Abstract
In a visual search task, when the spatial configuration of the distractor is associated with the target position (i.e. context), incidental learning of the context reduces the target search time, known as a contextual cueing effect (CCE). Guiding the attention towards the relevant target position by the context has been regarded as the mechanism of the CCE. Here, we show that such guidance of attention is not only induced by the presentation of the context, but that after learning, attention is guided towards spatial locations associated to context-relevant-targets, regardless of whether the context is presented or not. Participants (n=50) were required to search for a target under two distractor-configuration conditions. In the repeated condition, a specific spatial configuration of distractors was associated with the specific target position (CT; context-target). In the random condition, targets were positioned axisymmetric to the CT along with a random configuration of distractors that changed each trial (nCT; non-context-target). The search time reduced in the repeated condition, confirming the CCE. In the repeated condition, as learning progressed, the landing position of the first saccade (first fixation point) from the trial-start deviated from a central fixation point toward the CT, suggesting attentional guidance towards the CT. In the random condition, although the relevant target position was nCT instead of CT, the first fixation point also deviated towards the CT. The amount of deviation correlated with the individual strength of the CCE, both for the repeated and random conditions, indicating the relationship between the context-facilitated search and the first-fixation bias. Our results point to a novel attention guiding mechanism that generalizes the effect of the incidental learning to the whole task-set. Processing of the target associated with the context may be prioritized in the visual search task, which suggests a preference for regularity in human information processing.