Abstract
The ability to accurately perceive and remember object features depends greatly on where attention is focused (Treisman, 1998). Past work has examined how top-down or bottom-up attentional guidance influences object feature reports. For example, when attention is divided between objects at two locations and participants later report the color of one of the objects, responses reflect some mixing of the objects’ colors, while when attention is shifted to or captured by another object, responses instead reflect some swapping of a distractor color (Golomb, L’Heureux, & Kanwisher, 2014; Chen, Leber, & Golomb, 2019). The present study sought to examine whether manipulating spatial attention with an experience-dependent cue would induce feature errors more similar to goal-directed or bottom-up attentional manipulations. We used a spatial probability cue (Geng & Behrmann, 2005) to guide attention to one of four locations that colored squares appeared in; the target was biased to appear in one of the locations (high-probability) most frequently. Participants reported the color of the target square on a continuous color wheel. Across several experiments we tested how the presentation of the target probe (simultaneous with or following the array) and a probabilistic pre-cue affected the impact of the spatial probability cue. Our results showed that a spatial probability cue does not have the same effect on feature response errors as either goal-directed or bottom-up guidance. Instead, participants tended to avoid the feature in the high-probability location when it was not the target; i.e., report a color repulsed away from the color of the high-probability distractor. Additional experiments examined if this feature avoidance emerged due to an attentional bias strong enough that only the high-probability item was encoded, and subsequently avoided during target report. Overall, we conclude that experience-driven attentional guidance has a unique impact on how color features are encoded and remembered.