Abstract
In visual search tasks, attention can be guided to items that share features (e.g. color) with the desired target. Is feature guidance available immediately from the onset of a stimulus? To address this question, observers reported whether the gap in a target C was on the left or right among distractor Cs with gaps facing up or down. Stimuli were presented within circular placeholders that changed color during the trial. Once colored, displays had 4 placeholders of one color and 12 of another. The target always appeared in the 4-item set. Performance was compared to two control conditions: 4 or 16 homogeneously colored items. We manipulated stimulus onset asynchrony (SOA) between the color cue and the Cs. If Cs appeared well before color, observers searched through 16 items at random. As SOA approached zero, color information became increasingly effective, as if observers searched initially through all 16 items before the color information made it possible to restrict search to just four items. Color guidance was not fully effective at zero SOA. Only when color preceded the Cs by 200–400 ms was guided search equivalent to search through just four items. This was true even when the target color was constant for an entire block. When different target colors were randomly intermixed during a block, guidance took up to 600 ms to become fully effective, suggesting a cost for changing guidance settings. When the target color became a distractor color on the next trial, RTs were slowed relative to the randomly intermixed condition at zero ms SOA, but these conditions became equivalent by 600 ms. This suggests that observers continued to guide by the previous color until guidance to the new color came online. Feature guidance is not available immediately.
NRSA grant #F32 EY016632 to EMP