Abstract
Speech audiometry has been practised for many years and has proved useful in the diagnosis of hearing loss. The techniques currently practised in clinics in this country are reviewed briefly in Chapter 2. Some date back many years and the aim of this thesis is to show how recent technology can be used to advance the use and utility of speech audiometry. The historical development of speech audiometry is discussed in Chapter 1 to outline the reasoning behind the traditional techniques. Up to date theories of speech perception and psycho-acoustic measurements methods are examined in Chapter 3 to establish which of these new ideas may be used in speech audiometry. Recent developments in speech and computer technology are discussed in Chapter 4, these allow the implementation of more complex testing methods based on these new ideas. The measurement of speech, including speech level, spectral characteristics and word usage is discussed in Chapter 5. The importance of these properties to the subjective intelligibility of a set of test words is investigated in Chapter 6, using subjective data generated from the experiments described in Chapter 7. A model of word intelligibility is described in Chapter 7 and is used in computer simulations to ascertain the most efficient type of adaptive paradigm to be used in subjective experiments. These experiments provided useful information about the intelligibility of individual words. A new test of individual speech hearing acuity is developed in Chapter 7 using computer simulations. The new test results from the work on computer simulations and relies on the use of sets of specially selected words and an adaptive algorithm chosen on the basis of data collected both from simulations and also from subjective experiments. Informal trials of the new test on volunteers show that it is likely to be of use in the diagnosis of hearing loss. The new test and existing speech and pure tone tests are implemented on an integrated low-cost computer based system. Preliminary trials carried out with this system in a clinical environment are reported in Chapter 8 and confirm its simplicity of use and utility in audiological investigations.