An Analysis of Individual Differences in Recognizing Monosyllabic Words Under the Speech Intelligibility Index...

Area of Science:

• Auditory perception
• Speech processing
• Psychoacoustics

Background:

• Individual variability impacts speech recognition.
• The Speech Intelligibility Index (SII) models speech audibility.
• Listener-specific factors may influence spectral weighting.

Purpose of the Study:

• To investigate individual differences in speech recognition using the SII framework.
• To develop and apply an adaptive psychophysical procedure for estimating listener-specific SII parameters.
• To assess the consistency of spectral weighting across different speech materials and frequency bands.

Main Methods:

• Developed a quick-band-importance-function procedure for adaptive SII fitting.
• Simultaneously estimated band importance and link functions with 200-300 trials.
• Collected data from 30 normal-hearing listeners on NU6 and SPIN tests.

Main Results:

• Observed significant individual differences in spectral weights across frequency regions for both NU6 and SPIN tests.
• Found correlations in spectral weights between NU6 and SPIN materials at frequencies with high intersubject variability.
• Demonstrated that listener-originated factors likely drive spectral weighting differences.

Conclusions:

• The developed procedure efficiently estimates listener-specific SII parameters.
• Individual differences in spectral weighting are a key factor in speech recognition variability.
• Listener characteristics, not just speech material, significantly influence auditory processing and speech intelligibility.