Academic journal article Attention, Perception and Psychophysics

Top-Down Deactivation of Interference from Irrelevant Spatial or Verbal Stimulus Features

Academic journal article Attention, Perception and Psychophysics

Top-Down Deactivation of Interference from Irrelevant Spatial or Verbal Stimulus Features

Article excerpt

Published online: 1 My 2014

(Q> The Psychonomic Society, Inc. 2014

Abstract The selective-attention model of Houghton and Tipper (1994) assumes top-down deactivation of (conflicting) distractor representations as a mechanism of visual attention. Deactivation should produce an inverted-U- shaped activation function for distractor representations. In a recent study, Frings, Wentura, and Wiihr (2012) tested this prediction in a variant of the flanker task in which a cue sometimes required participants to respond to the distractors rather than to the target. When reaction times and error rates were plotted as a function of the target-cue stimulus onset asynchrony, a quadratic trend emerged, consistent with the notion of distractor deactivation. However, in the flanker task, an alternative explanation for the quadratic trend in terms of attentional zooming is possible. The present experiments tested the deactivation account against the attentional-zooming account with the Stroop and the Simon task, in which attentional zooming should have minimal effects on distractor processing, because the target and distractor are presented at the same spatial location. Both experiments replicated the quadratic trend in the performance functions for responses to incongruent distractors, and additionally showed linear trends in the performance functions for responses to congruent distractors. These results provide additional support for the notion of top-down deactivation of distractor representations as a mechanism of visual selective attention.

Keywords Visual selective attention * Deactivation * Cue * Simon task * Stroop task

Human action is usually directed toward a small subset of those objects that are simultaneously present in our environ- ment. In order to support the control of actions, selective attention is assumed to facilitate processing of the action- relevant features of action-relevant objects (e.g., Allport, 1987; Neumann, 1987). Whereas it is widely agreed that attention facilitates the processing of relevant visual informa- tion by activation processes (e.g., Pashler, 1998), it is less clear how attention deals with the representation of irrelevant distractor stimuli. For example, several selection models as- sume that distractor representations become inhibited, at least for a short period of time (e.g., Houghton & Tipper, 1994; Schrobsdorff et al., 2007; Treisman, 1960). However, both the existence and the possible characteristics of attentional inhi- bition are issues of considerable debate (e.g., Dagenbach & Carr, 1994; Gorfein & MacLeod, 2007; MacLeod, Dodd, Sheard, Wilson, & Bibi, 2003). In this article, we apply a new method for measuring the processing of irrelevant fea- tures in different selective-attention tasks (see Frings, Wentura, & Wiihr, 2012). The basic idea follows the selective-attention model of Houghton and Tipper (1994), which offers a general approach for understanding selection between relevant and irrelevant information.

At the core of the selective-attention model of Houghton and Tipper (1994, 1996) are the activation values of cognitive codes representing the features of target and distractor stimuli. These stimulus codes receive activation and deactivation from different sources. The presence of the target and the distractor stimulus in the environment provides bottom-up activation for the respective codes. Moreover, the stimulus codes that match a stored template of target features receive additional activa- tion from top-down sources, whereas the nonmatching stim- ulus codes receive top-down deactivation. As a result, external and internal sources will increase the activation value of the target codes, whereas external activation and internal deactivation are assumed to keep the activation value of distractor codes at resting level. A decision or response is triggered when the difference between the activation levels of target and distractor representations exceeds a threshold value. …

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

Oops!

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.