Forward- and backward-masked intensity discrimination measured using forced-choice and adjustment procedures

A “midlevel” hump in the intensity jnd has been reported for pure tones preceded [e.g., Zeng et al., Hear. Res. 55, 223-230 (1991)] or followed [Plack and Viemeister, J. Acoust. Soc. Am. 92, 3097-3101 (1992)] by an intense masker where the signal-masker interval was 100 ms. These previous studies used forced-choice procedures, in which subjects were required to indicate the more intense tone. Plack and Viemeister [J. Acoust. Soc. Am. 92, 3097-3101 (1992)] have proposed that the task of judging the intensity of the short probe tone, when it is presented along with an intense masker, may lead to cognitive or central factors influencing the results. The present experiments attempted to reduce these possible effects by measuring intensity jnd's using two additional paradigms. First, a “multiple-look” forced-choice method, in which subjects listened to the stimulus pairs several times before responding, was used to obtain only forward-masked intensity jnd's. Second, the method of adjustment was used to obtain both forward- and backward-masked intensity jnd's. Both the standard forced-choice method and the multiple-look forced-choice method yielded jnd data with a midlevel hump, when compared to jnd's measured without a masker. In contrast, jnd's obtained with the method of adjustment yielded jnd data with no midlevel hump. The present results suggest that the traditional method of adjustment for intensity discrimination, where subjects adjust the signal level to a point of subjective equality, may measure a fundamentally different quantity than that measured by forced-choice procedures.