Exaggerated perception of change with greater sensory imprecision

Area of Science:

• Auditory perception
• Computational neuroscience
• Psychophysics

Background:

• Bayesian models posit precision (inverse variance) as crucial for perception.
• Higher precision is typically assumed to increase perceived change from sensory input.
• Empirical evidence on the behavioral impact of precision in perception remains limited.

Purpose of the Study:

• To investigate how stimulus precision influences the perceived salience of systematic changes in auditory stimuli.
• To test the hypothesis that higher precision leads to greater perceived salience of change.

Main Methods:

• Four human experiments were conducted using Gaussian sequences of pure tones.
• Participants reported Perceived Salience of Change (PSC) for frequency or intensity variations.
• Stimulus precision was systematically manipulated across experimental conditions.

Main Results:

• Contrary to the Bayesian hypothesis, higher Perceived Salience of Change (PSC) was observed in low-precision conditions.
• This counter-Bayesian effect was more pronounced with extreme stimulus values and frequency variations.
• Low precision's impact on reporting the direction of frequency change varied.

Conclusions:

• Human perception may employ counter-Bayesian strategies in certain low signal-to-noise scenarios.
• This suggests perception's flexibility beyond strict Bayesian inference.
• Further research is needed to explore mechanisms, evolutionary advantages, and clinical relevance.