Residual costs in task switching: testing the failure-to-engage hypothesis

Reaction time is typically longer on trials on which the task changes. Thisswitch cost is reduced by the opportunity to prepare for the change before the stimulus onset, but there remains aresidual cost that resists reduction by further opportunity for preparation. De Jong (2000) proposed a model for evaluating the contribution to the residual cost of (1) failure to achieve endogenous task-set reconfiguration on a proportion of trials, and (2) limitations to the completeness of reconfiguration attainable by endogenous means. We report good fits of the model to the data from one previous and one new task-switching experiment, suggesting that the residual switch cost may indeed be attributable to a probabilistic failure to complete advance preparation. But strong incentives for preparation only marginally increased the estimated preparation probability, suggesting some intrinsic limitation to the ability to achieve endogenous preparation for a task switch on every trial.     

Components of task switching: A closer look at task switching and cue switching

Research using the diffusion model to decompose task-switching effects has contributed to a better understanding of the processes underlying the observed effect in the explicit task cueing paradigm: Previous findings could be reconciled with multiple component models of task switching or with an account on compound-cue retrieval/repetition priming. In the present study, we used two cues for each task in order to decompose task-switch and cue-switch effects. Response time data support previous findings that comparable parts of the switching effect can be attributed to cue-switching and task-switching. A diffusion model analysis of the data confirmed that non-decision time is increased and drift rates are decreased in unpredicted task-switches. Importantly, it was shown that non-decision time was selectively increased in task-switching trials but not in cue-switching trials. Results of the present study specifically support the notion of additional processes in task-switches and can be reconciled with broader multiple component accounts.     

On the representation of task information in task switching: Evidence from task and dimension switching

Task switching research has revealed that task changes lead to a performance switch cost. The present study focuses on the organization of task components in the task set. Three different views of task set organization have been distinguished and evidence in favor of each of these has been reported in the literature. In four experiments, we orthogonally varied the categorization task (magnitude and parity) and the stimulus dimension on which the categorization was to be made. Experiments 1, 2, and 4 used Stroop-like number stimuli, whereas Experiment 3 used global-local stimuli to define the stimulus dimension. In Experiments 2-4, the cue-stimulus interval was also varied. The findings showed that a change of any component resulted in a cost, without any reliable difference in the size of these costs. These results are consistent with the flat view on task-set organization, which assumes that the task set binds all elements in an unstructured representation, which is completely reconfigured each time a change to the task set is required. The implications of these findings are discussed in relation to other findings and the different views on task-set organization.     

Sequential task predictability in task switching

Many studies of task switching have found that a prolonged preparation time reduces switch costs. An alternative manipulation of task preparation is based on sequential task predictability, rather than preparation time. In Experiments 1 and 2 of the present study, participants performed explicitly instructed task sequences (i.e., AABB) and were then transferred to a random sequence. The observed benefit of predictability-based task preparation was not switch specific. In Experiment 3, the participants changed from random to predictable tasks. The observed predictability benefit again was not switch specific. The data thus suggest that task switching does not necessarily require a switch-specific reconfiguration process. Rather, task-specific control processes may be needed in both task switches and repetitions.     

A chain-retrieval model for voluntary task switching

To account for the findings obtained in voluntary task switching, this article describes and tests the chain-retrieval model. This model postulates that voluntary task selection involves retrieval of task information from long-term memory, which is then used to guide task selection and task execution. The model assumes that the retrieved information consists of acquired sequences (or chains) of tasks, that selection may be biased towards chains containing more task repetitions and that bottom-up triggered repetitions may overrule the intended task. To test this model, four experiments are reported. In Studies 1 and 2, sequences of task choices and the corresponding transition sequences (task repetitions or switches) were analyzed with the help of dependency statistics. The free parameters of the chain-retrieval model were estimated on the observed task sequences and these estimates were used to predict autocorrelations of tasks and transitions. In Studies 3 and 4, sequences of hand choices and their transitions were analyzed similarly. In all studies, the chain-retrieval model yielded better fits and predictions than statistical models of event choice. In applications to voluntary task switching (Studies 1 and 2), all three parameters of the model were needed to account for the data. When no task switching was required (Studies 3 and 4), the chain-retrieval model could account for the data with one or two parameters clamped to a neutral value. Implications for our understanding of voluntary task selection and broader theoretical implications are discussed.     

