We examined the hidden costs of intermittent multitasking. Participants performed a pursuit-tracking task (Experiment 1) or drove in a high-fidelity driving simulator (Experiment 2) by itself or while concurrently performing an easy or difficult backwards counting task that periodically started and stopped, creating on-task and off-task multitasking epochs. A novel application of the Detection Response Task (DRT), a standardized protocol for measuring cognitive workload (ISO 17488, 2016), was used to measure performance in the on-task and off-task intervals. We found striking costs that persisted well after the counting task had stopped. In fact, the multitasking costs dissipated as a negatively accelerated function of time with the largest costs observed immediately after multitasking ceased. Performance in the off-task interval remained above baseline levels throughout the 30-s off-task interval. We suggest that loading new procedures into working memory occurs fairly quickly, whereas purging this information from working memory takes considerably longer. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

中文翻译：

---

分心的持久性：间歇性多任务处理的隐性成本。

我们研究了间歇性多任务处理的隐性成本。参与者自己执行追踪跟踪任务（实验 1）或驾驶高保真驾驶模拟器（实验 2）或同时执行简单或困难的倒数任务，该任务周期性地开始和停止，创建任务中和关闭任务多任务时代。检测响应任务 (DRT) 的一种新应用，一种用于测量认知工作量的标准化协议 (ISO 17488, 2016)，用于测量在任务和非任务间隔中的性能。我们发现，在计数任务停止后，惊人的成本仍然存在。事实上，多任务处理成本作为时间的负加速函数消散，在多任务处理停止后立即观察到最大成本。在整个 30 秒的非任务时间间隔内，非任务时间间隔内的性能保持在基线水平之上。我们建议将新程序加载到工作记忆中的速度相当快，而从工作记忆中清除这些信息需要更长的时间。（PsycInfo 数据库记录 (c) 2022 APA，保留所有权利）。