An Analysis of the Current Clinical Procedure for Measuring Hearing Thresholds and Recommendations for Improving the Measurement

Abstract

Measuring hearing thresholds is part of virtually all hearing evaluations. The standard clinical procedure for measuring hearing thresholds is easy to use, efficient, and its outcome is relatively reproducible (Levitt, 1971). However, the procedure has the downside that the task performance (proportion of ‘yes’ response, or PYes) targeted by the clinical thresholds is unspecified. Furthermore, it remains uncertain whether the procedure extracts information about threshold in an optimal way. The purpose of this paper is to address both issues. Clinical thresholds (THHW) were measured at 1 and 4 kHz for 22 normal-hearing adult listeners with 10 repetitions per listener at each frequency. For determining the target PYes for THHW, or PYesHW, psychometric functions, which relate PYesHW to signal level, were also measured. From the measured psychometric functions, both PYesHW and the threshold corresponding to PYes=0.5, or TH0.5 were derived. On average, PYesHW=0.93 and TH0.5=THHW- 3 dB. With minimal modifications to the current procedure, better estimates of thresholds were obtained either by fitting a psychometric function to the trials in the adaptive track, or by averaging the turning points (reversals) in the adaptive track. Therefore, the above simple modifications to the current clinical protocol could be utilized by clinicians for threshold estimation.