A re-evaluation of tinnitus reliability testing

The purposes of these investigations were to (1) evaluate tinnitus loudness measures of unskilled normal listeners asked to imagine that they are experiencing a constant monaural tonal tinnitus and (2) compare the performance of these listeners to that of a sample of patients with tinnitus. Subgroups of 24 patients participated in two investigations. Results suggest that (1) normal subjects asked to imagine a high-pitched, tonal tinnitus show significantly greater tinnitus loudness matching levels (a) when they do not have an external reference, (b) after a 1-week interval, and (c) for low-frequency matching tones; (2) patients with high-pitched tinnitus did not demonstrate statistically significant differences in tinnitus loudness judgments within or between sessions or frequencies; (3) tinnitus patients do not differ significantly from normals feigning tinnitus in the variability of tinnitus loudness matching levels within a single session (two measures) or after a 1-week interval (one measure); and (4) normals feigning tinnitus tended to choose significantly greater loudness matching levels than did tinnitus subjects.