Abstract
Can top-down control be used to overcome cross-dimensional attentional capture by a salient singleton? Here, we test an alternative to the hypothesis that training people to search for a specific target (feature search) eliminates capture by an irrelevant singleton. Specifically, we investigate whether feature-trained individuals are simply able to more efficiently disengage from an irrelevant singleton and reorient towards a target, compared to a group previously trained to look for a unique item (singleton search). In Experiment 1, we replicated the feature and singleton training methods used by Leber and Egeth (2006). In the test displays of our transfer session, participants were presented with 5 shapes (1 circle, 4 squares), enclosing differently oriented lines. Participants were instructed to identify the orientation of the line within the circle. This display was presented for 30, 60, 90, or 120 ms, and the lines appearing within the shapes either disappeared or were replaced by masks. Response accuracy was measured. On a subset of trials, one of the non-targets was a color-singleton distractor. We expected that at short display durations, feature- and singleton-trained individuals would perform less accurately when the distractor was present, due to insufficient time to disengage from the distractor and reorient towards the target. We found a difference in accuracy at 90 ms between distractor present and absent trials for both groups when a mask was presented. To determine whether this difference resulted from a lack of rapid disengagement and reorienting (and not a filtering cost), we conducted a similar investigation (Experiment 2) with feature-trained participants, in which the orientation of the line within the distractor matched the identity of the line appearing within the target on half of all trials. We observed a congruency effect at 60 ms, suggesting that feature-trained participants processed the information within the singleton distractor.
Meeting abstract presented at VSS 2014