A novel bias-free approach for robust perceptual threshold estimation

Area of Science:

• Psychology and Neuroscience
• Sensory Perception
• Psychophysics

Background:

• Perceptual threshold estimation is crucial for understanding sensory processing.
• Classical methods like constant stimuli and staircase procedures often yield unreliable estimates.
• These methods fail to account for response biases, conflating sensitivity with decisional criteria.

Purpose of the Study:

• To develop and validate bias-free methods for perceptual threshold estimation.
• To demonstrate the overestimation of thresholds by classical procedures.
• To show the effectiveness of bias-free methods in mitigating decisional bias.

Main Methods:

• Development of bias-free versions of constant stimuli and staircase procedures.
• Inclusion of target-absent trials to measure decisional bias.
• Testing across 74 participants to evaluate method reliability.

Main Results:

• Classical procedures significantly overestimate perceptual thresholds.
• Interindividual variability in participant criteria contributes to overestimation.
• The bias-free staircase procedure provides reliable sensitivity thresholds, free from decisional bias.

Conclusions:

• Bias-free threshold estimation methods are essential for accurate perceptual research.
• The developed bias-free staircase procedure offers a reliable and practical solution.
• Adoption of these methods will enhance scientific investigation reliability across fields.