Abstract
Previous research has shown that spatial attention benefits from multimodal cues. When cues predict visual features such as location, shape, and color of targets, perceptual and motor processing for the targets are facilitated. In real-world settings such as on-road driving, however, there is no guarantee that cues are always informative and informative cues in the past are still informative now. Given these changes in cue reliability, we tested how the reliability of multimodal, visual and auditory cues affects the cue utilization for a visual task. In the first learning session, participants were asked to report the shape of the target (L or T) appearing either left or right of the fixation point. Visual (color) and/or auditory (pitch) cues preceded the target and perfectly indicated the location of target. In the second session, we made either visual or auditory cue uninformative by reducing the reliability of either cue to 50% while keeping the reliability of the other cue 100%. Our results indicated that even when the cues did not predict the target location in the second session, the participants were not able to completely ignore the uninformative cues. Furthermore, the adjustment for removal of informational value differed between auditory and visual cues; The participants were faster when the preceding visual cues lost any informational value than when the preceding auditory cues became irrelevant to the task. These results suggest that the utilization and combination of multimodal sensory cues are not simply determined by the reliability of cues at present and the informational values learned in the past but also weights on or biases toward sensory modalities.
Meeting abstract presented at VSS 2015