说，本研究旨在更好地理解(1)任务和映射是如何表征的(结构问题)，(2)任务和映射是如何在自由选择中自愿选择的(灵活性)，以及(3)任务集如何随着短期和长期的练习而变化(可塑性)。在第二个项目阶段，我们的目标是加深我们对多任务处理中特定通道交互的理解。首先，我们将MC的概念推广到不同的加工要求和实验范式，包括在任务分担的情况下将通道映射在参与者之间进行划分的条件。其次，我们探讨了MC在自由(“自愿”)反应方式选择中的作用。第三，我们考察了在连续任务(即映射)转换和同时双重任务中，从单一映射和混合映射练习中迁移的边界条件。最后，我们还打算扩大现有的合作关系，并发展新的合作关系。

Experimental studies revealed robust performance costs in human multitasking. In the present project in the context of the Priority Program "Human Performance under Multiple Cognitive Task Requirements: From Basic Mechanisms to Optimized Task Scheduling", we focus on serial multitasking using variations of the task-switching paradigm. In this paradigm, the basic finding is that performance is worse in task switches than in repetitions. These switch costs have been taken to index the involvement of cognitive control processes at the level of higher-order task representations ("task sets"). The present individual PP project aims at contributing to the goals of the PP by examining issues pertaining to the structure, flexibility, and plasticity of the task representations underlying serial task switching, with a specific emphasis on modality-specific effects ("modality compatibility", MC). We define MC as occurring to the degree that the stimulus modality corresponds to the modality of the intended response effect (e.g. auditory-vocal task mappings would be modality compatible because vocal responses are typically intended to produce auditory effects). In the first funding period, we found that switch costs are larger when switching between two incompatible modality mappings (e.g. auditory-manual and visual-vocal) than when switching between two compatible mappings. Our theoretical account proposes that such effects of MC in multitasking derive from "ideomotor" learning (i.e. learning response-effect contingencies and anticipation of response effects). Rooting the influence of MC in the functional constraints of human action control invites a broad empirical approach that raises research issues referring to the context-specific flexibility and learning-dependent plasticity of human action control and task set. Specifically, this research aims at understanding better (1) how tasks and mappings are represented (structural issue), (2) how tasks and mappings can be selected voluntarily in "free choices" (flexibility), and (3) how task sets can change with short-term and long-term practice (plasticity). In the second project phase, we aim at deepening our understanding of modality-specific interactions in multitasking. First, we generalize the notion of MC to different processing requirements and experimental paradigms, including conditions in which the modality mappings are divided across participants in a “task sharing” situation. Second, we explore the role of MC in free ("voluntary") response modality choices. Third, we examine the boundary conditions for transfer from single-mapping and mixed-mapping practice both in serial task (i.e. mapping) switching and simultaneous dual tasks. Finally, we also aim to extend existing and develop new cooperative links.