Abstract
Knowledge about the prior probability of target location may be exploited by the ocular motor system to enhance the efficiency of saccadic programming. However, little is known about whether these effects vary with saccadic programming demands or about their spatial organization.
Our first goal was to determine if the effect of prior probability differed significantly between prosaccades and antisaccades. Our second goal was to determine if prior probability effects differed between trials in which two targets were processed by the same cerebral hemisphere or by different hemispheres.
We performed two experiments. In the first, we contrasted a low-probability condition (0.125) consisting of eight targets, with two high-probability conditions (0.50) consisting of two targets, one in which the targets spanned the horizontal meridian, and one in which the targets spanned the vertical meridian. In a second experiment we replicated this design and added a third high-probability condition, in which the two targets were confined to a single quadrant.
We found that the effects of increased prior probability on prosaccades were minimal, with at most a slight reduction in reaction time. Antisaccades showed a much larger reduction in reaction times and improved directional accuracy, with less reflexive prosaccade errors. Both the directional accuracy and spatial precision of antisaccades were better for targets that spanned the vertical meridian, indicating a between-hemisphere advantage.
We conclude, first, that prior probability effects are larger for the more attentionally demanding antisaccade task. Second, the effects of prior probability on the computation of antisaccade trajectory show an inter-hemispheric advantage, consistent with theories of hemispheric interactions in attentional processing, through either divided resource allocation or selective filtering by the corpus callosum.