Effects of standard tone range and intersignal interval on frequency discrimination

Psychometric functions and reaction times were obtained for 4 listeners in a pure tone frequency discrimination task using a same - different paradigm. In the fixed standard conditions, each interval contained either a 1000 Hz tone (A) or a tone higher in frequency (B), such that all 4 combinations (AA, AB, BA, BB) were equally likely. In the roving standard conditions, the value of A was randomly selected from a set of 6 logarithmically spaced frequencies ranging from 725 to 1378 Hz. In separate 60 or 120 trial runs, the duration of the interval separating the 2 tones was varied from 600 msec to 8 sec. The discrimination results can be interpreted in terms of the memory assumptions in the model of intensity resolution proposed by Durlach and Braida. Increasing the intersignal interval had less effect in the fixed than in the roving standard conditions, presumably because context coding was possible. This interpretation is supported by the relative magnitudes of the RTs for decisions of same and different.