Instruction effects in task switching

The present study examined the effect of instructions on sequential task preparation using a cuing paradigm with three tasks. All task transitions were predictable, whereas task identity was unpredictable in switches but predictable in repetitions. In Experiment 1, predictability (predictable vs. random) was manipulated while preparation time (i.e., the cue-stimulus interval, or CSI) remained constantly short. In Experiment 2, CSI was manipulated for predictable task transitions. Both experiments showed clear instruction effects, but these were restricted to predictable task repetitions, for which predictability determined the identity of the upcoming task. Predictability effects were small in task switches and were not modulated by instruction, suggesting that preparation is mainly task-specific rather than switch-specific. Together, these results suggest that intentional processes contribute to predictability benefits in task repetitions, probably by enhancing the monitoring of sequential transitions in working memory in order to maintain activation in task repetitions.     

Stimulus-related priming during task switching

Task switch cost (the deficit of performing a new task vs. a repeated task) has been partly attributed to priming of the repeated task, as well as to inappropriate preparation for the switched task. In the present study, we examined the nature of the priming effect by repeating stimulus-related processes, such as stimulus encoding or stimulus identification. We adopted a partial-overlap task-switching paradigm, in which only stimulus-related processes should be repeated or switched. The switch cost in this partial-overlap condition was smaller than the cost in the full-overlap condition, in which the task overlap involved more than stimulus processing, indicating that priming of a stimulus is a component of a switch cost. The switch cost in the partial-overlap condition, however, disappeared eventually with a long interval between two tasks, whereas the cost in the full-overlap condition remained significant. Moreover, the switch cost, in general, did not interact with foreknowledge, suggesting that preparation on the basis of foreknowledge may be related to processes beyond stimulus encoding. These results suggest that stimulus-related priming is automatic and short-lived and, therefore, is not a part of the persisting portion of switch cost.     

Component processes in task switching

Participants switched between two randomly ordered, two-choice reaction-time (RT) tasks, where an instructional cue preceded the target stimulus and indicated which task to execute. Task-switching cost dissipated passively while the participants waited for the instructional cue in order to know which task to execute (during the Response-Cue Interval). Switching cost was sharply reduced, but not abolished, when the participants actively prepared for the task switch in response to the instructional cue (during the Cue-Target Interval). The preparation for a task switch has shown not to be a by-product of general preparation by phasic alertness or predicting target onset. It is suggested that task-switching cost has at least three components reflecting (1) the passive dissipation of the previous task set, (2) the preparation of the new task set, and (3) a residual component.     

Cue-task associations in task switching

Cognitive flexibility can be studied using the task-switching paradigm. This paradigm requires subjects to adapt behaviour to changing contexts as indicated by a cue. In our study, we addressed the question of how cue-based implementation of mental “task sets” occurs. We assumed that cues build up associations to the tasks that they indicate. These associations lead to retrieval of the associated task set once the cue shows up again. In three experiments, we tested this assumption using a negative transfer paradigm. First participants were exposed to one cue-task mapping. After a training phase, the cue-task mapping changed in either of two ways. Whereas one group of participants got new cues, the other experienced a reversal of the learnt cue-task mapping. Our results show that participants build up cue-task associations and that these formerly learnt associations can hamper the implementation of new cue-task mappings (particular with mapping reversal). Prolonged preparation time decreased the cost of changing the cue-task mapping but did not change the overall pattern of results.     

