How to define thresholds for level and interaural-level-difference discrimination: Insights from scedasticities and...

Area of Science:

• Auditory perception
• Psychoacoustics
• Signal detection theory

Background:

• Sensitivity to sound level and interaural level difference (ILD) is crucial for hearing.
• Various methods exist for defining and averaging auditory thresholds, but their statistical suitability is unclear.

Purpose of the Study:

• To determine the most statistically suitable definition and averaging method for auditory discrimination thresholds.
• To assess homoscedasticity and normality of different threshold definitions.

Main Methods:

• Measured thresholds in human listeners across six conditions using an adaptive two-alternative forced-choice paradigm.
• Investigated thresholds defined as log-ratios, log-Weber fractions for intensity, and log-Weber fractions for amplitude.
• Analyzed thresholds for homoscedasticity and adherence to normal distribution.

Main Results:

• Log-ratio thresholds were heteroscedastic; log-transformation did not improve this.
• Log-Weber fraction thresholds showed better homoscedasticity.
• Log-Weber fraction for stimulus amplitude best approximated homoscedasticity and normal distribution.

Conclusions:

• Auditory discrimination thresholds should be defined as the logarithm of the Weber fraction for stimulus amplitude.
• Arithmetic averaging of these thresholds across listeners is recommended for statistical analysis.
• This approach enhances the reliability and interpretability of auditory sensitivity data.