Abstract
Irrelevant distracters capture attention to a greater degree when they share a feature (e.g., the color red) with an attentional set that defines targets (e.g., “red letters”) than when they do not possess such a target-defining feature, a phenomenon known as contingent attentional capture. However, it is unclear whether the magnitude of such capture varies with whether distracters and targets possess features that match the same or different attentional sets. Given prior studies indicating that set switching is a time-consuming process, we hypothesized that increasing attention to a distracter whose color matches one attentional set (e.g., the color red) would make it harder to detect a subsequent target whose color matches a different (versus the same) attentional set (e.g., the color green). In Experiment 1, we asked participants to search for target letters presented in either of two colors (e.g. “red” or “green”) within a central rapid serial visual presentation (RSVP) stream, while ignoring distracters in two peripheral streams. Because the central RSVP stream also contained letters whose colors were task-irrelevant (e.g., purple, blue, orange, etc.), participants needed to actively maintain distinct attentional sets for the two task-relevant colors (e.g., “red” and “green”) in working memory. As predicted, target detection performance in the central RSVP stream was worse when the target's color (e.g., red) matched a different (versus the same) attentional set than the distracter's color (e.g., green). In Experiment 2, this color switching effect was eliminated when all colors were grouped into the same attentional set (i.e., when subjects were instructed to identify letters of any color among grey items). These findings demonstrate that the need to switch between distinct attentional sets enhances contingent attentional capture effects, and extend prior findings from the attentional blink, task switching, and working memory literatures indicating that set switching incurs behavioral costs.
We thank Anna Grummon, Erika Pinsker, and Melanie Sottile for their help on this project.