Benefits of facial and textual information in understanding of vocoded speech

Exposure to audiovisually presented vocoded speech is more effective than exposure to auditory-only vocoded speech in improving the subsequent ability to understand vocoded speech [1]. In addition, improvements in the audiovisual training condition were more rapid and greater in magnitude than in the auditory-only condition. The current study was conducted to establish whether exposure to concurrent textual information also results in improvements in the ability to recognize vocoded speech. Baseline measures of identification performance with auditory-only vocoded speech sentences were assessed for 45 participants. Participants then performed a speech identification task, where they were exposed to vocoded speech in either audiovisual (Group 1), auditory-only (Group 2), or auditory-only with concurrent text conditions (Group 3). Following exposure, participants were tested again on identification performance with auditory-only vocoded speech. Exposure to concurrent text improved subsequent understanding of vocoded speech, to a level similar to that seen with audiovisual speech exposure. In a second experiment, groups of normal hearing adults were exposed to vocoded non-lexical nonsense words in auditory-only with text and audiovisual speech presentation conditions. Exposure to both nonsense audiovisual and concurrent text conditions improved subsequent understanding of lexical auditoryonly vocoded speech, and there was no difference between the levels of improvement. In summary, to computerbased audiovisual and concurrent text exposure improves the ability to recognize vocoded speech over exposure to auditory stimuli alone. This effect does not appear to be dependent on exposure to lexical items.