Abstract
This paper extends the findings of a previous report (Bamford et al., 1980) which described a principal components analysis of pure tone audiometric data. Using a similar approach to the analysis of a large number of pure tone audiograms, the present study confirms the previous work. By far the greater part of the the audiogram variance could be ascribed to two components, one associated with the degree of, and the other with the slope of, the hearing loss. Further, it is shown that the proportions of the variance attributed to the two components change in a meaningful way according to the type of hearing loss: middle-ear, inner-ear, and mixed.