Related Experiment Videos
Perceptual evaluation of voice quality: review, tutorial, and a framework for future research
J Kreiman1, B R Gerratt, G B Kempster
1VA Medical Center, West Los Angeles, CA.
Summary
Listener ratings for voice quality can be unreliable. This study proposes a new framework to explain variations in voice quality ratings, aiming for more consistent clinical evaluations of pathological voices.
Area of Science:
- Speech and Hearing Sciences
- Clinical Linguistics
- Perceptual Acoustics
Background:
- Listener ratings of voice quality are crucial in clinical voice research and serve as a benchmark for other assessment methods.
- Existing literature reveals significant variability in both intrarater and interrater reliability of voice quality ratings across studies.
- Clinical data show wide discrepancies in vocal roughness ratings among individual clinicians for the same voice sample.
Purpose of the Study:
- To address the lack of a theoretical framework explaining variability in voice quality ratings.
- To propose a model that can guide the development of standardized and valid clinical protocols for voice evaluation.
- To enhance the reliability of perceptual ratings in clinical voice assessment.
Main Methods:
- Literature review on the reliability of voice quality ratings.
- Analysis of existing clinical data on vocal roughness ratings.
- Development of a theoretical framework to explain rating variability.
Main Results:
- Identified significant fluctuations in intrarater and interrater reliability in previous research.
- Observed wide variations in individual clinician ratings of vocal roughness.
- Proposed a theoretical framework attributing rating variability to listener factors, task design, listener-task interactions, and random error.
Conclusions:
- The proposed theoretical framework offers an explanation for the inconsistency in voice quality ratings.
- This framework can inform the creation of more efficient, valid, and standardized clinical voice and speech evaluation protocols.
- Ultimately, this research aims to improve the reliability of perceptual voice ratings and deepen the understanding of pathological voice characteristics.