The test-retest reliability of suprathreshold values measured with a loudness-judgement task
Abstract
A variety of methods have been used to assess the loudness perceptions of listeners. The resulting suprathreshold values, however, have been highly variable. The poor test-retest accuracy of intensity levels judged to be comfortable and uncomfortable has been attributed to differences in procedures, stimuli and instructions. Data regarding the stability of loudness measurements should assist in the fitting of contemporary amplification. If decisions about hearing aid gain, output, and compression characteristics are to be based on suprathreshold values, it is important that they remain reasonably stable over time. The use of a well-defined loudness-measurement task should decrease the intrasubject variability of suprathreshold values. The purpose of this study was to examine the test-retest reliability of intensities judged to be comfortable and uncomfortable, as measured using a category-rating task called the Contour Test. Using the computer-driven Contour Test, loudness discomfort levels (LDL) and range of comfort (RoC) values were measured during two separate test sessions (separated by a period of days) for 62 hearing-impaired subjects. The Contour Test, developed at the University of Memphis and adopted for use in the Independent Hearing Aid Fitting Forum's (IHAFF) fitting protocol, is a category-rating loudness-judgment task. After reading instructions, subjects rated the loudness of tones using 7 descriptors displayed on a large-print card. Tonal stimuli, at 500 and 3000 Hz, were delivered through insert earphones and presented using an ascending method of limits. Subjects (33 males; 29 females), ranging in age from 27 to 85 years (mean 71.4; SD = 9.4), showed mild to moderate cochlear hearing loss and reported hearing aid use for at least six months. As expected, results indicated large intersubject variability for suprathreshold values. However, loudness measures were found to be stable over a period of days. Most subjects showed test-retest differences of only $\pm$5 dB. Paired t-test results suggested that, in general, LDL and range of comfort values did not vary significantly over time. Thus, the test-retest accuracy of suprathreshold values was found to be quite good.
Degree
Ph.D.
Advisors
Goldstein, Purdue University.
Subject Area
Audiology|Psychological tests
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.