Investigating task order control in dual-task situations: The influence of instruction, task complexity, and practice

点击翻译按钮获取中文摘要

Performance costs (in reaction times and/or error rates) occur in situations in which people perform two tasks simultaneously compared to their isolated performance. Many theorists explain these dual-task costs with a bottleneck mechanism that requires serial processing of certain processes in dual-task situations. Many previous studies have provided valuable knowledge about the nature of the bottleneck, i.e. whether it is caused by structural capacity limitation or a strategic choice and whether it is an all-or-none or a shared bottleneck. However, despite this knowledge, it is still open which mechanisms regulate and control the order of the tasks before they enter the bottleneck and how subjects can flexibly change the order of the tasks at and before the bottleneck.While the traditional bottleneck model assumed a passive mechanism for determining task order, which predicts that that task is processed first which arrives first the bottleneck and which does not involve any further control mechanisms, the current project is based on an alternative approach. As a main presumption, we assume that the involvement of a bottleneck leads to competition between processes for access to the bottleneck and this competition requires the recruitment of additional control mechanisms. The appropriate order of the processes at the bottleneck needs to be planned, controlled, and regulated by additional processes. How do these control processes work? What are their specific mechanisms and principles? The aim of the current project is to investigate these questions. For that purpose the research on the first project phase was based on a task order control model, which allowed us formulating different predictions about different mechanisms regulating task order.In the current proposal we will continue to investigate task order control in dual tasks. The overarching goal of the new studies is the elaboration of a proposed task order control model that combines assumptions about bottleneck processing, executive task order control processes, and working memory mechanisms. Prior work in the first phase showed suitability of the model for understanding important aspects of task order control in dual tasks, but, additionally, provided new findings, which require further specification and testing in the second phase.For that purpose we aim to conduct several series of experiments, which are aimed at testing the predictions of the task order control model, 1) concerning the content and the structure of task order control representations guiding task processing in dual tasks, and 2) concerning the influence of different factors on the efficiency of task order control, such as the order instruction, the amount of working memory load, the type of the order decision criterion, and of individual task order preferences. Further, 3) we want to investigate to which degree task order control can be optimized by training and what the neural implementation of the control is.