The shielding function of task rules in the context of task switching

There is increasing evidence that task rules help shield the response against distractor interference. Here, the authors investigate the cognitive mechanisms underlying this assumed shielding function of task rules and how it is adjusted to changing task demands. In two experiments, participants switched between a noun categorization and an adjective categorization task. Target words were superimposed on distractor pictures. These pictures were always irrelevant and depicted either objects also used as target words in the noun task (noun distractors) or objects that were not part of the noun target-set but could be categorized according to the noun task (noun-related distractors). Results show that (a) on task repetitions shielding prevents interference from any distractors associated with a competing task; this is indicated by the lack of interference on adjective task repetitions; and (b) shielding is reduced on task switches. In the noun task, this reduction resulted in attenuated interference by noun-related distractors. In the adjective task, spatial distractors did not interfere despite the reduction. This result suggests that shielding is supported by a processing advantage for task-related information and not by distractor suppression.     

Effects of response selection on the task repetition benefit in task switching

A task switch typically leads to worse performance than a repetition does. This shift cost can be reduced with sufficient task preparation time, but a residual cost usually remains. We propose that a large part of this residual cost is caused by an activation bias produced by response selection processes in the preceding trial. In our experiments, we manipulated response selection requirements using a go/no-go methodology. The residual shift cost disappeared after no-go trials, suggesting that response selection is crucial to establish an activation bias for the current category-response rules and that this bias persists into the next trial. A comparison with a go-only group confirmed this analysis by revealing no differences in preparatory strategy due to the inclusion of no-go trials. In addition, no-go trials had no significant effects on subsequent trials in a single-task experiment, suggesting that no-go trials are not coded as a task different from go trials and that there is no inhibition of the prepared task in a no-go trial. We thus conclude that a persisting activation bias of response rules plays a major role in task switching.     

Working memory costs of task switching

Although many accounts of task switching emphasize the importance of working memory as a substantial source of the switch cost, there is a lack of evidence demonstrating that task switching actually places additional demands on working memory. The present study addressed this issue by implementing task switching in continuous complex span tasks with strictly controlled time parameters. A series of 4 experiments demonstrate that recall performance decreased as a function of the number of task switches and that the concurrent load of item maintenance had no influence on task switching. These results indicate that task switching induces a cost on working memory functioning. Implications for theories of task switching, working memory, and resource sharing are addressed.     

Exogenous influences on task set activation in task switching

To explore the effect of exogenous processes on cognitive control, we used a cueing task-switching paradigm with two spatial judgement tasks and added an irrelevant colour attribute to the task-relevant spatial attribute of the target. The colour was not related to any specific Stimulus-Response relation in the tasks. A correlation was created between stimulus colour and task identity. This correlation was strong but imperfect in Experiment 1 and perfect in Experiment 2. As a result of the colour-task correlation, stimuli contained redundant information about task identity. By changing the correlation pattern every few blocks we caused this information to be sometimes invalid. In both experiments, performance was worse when the information carried by the target was invalid than when it was valid. However, this effect was exclusive to conditions with short task preparation time. By comparing performance with a control group, which had no colour-task correlation (in Experiment 2) we established that the colour manipulation did not cause a qualitative change in preparation strategy, and that the exogenous effect was stronger in switch trials than in repetition trials. We conclude that exogenous processes that are related to task set affect performance primarily if they are presented before endogenous processes of task set preparation have been launched.     

Generic cognitive adaptations to task interference in task switching

The present study investigated how the activation of previous tasks interferes with the execution of future tasks as a result of temporal manipulations. Color and shape matching tasks were organized in runs of two trials each. The tasks were specified by a cue presented before a task run, cueing only the first trials of each run. Response times (RTs) and error rates were measured for task switching and task repetition conditions. Task interference was varied as a function of response-cue interval (RCI of 300 and 900 ms), that is, the interval between the task runs. Keeping the response-stimulus interval within the task runs constant at 300 ms allowed the disentangling of the direct effects of RCI manipulation on performance (first trials) from the general effects on performance (both trials in the run). The data showed similar performance improvement due to RCI increase on both trials in the task run. Furthermore, increasing RCI improved both switch and repetition performance to a similar extent. Together, our findings provide further evidence for accounts stressing generic effects of proactive task interference in task switching.     

