Abstract
We recently demonstrated that contingent attentional capture costs are doubled or tripled under certain conditions when participants maintain more than one goal at a time (Moore & Weissman, 2010, 2011). Specifically, such "set-specific capture" occurs when participants must switch between multiple concurrent goals in a visual search task. Here, we found that set-specific capture effects are modulated by associative learning training. Participants searched for target letters presented in either of two colors (e.g. "orange" or "green") in a central rapid serial visual presentation stream, while ignoring distractors in the periphery. On same-set trials, a target-colored peripheral distractor appeared just prior to the central target of the same color. On set-switch trials, the peripheral distractor was a different target color (e.g. green) than the subsequent target (e.g. orange). Half of participants were trained to expect mostly same-set trials, and half were trained to expect mostly set-switch trials. Across all participants, target-colored distractors of both types impaired performance as compared to a no-distractor condition, replicating contingent attentional capture findings. Also, performance was worse in set-switch trials than same-set trials, replicating set-specific capture findings. Critically, these findings were modulated by training: participants with set-switch training showed greatly reduced (but still present) set-specific capture effects during a post-test phase, whereas those with same-set training showed dramatically increased set-specific capture effects. Thus, participants with set-switch training were able to anticipate a necessary set switch when they saw a target-colored peripheral distractor. Two follow-up experiments demonstrate that the training effect is due primarily to rapid and specific associative learning. Increased ability to task-switch more flexibly plays a lesser role.
Meeting abstract presented at VSS 2013