Abstract:

Discrimination of target frequency increments co-occurring with irrelevant increments in frequency, duration, or level was studied. Pairs (t[inf 1],t[inf 2]) of 1500-Hz, 80-ms pure tones separated by 20--120 ms of silence (ITI) were presented in a three-interval, two-alternative, forced-choice (3I/2AFC) task. On each trial, a ``standard'' pair of tones (interval 1) was followed by two ``comparison'' pairs (intervals 2 and 3), with 500 ms of silence between observation intervals. One (randomly chosen) comparison pair contained a target increment in the frequency of either t[inf 1] or t[inf 2]; the size of the target increment was controlled by adaptive rules. In addition to the target increment, both comparison stimuli contained identical irrelevant increments in either the duration, frequency, or level of either t[inf 1] or t[inf 2]; there was no correlation between the observation intervals containing target and irrelevant increments. Listeners had to determine which observation interval contained the target increment. Discrimination of t[inf 1] frequency was degraded by irrelevant increments in t[inf 2] frequency or t[inf 2] level, particularly at the shortest ITI of 20 ms; prolonged discrimination training lessened the effects. Discrimination of t[inf 2] frequency was not affected by irrelevant increments. [Work funded by OCAST Project HR4-064.]