Selective attention and simple auditory recognition: Methodological considerations

We have reported at recent meetings a severe performance decrement in two-channel frequency difference limen tasks. In these studies, d's in both the selective attention and divided attention conditions were significantly depressed relative to monaural levels. When the signals to the two earphone channels were widely separated in frequency (by nine to ten critical bands), the selective attention condition was improved relative to the divided attention condition, yet still significantly below monaural performance levels. When all signals were within the same critical band (or were separated by only two or three critical bands), there were no significant differences between the two dichotic conditions. The present study generated psychometric functions under the selective and divided attention conditions for signals within the same critical band. Using this procedure, differences between the two dichotic listening conditions are evident in both absolute performance levels and in the psychometric functions. The results of the present experiment clearly indicate the need for greater use of adaptive psychophysical methods in the study of dual channel listening. [Research supported by grant NS 10995 from N.I.N.D.S.]