A comparison of two perceptual voice evaluation training programs for naive listeners

Area of Science:

• Speech and Hearing Sciences
• Acoustic Phonetics
• Perceptual Voice Evaluation

Background:

• Traditional clinical perceptual voice evaluation faces reliability concerns.
• Methods like references, training, and analysis-by-synthesis have been proposed to enhance reliability.
• Direct comparisons of these reliability-enhancing methods are lacking.

Purpose of the Study:

• To compare the effectiveness of two listener training programs (anchors method vs. paired comparison method).
• To improve naive listeners' ability to detect subtle differences in voice breathiness.
• To contribute to the theoretical framework of voice quality perception.

Main Methods:

• Two distinct listener training programs were implemented: an anchors method and a paired comparison method.
• Trained naive listeners were assessed on their ability to detect subtle voice breathiness.
• Statistical analysis was used to compare listener performance before and after training.

Main Results:

• Listeners demonstrated significant improvement in detecting voice breathiness after undergoing training.
• The study yielded equivocal results regarding the comparative effectiveness of the anchors method versus the paired comparison method.
• A need for listener training prior to employing the analysis-by-synthesis method was suggested.

Conclusions:

• Listener training enhances perceptual voice analysis, particularly for subtle acoustic cues like breathiness.
• Further research is needed to determine the optimal training methodology for voice perception tasks.
• Findings support an exemplar-based approach to understanding the internal representations of voice quality.