The Effect of Experience on Classification of Voice Quality

Objectives/Hypothesis. The purpose of this study was to compare the agreement among several groups of listeners with different types of experience in regard to classifying voice quality. Study Design. This is a retrospective cross-sectional quasi-experimental design. Method. This study compared three groups: speech-language pathologists who specialize in voice, singing voice teachers, and inexperienced listeners. All groups were asked to classify voice samples as breathy, rough, or normal. Results. Results show a significant difference across all groups with speech-language pathologists demonstrating a substantial interrater agreement, k = 0.67, z = 103.07 (P < 0.01); singing voice teachers demonstrating a moderate interrater agreement, k = 0.53, z = 79.10 (P < 0.01); and inexperienced listeners demonstrating a fair interrater agreement, k = 0.24, z = 35.82 (P < 0.01). Conclusions. Experienced listeners demonstrated a higher interrater agreement as compared with inexperienced listeners, with speech-language pathologists demonstrating a superior agreement as compared with all groups.