Tone-glide direction identification revisited

Dawson and Feth (2004) reported that listeners showed an apparent difference in sensitivity for the direction of frequency change when listening to virtual frequency glides or frequency modulated (FM) tones. Their work was a partial replication and extension of the work by Gordon and Poeppel (2002) using a one-interval fixed-block paradigm. In each interval, the listener was asked to report whether the frequency was rising or falling using buttons labeled up and down. Psychometric functions were reported but the original data were not examined for interval response bias. This study addresses possible bias in two ways. The original data were reexamined for response bias, and direction identification was tested using FM glides in the single-interval adjustment-matrix (SIAM) procedure. The SIAM procedure [Kaernbach, (1990)] is designed to control response bias in a one-interval paradigm. Contrary to the previous work, data collected with the SIAM procedure do not show a consistent pattern of sensitivity toward change in either direction. Additionally, psychometric functions using this procedure are steeper than those of the prior studies. Implications of these results in comparison with previous work will be discussed. [Work supported by a grant from NIH/NIDCD R01-DC006879].