A stable capacity limit in interaural time‐difference discrimination across training and individual listeners.

Abstract

An interesting learning pattern was previously observed for ITD‐discrimination training with a 500‐Hz tone. While the highest of the twelve threshold estimates obtained each day from each listener (n=8) decreased significantly over days, the lowest did not. However, the lowest values averaged ∼6 microseconds, raising the question of whether they represented actual discrimination ability or resulted from random walks in the adaptive testing procedure. In an attempt to address this question, all twelve estimates of each listener were ordered from lowest to highest and the effect of training was examined separately for each rank. There was no improvement for the four lowest ranks, which ranged from ∼6–25 microseconds. These values represent 33% of the data, a seemingly unlikely rate for random walks. In addition, in naive listeners (n=32), despite marked inter‐individual variation in the average and highest thresholds, the lowest thresholds were constant across listeners. Though these analyses do not preclude the potential contribution of random walks within this data set, they keep open the possibility that the invariance in the lowest thresholds through training and across listeners was not an artifact of the adaptive procedure, but rather reflects a stable capacity limit in ITD processing. [Work supported by NIH.]