Equal appearing interval and visual analogue scaling of perceptual roughness and breathiness

Abstract

One of the factors that affects the reliability of perceptual voice evaluation is the rating scale. Equal-appearing interval (EAI) and visual analogue (VA) scales are the two most common scales used and have attracted much attention in recent studies of perceptual voice evaluation. Available findings are contradictory, with one study finding the EAI scale to be more desirable while an earlier study showed equivocal results. The objective of the present study was to compare the reliability of using an 11-point EAI scale and a VA scale, each of 10 cm long, for perceptual voice quality evaluation. Thirty listeners with no prior experience in perceptual voice evaluation were asked to rate the breathy and rough qualities of 28 voice samples. The results showed that the listeners demonstrated a significantly higher intra-rater agreement and lower inter-rater variability in rating the two perceptual voice qualities using the EAI scale, when compared to the VA scale. However, listeners tend to show more bias in using certain points on the EAI scale than on the VA scale. In addition, a linear relationship was found between the EAI and VA ratings, suggesting the psychoperceptual characteristics of breathy and rough qualities could be captured equally well by the EAI and VA scales. Since the ease of use of the rating scale is an important consideration in clinical situations, the 11-point, 10 cm long EAI scale is therefore more preferable, although not necessarily more superior, than the VA scale for evaluating breathy and rough qualities. © 2004 Taylor & Francis Ltd.