Abstract
Introduction. The complexity of our visual environment entails selectivity in visual processing. Previous research suggests that spatial orienting is driven by salient stimuli or an observer’s goal. However, in the real world, one's previous experience frequently directs spatial attention, often in an implicit manner. How does implicit learning affect the spatial orienting of attention, and is implicitly learned attention a distinctive form of attention? Methods. To characterize implicitly learned attention, we examine its long-term persistence and sensitivity to cue validity. In a location probability learning task, participants searched for a T target among L distractors. Unbeknownst to them, across multiple trials, the target was more often located in one region of the screen than in others. Learning was reflected by faster RT when the target fell in the rich rather than the sparse locations. Results. An attentional bias toward the rich locations developed rapidly over dozens of trials at the beginning of training. However, once a bias existed, it was slow to adjust to new statistical contexts. These biases persisted for at least a week and for hundreds of trials after the target's location became evenly distributed, even when new visual statistics made use of the learned attentional bias costly, and despite explicit knowledge that the learned bias was no longer valid. However, learned spatial biases did not always dominate spatial attention once they were acquired. The learned spatial bias did not transfer between different tasks, such as visual search and change detection. Furthermore, the learned spatial bias was substantially weakened by an endogenous cue (such as a central arrow) that varied from trial to trial. Conclusion. Long-term persistence differentiates implicitly learned attention from the more flexible goal-driven attention. It also suggests that visual statistical learning does not always reflect the true statistics of the environment.
Meeting abstract presented at VSS 2012