The persistence of distraction: The hidden costs of intermittent multitasking.

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).

Multiple processing limitations underlie multitasking costs.

Human multitasking is typically defined as the practice of performing more than one task at the same time (dual task) or rapidly alternating between multiple tasks (task switching). The majority of research in multitasking has been focusing on individual paradigms, with surprisingly little effort in understanding their relationships. We adopted an individual-difference approach to reveal the limitations underlying multitasking costs measured in different paradigms. Exploratory factor analyses revealed not a general multitasking factor but instead three different processing limitations associated with response selection, retrieval and maintenance of task information, and task-set reconfiguration. The three factors were only weakly correlated with and thus not reducible to common measures of processing speed, working memory capacity and fluid intelligence. Males and females excelled in different aspects of multitasking, demonstrating the benefit of using a multifaceted view of multitasking competency in group comparison. Findings of the current study help resolve conflicting results between studies using different paradigms, and form the basis of more comprehensive measurement tools and training protocols covering different aspects of multitasking limitations. The study will also help future integration of multitasking abilities into the theoretical framework of executive function.

Putting a stereotype to the test: The case of gender differences in multitasking costs in task-switching and dual-task situations.

According to a popular stereotype, women are better at multitasking than men, but empirical evidence for gender differences in multitasking performance is mixed. Previous work has focused on specific aspects of multitasking or has not considered gender differences in abilities contributing to multitasking performance. We therefore tested gender differences (N = 96, 50% female) in sequential (i.e., task switching) and concurrent (i.e., dual tasking) multitasking, while controlling for possible gender differences in working memory, processing speed, spatial abilities, and fluid intelligence. Applying two standard experimental paradigms allowed us to test multitasking abilities across five different empirical indices (i.e., performance costs) for both reaction time (RT) and accuracy measures, respectively. Multitasking resulted in substantial performance costs across all experimental conditions without a single significant gender difference in any of these ten measures, even when controlling for gender differences in underlying cognitive abilities. Thus, our results do not confirm the widespread stereotype that women are better at multitasking than men at least in the popular sequential and concurrent multitasking settings used in the present study.

Multitasking in aging: ERP correlates of dual-task costs in young versus low, intermediate, and high performing older adults.

With large inter-individual variability, older adults show a decline in cognitive performance in dual-task situations. Differences in attentional processes, working memory, response selection, and general speed of information processing have been discussed as potential sources of this decline and its between-subject variability. In comparison to young subjects (n = 36, mean age: 25 years), we analyzed the performance of a large group of healthy elderly subjects (n = 138, mean age: 70 years) in a conflicting dual-task situation (PRP paradigm). Based on their dual-task costs (DTCs), the older participants were clustered in three groups of high, medium, and low performing elderly. DTCs differed between groups and increased linearly from young subjects to low performing elderly. The groups did not differ with respect to ERP-components related to task preparation (CNV) and recall of stimulus-response mappings (P2). Peak latencies of the frontocentral P2 and N2 were shorter in young as compared to older adults but did not differ between elderly performance groups. However, differences in N2 amplitude between short and long SOA were correlated with the corresponding DTCs, suggesting more efficient S-R implementation in subjects with lower DTCs. Based on our results, between-subject differences in dual-task interference can be explained in terms of individual differences in selection of an appropriate response in dual-task situations.

Sources of interference in cross-modal action: response selection, crosstalk, and general dual-execution costs.

Performing several actions simultaneously usually yields interference, which is commonly explained by referring to theoretical concepts such as crosstalk and structural limitations associated with response selection. While most research focuses on dual-task scenarios (involving two independent tasks), we here study the role of response selection and crosstalk for the control of cross-modal response compounds (saccades and manual responses) triggered by a single stimulus. In two experiments, participants performed single responses and spatially compatible versus incompatible dual-response compounds (crosstalk manipulation) in conditions with or without response selection requirements (i.e., responses either changed randomly between trials or were constantly repeated within a block). The results showed that substantial crosstalk effects were only present when response (compound) selection was required, not when a pre-selected response compound was merely repeated throughout a block of trials. We suggest that cross-response crosstalk operates on the level of response selection (during the activation of response codes), not on the level of response execution (when participants can rely on pre-activated response codes). Furthermore, we observed substantial residual dual-response costs even when neither response incompatibility nor response selection requirements were present. This suggests additional general dual-execution interference that occurs on a late, execution-related processing stage and even for two responses in rather distinct (manual and oculomotor) output modules. Generally, the results emphasize the importance of considering oculomotor interference in theorizing on multiple-action control.

Action control in task switching: do action effects modulate N - 2 repetition costs in task switching?

Ideomotor theory posits that actions are controlled by the anticipation of their effects. In line with this theoretical framework, response-contingent action effects have been shown to influence performance in choice-reaction time tasks, both in single-task and task-switching context. Using a task-switching paradigm, the present study investigated whether task-contingent action effects influenced N - 2 repetition costs in task switching. N - 2 repetition costs are thought to be related to task-switch costs, and reflect inhibitory control in task switching. It was expected that task-contingent action effects reduce between-task interference, leading to reduced N - 2 repetition costs. An experimental group (N = 24) performed eight blocks of trials with task-contingent action effects, followed by one block with non-contingent action effects; a control group (N = 24) performed nine blocks of trials with non-contingent action effects. In line with our expectations, a three-way interaction of group, block, and task sequence was obtained, indicating differential data patterns for the two groups: In error rates, the group who had received contingent action effects throughout blocks 1-8 showed larger N - 2 repetition costs in the random block 9 than in block 8, whereas the control group showed a reversed data pattern. The RT data pattern was in the same direction, although no significant three-way interaction was obtained. Taken together, we tentatively conclude that task-contingent action effects reduce task inhibition in task switching, and we outline directions for future research on the role of action effects in multitasking performance.

"Optimal suppression" as a solution to the paradoxical cost of multitasking: examination of suppression specificity in task switching.

Switching between tasks necessitates maintaining tasks in high readiness, yet readiness creates paradoxical interference from these tasks when they are not currently required. "Optimal suppression", which targets just the interfering information, provides a partial solution to this paradox. By examining the carryover of suppression of a competitor stimulus-response (S-R) set from Trial N - 1 to Trial N, Meiran, Hsieh  and colleagues (Meiran  et al., J Exp Psychol Learn mem cognit 36:992-1002, 2010; Cognit Affect Behav Neurosci 11:292-308, 2011, and Hsieh et al., Acta Psychol 141:316-321, 2012) found that only the competing stimulus-response (S-R) set of rules is suppressed. Specifically, they found that a competitor S-R set in Trial N - 1 incurs cost when it becomes the relevant set in Trial N [competitor becomes relevant (CbR)]. Extending this logic, we predicted performance benefit when the competitor S-R set in Trial N - 1 remains the competitor S-R set in Trial N [competitor remains competitor (CrC)]. Here, we examined the question of whether what is being suppressed when encountering a response conflict is the entire S-R set of rules (e.g., "IF pink PRESS right", and "IF blue PRESS left") or an even more specific representation, namely, the currently interfering S-R rule (e.g., just "IF blue PRESS left"). We show that both CbR and CrC interact with Response (i.e., left or right key), suggesting that the system can recognize the exact source of interference (the competing S-R rule), and inhibit only this source.