Abstract
Many studies have shown that statistical learning of target or distractor distributional regularities can bias attentional selection. A recent study demonstrated that participants could also extract spatiotemporal regularities that occurred across trials: participants were faster to find the target when its location was predicted by the target location on the previous trial. However, this across-trial statistical learning was only demonstrated for parallel search involving a “pop-out” shape singleton target. The current study investigated whether there is also learning of across-trial regularities when search is serial. Participants searched for a rotated “T” target among seven rotated “L” distractors that were positioned in an imaginary circle. The target was equally likely to appear at one of eight locations. Unbeknownst to participants, two spatiotemporal regularities involving two target locations each (e.g., T1->T2) were built in, which meant the target position T1 on trial N was 100% predictive of the target position T2 on trial N+1. On other trials, targets appeared at random locations across trials. In Experiment 1 where all search items were grey, we found that participants were unable to learn the existing regularities. In Experiment 2 we employed the same display and the same regularities except that during the first half of the experiment targets were colored red, allowing feature search to find the target. Critically, participants did learn the across-trial regularities during “pop-out” feature search and the learned biases persisted when search became serial. Participants were unaware of these regularities suggesting that learning was automatic and implicit. We propose that across-trial target-target associations learned during feature search, shaped a flexible priority map whereby the selection of the predicting location would result in up-weighting of the predicted location on the next trial. This flexible priority map remained active even when the search task changed from parallel feature search to serial search.