In a “same-different” psychophysical task three experienced listeners discriminated changes in frequency (Δf) that occurred in either of the following stimulus conditions: (1) a single 80- or 160-ms, 1.0-kHz tone burst; (2) in the first or in the second of two successive 80-ms, 1.0-kHz tones separated by intertone intervals of 5, 10, 20, 60, 100, or 150 ms; and (3) in both of the two successive tones described under (2). All stimulus conditions occurred randomly within each block of trials and training proceeded until performance reached asymptote. In the first condition, the value of Δf (d′ = 2.0) was ≈4.0 Hz regardless of tonal duration. For the second condition, the Δf in the first tone required for d′ = 2.0 ranged from 8 to 12 Hz, and that in the second tone ranged from 5 to 7 Hz, depending on the intertone interval. In the third condition the values of Δf (d′ = 2.0) were generally smaller than any of those of the second condition, but still larger than the √ d′ prediction for two independent observations. It appears that within the temporal intervals and experimental conditions of this study, the improvement in performance that is predicted with multiple observations is partly cancelled by temporal interferences that occur between the observations. [Supported by NRC Postdoctoral Fellowship to the author.]
Read full abstract