Integrating modality-specific expectancies for the deployment of spatial attention.

Area of Science:

• Cognitive Neuroscience
• Psychology
• Sensory Integration

Background:

• Spatial attention deployment is sensitive to stimulus predictability.
• Crossmodal links in spatial attention are known, but integration of divergent, concurrent predictions across senses is unclear.

Purpose of the Study:

• Investigate how modality-specific expectancies are processed and integrated.
• Examine the influence of cue predictability and context divergence on multimodal spatial attention.

Main Methods:

• Behavioral studies using a multimodal cueing paradigm.
• Independent manipulation of cue predictability for visual and tactile targets.
• Bayesian ideal observer model applied to response speed data.

Main Results:

• Integration of probabilistic contexts depended on predictability level and divergence.
• Highly divergent but equally predictable contexts favored separate processing, especially for visual targets.
• Evidence for flexible integration based on reliability.

Conclusions:

• Modality-specific predictions are integrated flexibly based on their reliability.
• Supports a model of separate, yet linked, modality-specific attentional systems.
• Efficient cross-sensory spatial attention deployment is achieved through this integration.