The interplay of uncertainty, relevance and learning influences auditory categorization

Area of Science:

• Cognitive Neuroscience
• Auditory Perception
• Decision Making

Background:

• Auditory categorization, essential for understanding speech and music, is influenced by acoustic properties.
• Factors like sensory uncertainty, task relevance, and environmental statistics affect auditory perception.
• The interaction between these factors in shaping auditory categorization remains largely unexplored.

Purpose of the Study:

• To investigate how sensory uncertainty, task relevance, and environmental statistics interact to influence auditory categorization.
• To develop a computational framework for understanding auditory decision-making under multiple behavioral demands.

Main Methods:

• Human participants performed a multi-tone auditory categorization task.
• Bayesian modeling was used to analyze individual participant behavior.
• Analysis focused on how sensory features, task relevance, and learned statistics influenced category choices.

Main Results:

• Task-relevant tones had a greater impact on category choice than task-irrelevant tones.
• Poor estimates of task-relevant tones or high sensory uncertainty negatively affected category choice.
• Learning of sound statistics over short and long timescales biased decisions toward overrepresented categories, inversely correlating with relevance estimates.

Conclusions:

• Individuals uniquely weigh sensory uncertainty, task relevance, and statistical learning in auditory categorization.
• The study provides a novel computational framework for understanding sensory decision-making under complex demands.
• Findings reveal how multiple factors interact to shape auditory perception and categorization.