Children and adults with significant communication challenges may rely on augmentative and alternative communication (AAC) systems to communicate. Electronic communication devices that produce synthesized/digitized speech offer significant advantages to interaction, but only if the speech is sufficiently intelligible to the user and his or her communication partners. The intelligibility of synthesized speech is consistently lower than natural speech, due to fewer cues available in the signal, and the limited attentional resources available to listeners to process the speech. One way to improve the intelligibility of speech output is by training the listener to better recognize and understand the output. The present application will address the question of whether a practice program will significantly improve the intelligibility of speech output for young children. A developmental perspective will be taken to investigate the performance of three- and four-year-old children.
The first aim of the application is to compare the effects of using words, sentences, and narratives as practice materials in increasing the intelligibility of synthesized speech for three- and four-year-old listeners. A Three Factor Mixed design will be used. Children will be assigned to one of four conditions (word, sentence, narrative-level practice, or no-training control group). Children in practice conditions will participate in four 20-minute practice sessions. The independent variables will include age, the practice materials, and time. The dependent measures will include intelligibility of single words and intelligibility of sentences.
The second aim i s to compare the effectiveness of two practice programs that differ in amount of listener engagement with the materials in increasing the intelligibility and comprehension of digitized and synthesized speech for three- and four-year-old listeners. The second experiment will use a Four Factor Mixed design, using age, speech type, the amount of engagement in the practice program, and time as independent variables. The dependent measures will include intelligibility of single words and comprehension of commands.
| Drager, Kathryn D R; Finke, Erinn H (2012) Intelligibility of children's speech in digitized speech. Augment Altern Commun 28:181-9 |
