Auditory-visual interaction in a dichotic listening task.

Abstract

Eleven normal adults identified consonant-vowel syllables (CV's) under three conditions after extensive training in the tasks. The auditory condition consisted of the simultaneous presentation of dichotic CV's. The Auditory-Visual I condition was the presentation of dichotic CV's with the simultaneous presentation of a visual CV to both eyes. The visual CV was always the same as one of the two auditory CV's. In the Auditory-Visual II condition, the auditory dichotic and visual CV's were again presented, but the visual CV was radomly varied so that it might or might not be identical to one of the auditory CV's. No significant difference between the right and left ears were found; however, there was a slight numerical superiority, especially in the auditory condition, in favor of the right ear. The presence of a visual stimulus tended to reduce the numerical superiority so that both ears performed more nearly the same. No significant differences were found in the performance of the right ear under the 3 test conditions. The left ear performance, however, tended toward enhancement with the presentation of a noncompeting visual stimulus but was significantly reduced when the visual stimulus competed with the auditory stimulus. A similar pattern was found under a double correct data analysis. Overall auditory performance was significantly poorer than was visual performance regardless of ear or test condition. Finally, error responses were more likely to be placed of articulation errors than voicing errors.