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.     

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.     

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.     

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.     

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.     

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.     

Consciousness and control in task switching

Participants were required to switch among randomly ordered tasks, and instructional cues were used to indicate which task to execute. In Experiments 1 and 2, the participants indicated their readiness for the task switch before they received the target stimulus; thus, each trial was associated with two primary dependent measures: (1) readiness time and (2) target reaction time. Slow readiness responses and instructions emphasizing high readiness were paradoxically accompanied by slow target reaction time. Moreover, the effect of task switching on readiness time was an order of magnitude smaller then the (objectively estimated) duration required for task preparation (Experiment 3). The results strongly suggest that participants have little conscious awareness of their preparedness and challenge commonly accepted assumptions concerning the role of consciousness in cognitive control.     

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.     

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.     

Sequential dependencies in the lexical decision task

Two experiments addressed whether response latency in a trial of the lexical decision task is independent of the lexical status of the item presented in the previous trial. In Exp. 1, it was found that both word and nonword responses were significantly slower when the previous trial had involved a nonword than when it had involved a word. In Exp. 2, which employed a different list composition, it was found that responses to nonwords and pseudohomophones were significantly slower when the previous trial had involved a nonword or a pseudohomophone than when it had involved a word. However, responses to words were not influenced by the nature of the previous trial. We concluded that sequential dependencies exist across consecutive trials in the lexical decision task even when there is no semantic, morphological, phonological, or orthographic relationship between the items presented during those trials.     

Components of competitor priming in task switching

Executing an action in response to a stimulus is thought to result in the creation of an event code that integrates stimulus and action features (Allport, 1987; Hommel in Visual Cognition 5: 183-216, 1998). When switching between tasks, competitor priming occurs if a distractor stimulus cues the retrieval of a previously established event code in which that distractor is bound to a competing task, creating a source of interference with the current task whereby the observer is encouraged to apply the competing task to the distractor. We propose a second aspect of competitor priming: the misapplication of the retrieved competing task to the target stimulus. We report two task-switching experiments in which tasks applied to picture-word compound stimuli were manipulated to create conditions in which this second aspect of competitor priming could be revealed and distinguished from other sources of task- and stimulus-based priming. A substantial increase in competitor priming was observed when subjects switched between tasks that required very different processing operations and the competing task was highly relevant to the target stimulus. These results are consistent with our claim that competitor priming can result from applying the competing task either to the distractor that cued it or to the target stimulus.     

Training and transfer effects in task switching

Performance on task switching, a paradigm commonly used to measure executive function, has been shown to improve with practice. However, no study has tested whether these benefits are specific to the tasks learned or are transferable to new situations. We report evidence of transferable improvement in a cued, randomly switching paradigm as measured by mixing cost, but we report no consistent improvement for switch cost. Improvement in mixing costs arises from a relative reduction in time to perform both switch and nonswitch trials that immediately follow switch trials, implicating the ability to recover from unexpected switches as the source of improvement. These results add to a growing number of studies demonstrating generalizable improvement with training on executive processing.     

Modulation of word-reading processes in task switching

The authors examined modulation of the simple act of word naming induced by the conflict arising when that task competes with color naming in a task-switching paradigm. Subjects alternated between naming a word printed in black and naming the color of a stimulus in 2 conditions. In the incongruent condition, the colored stimulus was an irrelevant word generating conflict, and in the neutral condition, color was carried by a row of asterisks. Subjects took substantially longer to name a word printed in black in the incongruent condition, implying a form of suppression. This modulation of the word-naming response was adaptive in that it led to more efficient color naming. The modulation effect was replicated using phoneme detection instead of word naming but not with lexical decision or visual comparison, implicating a phonological encoding process.     

Hierarchical switching in a multi-dimensional task space

We explored shift costs for a dimensionally organized set of tasks. Task dimensions were type of judgement (numerical vs. spatial) and judgement-to-response mapping (compatible vs. incompatible). Shift costs were determined as the difference in RTs between switch trials and repetitions in a situation in which the kind of task was unpredictable. Shift costs were greatest when the type of judgement was changed. A change in the mapping increased shift costs when the type of judgement remained unchanged, but reduced shift costs when the type of judgement was changed as well. Also, response alternations produced costs when both the type of judgement and SR mapping were unchanged, but reduced shift costs otherwise. In addition to the relation between successive tasks, shift costs were modulated by the specifics of the elementary tasks. The major characteristics of the pattern of shift costs could be accounted for by the hypothesis of a dimensionally organized task space in which hierarchically organized, relative switching operations are performed that affect the highest-level task dimension at which a change is required, and all lower-level dimensions in a first step. In a second step, lower-level dimensions are switched back when no change is required, and then the selected control structure is implemented. Received: 11 December 1997 / Accepted: 6 July 1998     

Inhibition of response mode in task switching

Task inhibition was explored in two experiments that employed a paradigm in which participants switched among three tasks. Two tasks required manual choice responses based on numerical judgment (parity or magnitude), whereas a third task required an unconditional double-press of both response keys. Both experiments showed that switching to a just-abandoned task (n-2 task repetition) generally leads to a performance cost relative to switching to another task. Specifically, this task inhibition effect also occurred for the double-press task, suggesting inhibition of response mode. Prolonging the task-cuing interval showed that advance task preparation reduced only inhibition of the double-press task but not of the choice tasks (Experiment 1). Prolonging the response-cue interval led to a decrease of the inhibition effect in all tasks (Experiment 2), suggesting a time-based release of task inhibition. Together, the experiments support the notion of a response-related component of task inhibition.     

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.