| Description |
This presentation discusses a study that used a fully automated approach to compare phonetic development of children who are deaf or hard of hearing, children with typical development, children who have language delays and children diagnosed with autism. Phonetic development is indicative of how a child is progressing towards conversational skills. In the study, a digital recorder (LENA) worn by each child collected daylong recordings of speech utterances and conversations. The daylong audio recordings were analyzed to identify child utterances, as well as the type and amount of utterances that were consonant-like units, vowel-like units, non-speech-like units and pauses. Results showed that consonant production positively correlates with age, and that children with hearing loss showed similar production patterns as children with typical hearing, and they produce more consonants than their peers with language delays or autism. Further, children with hearing loss produced more non-speech-like sounds and pauses than any other children in the study. This presentation will demonstrate the potential of using automated audio recording and transcripting to develop more research protocols and enhance intervention plans.
|