Abstract
How does past experience guide visual search strategy (i.e., attentional set)? Leber and Egeth (in press) found that observers given the option to use more than one attentional set in a “test phase” consistently used the set they established in a directly preceding “training phase”. While this result demonstrates a clear impact of past experience on attentional set, the question remains whether it reflects simple persistence with pre-established set or the formation of lasting associations that can be reactivated after intervening tasks. To examine these alternatives, and to probe the strength and duration of the learning, we introduced one-day and one-week delays between training and test. In a 320-trial training phase, observers assigned to a “singleton search” group were required to search for oddball color targets (selected randomly on each trial) in a rapid serial visual presentation of letters, while those assigned to a “feature search” group were required to search for targets of a specific, consistent color. In the 320-trial test phase, both groups of observers searched for a consistent color that was also a singleton; thus, both singleton and feature search strategies had become available to all observers. Test phase results revealed that observers in both one-day and one-week delay conditions used the same strategy that was required of them in the training phase. This long-term learning effect, arising from relatively short training (∼30 minutes), could not have resulted from simple persistence; rather, attentional set can be reactivated when triggered by the appropriate task context.