Attentional and Contextual Priors in Sound Perception

Area of Science:

• Cognitive psychology
• Auditory perception
• Computational neuroscience

Background:

• Explicit attentional cues significantly impact perception and performance.
• Implicit contextual cues from the environment also guide expectations, but their influence and relation to explicit cues are unclear.

Purpose of the Study:

• To investigate the independent and combined effects of explicit attentional cues and implicit contextual cues on auditory perception.
• To determine if listeners utilize contextual statistical information and how they update these priors.

Main Methods:

• Simultaneously manipulated explicit (attentional) and implicit (contextual) prior probabilities of acoustic frequencies in a sound detection task.
• Utilized a Bayesian Observer model to analyze listener behavior and adaptation.

Main Results:

• Both attentional and contextual priors had significant and independent effects on sound detectability.
• Listeners effectively tracked and updated contextual priors rapidly and optimally across various sound classes.
• A Bayesian Observer model successfully explained the observed attentional and contextual adaptations.

Conclusions:

• Perception can be robustly modeled as Bayesian inference, integrating both explicit and implicit prior information.
• Effects previously attributed solely to selective attention may arise from the integration of contextual priors.