The intelligibility of speech produced by young children with cochlear implants

Objective: The purpose of the current study was to examine the intelligibility of speech produced by young children with cochlear implants. Specifically the questions posed was, does intelligibility vary across different sampling methods, by post-implantation age, and by listener familiarity? Participants: Six preschool children participated in the study. These children were selected because they met the following criteria: 1) had pre-lingual deafness; 2) had severe to profound binaural hearing loss; 3) fitted with either a Clarion or Nucleus-24 multi-channel cochlear implant; 4) used verbal communication rather than signing during treatment; 5) had a receptive vocabulary within 2 standard deviations of the mean according to their age; 6) were implanted before age 4 years; and 7) had postimplantation age of at least 18 months. Method: Data was gathered through a conversational speech sample, the Children's Speech Intelligibility Measure (CSIM), and rating scales. To assess percent intelligible in conversational speech, a thirty-minute language sample was collected and audio taperecorded. Two experienced listeners, who were unfamiliar with the speakers, listened to each tape individually and orthographically transcribed the samples. The listeners then developed one final transcription per child through a consensus method. Percent intelligible in conversation was then determined using a procedure described by Shriberg (1986). The CSIM was used to obtain percent intelligible in single words. Each child imitated 50 words. Their utterances were audio taped and played back to a panel of 3 inexperienced listeners. There were 18 listeners (3 per sample) total. The listeners were asked to identify the word they thought the child said. In addition, rating scales were filled out by both parents and the primary clinician of each child. These individuals reported how much they understood and how much they believed others understood of the child's speech. Data Analysis: Spearman Rank-Order Correlation Coefficients were used to determine relationships among the variables. Results: The only factor to reach statistical significance was post-implantation age. The failure to find other statistically significant correlations may have been due to the small sample size used in the current study.