Abstract

this study concerns the subjective perception of the quality of the voice, more particularly in the case of dysphonia. Our general objective is to study the perceptual mechanisms, which constitute Hirano's GRBAS multidimensional perceptual rating scale. We tested the reliability of the GRB perceptual scale both in terms of test-retest reliability (intra-listener stability) and in terms of inter-listener agreement. We also studied how listeners process the three different parameters of the scale by varying the perceptual rating conditions of G, R, and B. The three attributes were rated simultaneously in one condition and in isolation in a second condition. The experiment was based on texts read by fifty dysphonic speakers. We selected five listeners specialized in voice assessment to provide perceptual judgments of the voices. Our results show that G is strongly correlated with R and/or B. When the GRB rating process is simultaneous, G is almost equal to the maximum score assigned to the R and/or B parameter (R²=0.97). Conversely, R and B are poorly correlated. While intra-listener variability was limited, inter-judge variability was important even in a homogeneous panel of judges.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call