The capacity to detect synchronous audiovisual events is severely limited: Evidence from mixture modeling

Visual attention serves to select salient and relevant events from the visual input. Selective attention to a visual event can be driven by a synchronous sound. Interestingly, recent evidence suggests that a sound can only drive selection of 1 concurrent visual event, suggesting that attentional capacity is much lower for audiovisual events than for purely visual events. Here we corroborate and extend this finding using a mixture modeling technique that distinguishes between the probability and precision of perception. Observers were presented with displays of multiple continuously flickering objects, of which either 1 or 2 were coupled to a single sound. In 2 experiments, we found that the probability of correctly reporting an object was almost halved when the number of synchronized visual objects increased from 1 to 2. Precision, however, was not affected. This indicates that rather than attention being distributed across multiple simultaneous audiovisual events, just 1 of them is singled out for attentional selection. This was not due to a capacity limit for selecting the visual objects per se; pure visual cues elicited a much higher probability of report and in that case there were clear declines in precision at larger set sizes, indicating the concurrent selection of multiple items. The results point toward a dissociation in capacity for visually and aurally cued prioritization of visual objects. (PsycINFO Database Record