Exogenous influences on task set activation in task switching

To explore the effect of exogenous processes on cognitive control, we used a cueing task-switching paradigm with two spatial judgement tasks and added an irrelevant colour attribute to the task-relevant spatial attribute of the target. The colour was not related to any specific Stimulus–Response relation in the tasks. A correlation was created between stimulus colour and task identity. This correlation was strong but imperfect in Experiment 1 and perfect in Experiment 2. As a result of the colour–task correlation, stimuli contained redundant information about task identity. By changing the correlation pattern every few blocks we caused this information to be sometimes invalid. In both experiments, performance was worse when the information carried by the target was invalid than when it was valid. However, this effect was exclusive to conditions with short task preparation time. By comparing performance with a control group, which had no colour–task correlation (in Experiment 2) we established that the colour manipulation did not cause a qualitative change in preparation strategy, and that the exogenous effect was stronger in switch trials than in repetition trials. We conclude that exogenous processes that are related to task set affect performance primarily if they are presented before endogenous processes of task set preparation have been launched.     

The role of inhibition in task switching: A review

The concept of inhibition plays a major role in cognitive psychology. In the present article, we review the evidence for the inhibition of task sets. In the first part, we critically discuss empirical findings of task inhibition from studies that applied variants of the task-switching methodology and argue that most of these findings— such as switch cost asymmetries—are ambiguous. In the second part, we focus on n-22 task-repetition costs, which currently constitute the most convincing evidence for inhibition of task sets. n-22 repetition costs refer to the performance impairment in sequences of the ABA type relative to CBA, which can be interpreted in terms of persisting inhibition of previously abandoned tasks. The available evidence suggests that inhibition is primarily triggered by conflict at selection of stimulus attributes and at the response level. Author Note     

Effects of anxiety on task switching: Evidence from the mixed antisaccade task

According to the attentional control theory of anxiety (Eysenck, Derakshan, Santos, & Calvo, 2007), anxiety impairs performance on cognitive tasks that involve the shifting function of working memory. This hypothesis was tested using a mixed antisaccade paradigm, in which participants performed single-task and mixed-task versions of the paradigm. The single task involved the completion of separate blocks of anti- and prosaccade trials, whereas in the mixed task, participants completed anti- and prosaccade trials in a random order within blocks. Analysis of switch costs showed that high-anxious individuals did not exhibit the commonly reported paradoxical improvement in saccade latency, whereas low-anxious individuals did. The findings are discussed within the framework of attentional control theory.     

Adult age differences in task switching

Age differences in 2 components of task-set switching speed were investigated in 118 adults aged 20 to 80 years using task-set homogeneous (e.g., AAAA ...) and task-set heterogeneous (e.g., AABBAABB ... ) blocks. General switch costs were defined as latency differences between heterogeneous and homogeneous blocks. whereas specific switch costs were defined as differences between switch and nonswitch trials within heterogeneous blocks. Both types of costs generalized over verbal, figural, and numeric stimulus materials; were more highly correlated to fluid than to crystallized abilities; and were not eliminated after 6 sessions of practice, indicating that they reflect basic and domain-general aspects of cognitive control. Most important, age-associated increments in costs were significantly greater for general than for specific switch costs, suggesting that the ability to efficiently maintain and coordinate 2 alternating task sets in working memory instead of 1 is more negatively affected by advancing age than the ability to execute the task switch itself.     

Task conflict effect in task switching

A part of action preparation is deciding what the relevant task is. This task-decision process is conceptually separate from response selection. To show this, the authors manipulated task conflict in a spatial task-switching paradigm, using conflict stimuli that appeared during trials with univalent targets (affording 1 task). The conflict stimuli afforded task identity because they were used as task cues with bivalent targets (affording 2 tasks) that were intermixed with the univalent targets. Thus, for univalent targets, irrelevant stimuli either caused low task conflict or high task conflict. In three experiments, the authors found poorer performance in high task conflict trials than in low task conflict trials. Task conflict was introduced during target appearance (Experiment 1) or task preparation (Experiments 2 and 3). In the latter case, the task conflict effect decreased with increasing task preparation time showing that task preparation involves task decision.