Abstract
Aim: Visual attention can benefit from both short and long term learning: if we're looking for a cat who has climbed up a tree, and we looked at the tree before he climbed, we'll ignore the tree and find the cat more quickly (as in ‘preview search’). However, if Whiskers always climbs up the same five trees, and always goes to the same branch on a given tree, we'll also get better at finding him (as in ‘contextual cueing’). Are these two forms of attentional learning additive, or do they interfere with each other? Methods and Results: We used a hybrid paradigm combining preview search with contextual cueing. Subjects were asked to look for a T among L's. Each trial began with a display of scattered black L-shaped distractors, followed one second later by a display with more distractors and a single black T. On valid preview trials, the distractors on the second display were added to those present on the first, while on invalid preview trials, all of the distractors in the second display were in new positions. In addition, half of all trials benefited from contextual cueing, as the same spatial layouts were repeated over the course of the experiment. As expected, valid preview trials were faster than invalid trials, but there was a contextual cueing benefit only on invalid trials. In a second study, a color cue segregated the old and new items: the first display items were red and the second were blue. In this case, contextual cueing benefits were found on both valid and invalid preview trials. Conclusion: Long term learning effects of contextual cueing are reduced in a typical preview paradigm, perhaps because subjects do not search exactly the same set of items on a given valid preview trial. They fail to ignore all of the previewed distractors, and instead search through a different subset of them from trial to trial, changing the attended set on a given layout. Color cues eliminate this variability by allowing consistent segregation of old and new items.