Visual attention and co-activation of response decisions for features from different dimensions

The role of visual attention in task performance has been extensively debated. Several studies have offered different loci in the stream of visual processing, where attention may influence the performance (e.g., Allport, 1989; Tsal & Shalev, 1996). Following the Dimensional-Action model (e.g., Cohen & Shoup, 1997), we hypothesize that a major role of attention is to transfer response decisions from targets on which it is focused to high level centers dealing with response execution. This hypothesis predicts that response decisions for two targets will interact only when attention is focused on both targets, and only when the response to the targets is defined by different dimensions. Three experiments, using the redundancy gain paradigm, tested and confirmed this prediction. A special aspect of our study was the analysis of the cumulative probability density function (CDF) of the RT. This method allowed us to explore the data beyond the RT means, and study the entire distribution. Experiment 1 showed that co-activation of two cross-dimensional targets occurs only when the targets are positioned in the same location. Experiment 2 manipulated the focus of attention and showed that co-activation could occur even for targets positioned in different locations if they are both within the attentional focus. Experiment 3 showed that this attention-induced co-activation does not occur for targets from the same dimensional module. These results suggest that a major role of attention is post-perceptual and involves gating of selected responses to executive functions.