Academic journal article Psychonomic Bulletin & Review

Proactive versus Reactive Task-Set Inhibition: Evidence from Flanker Compatibility Effects

Academic journal article Psychonomic Bulletin & Review

Proactive versus Reactive Task-Set Inhibition: Evidence from Flanker Compatibility Effects

Article excerpt

Two converging tests were used to determine whether people proactively inhibit recently performed tasks when switching to new tasks. A task-cuing paradigm was used. In each trial, the relevant stimulus was accompanied by flankers belonging either to the task performed on the immediately preceding trial (lag 1) or a more distant trial (lag 2+). If the just-performed task is inhibited when switching to another task, and this inhibition declines across trials, then flanker interference should be smaller with lag 1 flankers than with lag 2+ flankers. Experiment 1, following the methods of Hübner, Dreisbach, Haider, and Kluwe (2003), failed to confirm this prediction. The prediction was confirmed in Experiment 2, however, using a design modified to provide greater incentives for task-set inhibition. The results provide evidence that inhibition can be applied proactively, to reduce the ability of an abandoned task to interfere with the performance of other tasks.

To better understand how the mind controls itself, researchers have utilized the task-switching paradigm. In one variant, known as the alternating-runs paradigm, participants receive a repeating task sequence, such as AABB throughout a block (e.g., Rogers & Monsell, 1995). In another popular variant, the task sequence is unpredictable but a task cue (often an explicit task name) indicates which task should be performed next (e.g., Meiran, 1996). Regardless of the paradigm, performance is slower and less accurate following a task switch. This switch cost is especially large when each stimulus is associated with two active tasks (bivalent), rather than just one (univalent).

Switch costs have been attributed to the carryover of a previously activated task set (called task-set inertia; see, e.g., Allport, Styles, & Hsieh, 1994), and the time needed to configure a new task (called task-set reconfiguration; see, e.g., Rogers & Monsell, 1995; Ruthruff, Remington, & Johnston, 2001). Another line of debate concerns whether switch costs reflect an inability to complete task-set preparation in advance of the stimulus (see, e.g., Lien, Ruthruff, Remington, & Johnston, 2005) or occasional failures to engage in preparation (see, e.g., De Jong, 2000). In the present study, we focus on another mechanism hypothesized to contribute to task-switch costs-namely, task-set inhibition (see, e.g., Mayr & Keele, 2000). Such inhibition would facilitate the immediate switch to a new task, but might impede a subsequent switch back to the inhibited task.

n-2 Repetition Costs As an Index of Task-Set Inhibition

To look for evidence of task-set inhibition, Mayr and Keele (2000) examined task switch performance as a function of how recently the task was performed (see also e.g., Arbuthnott, 2005; Schuch & Koch, 2003). The critical prediction is that response time (RT) should be longer for more recently performed tasks, because the inhibition of that task would have had less opportunity to decay (but see Gade & Koch, 2005). They compared performance on the third trial of an ABA task sequence (where inhibition of Task A is recent) versus the third trial of a CBA task sequence (where inhibition of Task A is less recent). Hereafter, we refer to these conditions as the n-2 repetition and n-2 switch conditions, respectively. Many studies have reported an n-2 repetition cost: slower responses in the n-2 repetition condition than the n-2 switch condition. The cost is somewhat counterintuitive, in that repetition generally produces a benefit, but makes perfect sense in light of task-set inhibition theory.

Is Inhibition Applied Proactively or Reactively?

Mayr and Keele (2000) proposed that inhibition is applied to a task proactively, just after it has been performed, to reduce interference from that task on a subsequent switch trial. In support of this claim, they reported n -2 repetition costs in an ABA sequence even when there was no Task A distractor to inhibit (reactively) in the intervening trial. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.