Speech Feature Discrimination in Deaf Children Following Cochlear Implantation

Speech feature discrimination is a fundamental perceptual skill that is often assumed to underlie word recognition and sentence comprehension performance. To investigate the development of speech feature discrimination in deaf children with cochlear implants, we conducted a retrospective analysis of results from the Minimal Pairs Test (Robbins et al., 1988) selected from patients enrolled in a longitudinal study of speech perception and language development. The MP test uses a 2AFC procedure in which children hear a word and select one of two pictures (bat-pat). All children were prelingually deafened, received a cochlear implant before 6 years of age, and used either oral or total communication. Children were tested once every six months to a year for seven years; not all children were tested at each interval. By two years post-implant, the majority of these children achieved near-ceiling levels of discrimination performance for vowel height, vowel place, and consonant manner. Most of the children also achieved plateaus but did not reach ceiling performance for consonant place and voicing. Finally, children’s performance was related to measures of speech feature discrimination, spoken word recognition, and sentence comprehension. Introduction Speech feature discrimination is assumed to be a fundamental perceptual skill that underlies word recognition and sentence comprehension performance in normal-hearing listeners. With the advent of cochlear implants in young children, it is important to understand how effectively deaf children use input from their cochlear implants to discriminate such speech features. Previous research has shown that deaf children with cochlear implants (2 years post-implant) perceive spoken words using “broader” phonetic categories than normal-hearing children normally do (Pisoni et al., 1999). However, do deaf children with cochlear implants eventually catch up with their normal-hearing peers in terms of discriminating phonetic features in words? To investigate the development of speech feature discrimination in deaf children with cochlear implants, we conducted a retrospective analysis of results from the Minimal Pairs Test (Robbins et al., 1988) administered to patients enrolled in a longitudinal study of speech perception and language development at the Indiana University School of Medicine. A minimal pair consists of two words that differ in sound by only one feature (e.g., key-pea, big-bug). The MP test uses a 2-alternative forcedchoice procedure in which children hear a spoken word and select one of two pictures that depict a minimal pair. If speech feature discrimination underlies word recognition and sentence comprehension performance, we would also expect the results from the Minimal Pairs Test to be correlated with measures of spoken word recognition and sentence comprehension. Thus, we also conducted a retrospective analysis of results from the Phonetically Balanced Kindergarten Words (PBK) test (Haskins, 1949), a test of spoken word recognition, and the Common Phrases test (Osberger et al., 1991), a test of sentence comprehension. In these tests, children are asked to repeat either a word or a sentence spoken by the clinician. All children in the present study were prelingually deafened, received a cochlear implant before 6 years of age, and used either oral or total communication. Children were tested once every 6 or 12 months for seven years. We predicted that the children’s performance on all tests would improve over time and that children with oral communication experience would perform more accurately compared to children with total communication experience, particularly on the tests with an open-set response format. Moreover, we predicted that children’s scores on tests of speech feature discrimination,