Variables affecting pure tone and speech audiometry in experimentally simulated hearing loss.

Previous studies have found that subjects simulating hearing losses can provide information which is useful to audiological clinicians when patients produce exaggerated hearing levels with no apparent organic aetiology. Two groups of normally-hearing subjects were used, designated unsophisticated or sophisticated in audiometric testing. Each subject was instructed to feign any amount of hearing loss in both ears prior to carrying out pure tone and speech audiometry using monosyllabic speech material. A number of variables was studied to identify factors influencing success or failure in simulating hearing loss and to investigate strategies used by the subjects. Analysis of the audiometric results revealed no typical degree of hearing loss for tones or speech. The majority of the subjects simulated sensori-neural pure tone losses with flat audiometric configuration. Approach mode, ascending or descending, had no significant effect on simulated tonal threshold. Four strategies for simulating a loss were identified. A fair degree of test/retest reliability in hearing loss for speech and tones was observed. Large discrepancies in hearing loss for speech and tones were recorded with the hearing loss for tones typically in excess of that for speech. Responses during speech audiometry showed a high occurrence of 'no response' errors. No significant differences were recorded between the sophisticated and unsophisticated groups, except in the pattern of response during speech audiometry, when the sophisticated subjects gave significantly fewer correct responses and 'no response' errors.