Choice of Japanese modifiers in noise annoyance scale during exposures to white noise

The linearity of a Japanese verbal scale of annoyance was estimated under real noise exposure. Young healthy adults, 32 male and 132 female were exposed to white noise of 30, 45, 60, 75 and 90 dBA in rooms where the exposure levels could be controlled according to contours. The series of noise exposure consisted of three components; white noise, pure-tone for conditioning and silence for evaluating noise. During the exposures the noisiness of each emission was evaluated with the verbal scale and a line-rating scale. The data showed a significant correlation of about 0.9 between the exposure level and the line length on the rating scale. The regression line of the noise level to the line length exhibited a reasonable locus of zero on the line length at the ambient noise level. The rank correlation between the selected modifier and the noise level was nearly 0.8 ( p < 0.001 ) . When the five modifiers were converted to 1–5 points, respectively, the regression of the line length to the selected modifier revealed an adequate and consistent fitness (coefficient of determination was > 0.8 ). The results suggest that the scale can be used as an equally divided linear scale in community noise research.