The transfer of learned distractor suppression: on the role of spatial configurations

It is well established that participants learn to suppress locations that more frequently contain distracting information through a process known as statistical learning. The present study examined the flexibility of this learned suppression using different stimulus configurations. Participants performed the additional singleton task, in which they searched for a unique target shape while ignoring a salient distractor singleton. The experiments included both biased and uniform search displays that differed in overall configuration, either by varying the spatial positions of the stimuli (Experiments 1 to 3) or by manipulating set size (Experiments 4 and 5). In the biased displays, the distractor appeared with a higher probability at one location than at all others, whereas in the uniform displays, distractors were uniformly distributed across locations. The central question was whether suppression learned in a biased configuration would transfer to a different, uniform configuration or would remain tied to the configuration in which it was acquired. The first three experiments demonstrated that suppression learned in one configuration generalized to other configurations that differed in the spatial positions of the stimuli. By contrast, only the last two experiments, in which the set sizes differed between configurations, provided evidence that the transfer of suppression was substantially reduced. Together, these findings suggest that learned distractor suppression is generally rigid and remains in place across different stimulus configurations. However, when configurations differ substantially, suppression may become at least partly configuration dependent.