Study of the hearing losses of industrial workers with occupational noise exposure, using statistical methods for the analysis of qualitative data.

The application of a polychotomous generalization of the linear probabilistic model of Cox is described, which allows the investigation of the effects of different noise levels with various durations of exposure on hearing. If the model assumptions are not rejected, it is possible to separate the effect of noise exposure from the effect of age. The results will thus show to what extent the hearing losses of an industrial worker are due to presbyacusis or to noise exposure. An empirical investigation was based on the evaluation of hearing losses at the audiometric frequencies 1, 2 and 3 kHz in 35 212 industrial workers of both sexes. It was attempted to predict the distribution of hearing losses in the left and the right ear as an additive function of the effect of noise exposure, age and sex. The parameters describing the effect of age increased linearly. Noise levels equal to and above 85 dB(A) have a noxious influence on hearing. Furthermore, hearing loss does not increase continuously with duration of exposure; from 15 years onwards, no essential increase in hearing loss appears which can be attributed to noise at work.