Does Contextual Cueing Guide the Deployment of Attention?

Melina, A. Kunar 1,2, Stephen J. Flusberg 2, Todd S. Horowitz 1,2, Jeremy M. Wolfe 1,2

1 - Harvard Medical School
2 - Brigham and Women's Hospital

Contextual cueing (CC) experiments show that when visual search displays are repeated, reaction times (RTs) to find a target decrease over time even when observers are not aware of the repetition. In other experiments, observers use information about features, like color, to guide attention to likely target locations. Do observers use implicit memory for the display to guide attention in the same manner in CC tasks? We compared CC effects to standard guidance. Firstly, guidance decreases the slope of RT x Set Size functions. When guidance is perfect (e.g. feature search for a red target among green distractors) slopes will be close to 0 msec/item. In contrast, contextually repeated trials showed little reduction in slope compared to trials that did not repeat context. Secondly, observers can use color to guide attention to multiple target locations. However, with the same training, observers could not prioritize more than two target locations in CC. Thirdly, in standard guided search there is little reduction in guidance if the location of a target on one trial is occupied by a distractor on the next. However, when a distractor appeared in a prior target location in a CC paradigm, a CC benefit was not found. Finally, even with perfect guidance (e.g., in a feature search), we found a small, but reliable additional CC effect. This suggests a contribution from other factors, such as response selection. Overall, the data suggest that the relationship between guidance and CC is weak.