Abstract：
This study investigated whether the interference between two tasks in dual-task processing stems from bottleneck limitations or insufficient cognitive resources due to resource sharing. Experiment 1 used tone discrimination as Task 1 and word or pseudoword classification as Task 2 to evaluate the effect of automatic versus controlled processing on dual-task interference under different SOA conditions. Experiment 2 reversed the task order. The results showed that dual-task interference persisted regardless of task type or order. Neither experiment found evidence that automatic tasks could eliminate interference. This suggests that resource limitations, rather than bottlenecks, may better explain dual-task costs. Specifically, when tasks compete for limited resources, the processing efficiency of both tasks is significantly reduced. Future research should explore how cognitive resources are dynamically allocated between tasks to better account for dual-task interference effects.
摘要：
本研究调查了双任务处理中两个任务之间的干扰是否源于瓶颈限制或由于资源共享而导致的认知资源不足。实验1以声调判别为任务1，以单词或伪词分类为任务2，评估不同SOA条件下自动与受控处理对双任务干扰的影响。实验2颠倒了任务顺序。结果表明，无论任务类型或顺序如何，双重任务干扰都会持续存在。这两个实验都没有发现自动任务可以消除干扰的证据。这表明资源有限，而不是瓶颈，可以更好地解释双重任务成本。具体来说,当任务争夺有限的资源时，这两个任务的处理效率显著降低。未来的研究应该探索认知资源如何在任务之间动态分配，以更好地考虑双重任务干扰效应。
