Academic journal article Memory & Cognition

Selective Attention and Response Set in the Stroop Task

Academic journal article Memory & Cognition

Selective Attention and Response Set in the Stroop Task

Article excerpt

Response set membership contributes much to the interference in the color-word Stroop task. This may be due to selective allocation of attention to eligible responses or, alternatively, to greater inhibition of distractors that are not responses. In the present article, we report two experiments that were designed to adjudicate between these accounts. In Experiment 1, membership was manipulated on a trial-by-trial basis by cuing the possible responses for each trial. Response time (RT) was longer for distractors that corresponded to a cued, eligible response than to an ineligible one. This cuing effect was independent of the number of different responses. In Experiment 2, the distractor was cued on half the trials. Cuing the distractor decreased RTs on both incongruent and congruent trials. Vincentile analyses in both experiments revealed that the effects were constant throughout the entire RT distributions. These results suggest that response set effects arise because of selective allocation of attention to eligible responses.

Since the inception of modern attention research in the 1950s (e.g., Broadbent, 1958), the issue of whether attentional selection is early or late in perceptually based responding has played a central role in experimental research (for reviews, see, e.g., Pashler, 1998; Styles, 2006). By the 1970s, it was generally accepted that attentional selectivity may be early or late, depending on the prevailing circumstances. This view was exemplified by Broadbent's (1970, 1971; Broadbent & Gregory, 1964) distinction between "stimulus set" and "response set." Stimulus set refers to selection on the basis of a perceptual attribute, such as spatial location, color, shape, or temporal order. Response set refers to selection on the basis of the vocabulary of eligible responses. Broadbent recognized that task performance may require one or both of these kinds of selective attention, depending on the situation. Whereas attention research in the past few decades has intensively investigated aspects of stimulus set (e.g., Pashler, 1998; Styles, 2006), response set has been much less examined.

One of the tasks that has demonstrated the importance of response set is the color-word Stroop task (Stroop, 1935). In a common version of this task, participants have to name the ink color of congruent or incongruent color words (e.g., the words green or red printed in green ink). Response time (RT) is typically longer in the incongruent than in the congruent condition (see MacLeod, 1991, for a review)-a phenomenon henceforth called the Stroop effect. A major part of the interference caused by incongruent stimuli is specific to the members of the response set. In a classic study, Klein (1964) observed that color words that were eligible responses produced approximately two times more interference than did color words that were not used as responses in the experiment. For example, if the ink colors were red and green, color-naming responses were much slower for the word green in red ink than for the word blue in red ink. This effect of response set has been replicated in several studies (e.g., Glaser & Glaser, 1989; Proctor, 1978).

In the present article, we examined two different explanations for the response set effect. First, the response set effect may be due to selective allocation of attention to eligible responses at the response selection level (see Broadbent, 1970, 1971; Deutsch, 1977). In a prominent model of the Stroop task that was developed by Cohen, Dunbar, and McClelland (1990), attention to particular responses is achieved by placing eligible responses in a more responsive part of their activation curve. Likewise, in the WEAVER++ model of Roelofs (1992, 2003), selective allocation of attention at the response level is achieved by restricting the "selection space" to eligible responses. Only responses within this selection space are competing for selection. In this model, words that do not correspond to eligible responses may still yield some interference because they activate alternative responses in the selection space via conceptual links, but this interference is less than that for words that are part of the response set. …

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.