Auditory spatial attention assessed in a flanker interference task

We report three experiments designed to investigate interference effects in an auditory analog of Eriksen's visual flanker interference task (e.g. [1, 2]). In Experiment 1, participants made speeded auditory 'bat' versus 'bed' discrimination responses to a series of target words presented from a central loudspeaker while simultaneously trying to ignore auditory flanker words ('bat' or 'bed') presented from two loudspeakers positioned 30 degrees to either side of fixation. Participants responded more slowly on trials where the flankers were incongruent with the target than when they were congruent. Experiment 2 revealed that the magnitude of this auditory congruency effect was relatively unaffected by increasing the azimuthal separation between the target and flanker stimuli from 30 to 90 degrees. In Experiment 3, the flanker words were always different from the target word, no matter whether they were congruent or incongruent with the target response. Despite the fact that any energetic masking effects from congruent and incongruent flankers should have been normalized under such conditions, a congruency effect was still reported, supporting an informational masking account of the flanker effect. Our results therefore provide the first empirical demonstration of an auditory analog of Eriksen's visual flanker interference effect.