Literacy development in children with cochlear implants

Abstract

In general, school-age children with significant hearing loss demonstrate poorer literacy skills than their typically hearing (TH) peers. However, it is often difficult to infer from research findings which individual factors contribute to overall learning outcomes, given the high degree of heterogeneity in the hearing-impaired population. Since the development of any specialised intervention is based on evidence about the strengths and weaknesses of the targeted group, gaps in research arising from population variability have clinical and educational significance. Of the broader population of children who are deaf and hard of hearing, the research presented in this thesis pertains to a homogeneous group of beginning readers with significant bilateral hearing loss, who use cochlear implants (CIs) and communicate via solely spoken language. A psycholinguistic approach, founded on existing theoretical models, informed the collection and analysis of research findings in this thesis, which has thus allowed for specific interpretations as to the skills underlying reading and writing outcomes in children with CIs.As described in Chapter 1, children with CIs have spoken and written language difficulties, relative to TH children. Chapters 2 and 3 of the present thesis expand on these findings, by comparing the behavioural performances of children with CIs with those of TH children, and by examining the underlying skills that contribute to observed literacy outcomes in each group. These results are complemented by research presented in Chapters 4 and 5, both of which use electroencephalography (EEG) to capture children’s neural responses to simple linguistic tasks. In the thesis conclusion (Chapter 6), all study findings are integrated to provide a thorough, yet contextually informed, understanding of how literacy develops in children with CIs during the early school years.The focus of Chapter 2 is on reading development in children with CIs who use spoken language. Results from a cohort of children with CIs were examined with reference to a TH control group, who were similar in age, gender and nonverbal reasoning ability. The CI group performed significantly worse than TH children on measures of reading accuracy and phonological processing. As expected, based on the Simple View of Reading model, word reading accuracy and listening comprehension contributed to reading comprehension for both groups. However, the predominant concurrent predictor of reading comprehension was word reading accuracy for the CI group and listening comprehension for the TH group. For all children, orthographic and phonological skills were found to contribute to word-level reading accuracy.Chapter 3 examined the spelling skills of CI and TH children. With respect to overall accuracy, children with CIs achieved similar irregular and nonsense word spelling scores to the TH control group. Yet, spelling errors were less phonologically plausible for children with CIs, suggesting that they applied phonics knowledge less effectively when producing written words. In support of this interpretation, letter-sound knowledge was significantly related to nonword spelling performance for the TH group, but not the CI group.In Chapter 4, EEG measures were used to examine semantic sensitivity to word-picture incongruence in the CI and TH groups. Comparisons were made on the basis that semantic processing skills – in combination with numerous other inter-related skills – contribute to overall word recognition and reading comprehension. Analysis of the results showed that a similar ‘N400 effect’, which is known to index sensitivity to lexical-semantic incongruence, was elicited in both the CI and TH groups. Hence, this fine-grained, semantically driven neural response was found to be normal in children with CIs.Chapter 5 examined on-line rhyme sensitivity in TH children and a small group of children with CIs, again implementing EEG methodology to capture participants’ neural responses. A letter rhyme judgement task elicited a significant ‘rhyme effect’ in children with typical hearing. Within this group, a larger amplitude of rhyme effect was significantly correlated with worse letter-sound knowledge, as measured behaviourally, and this relationship was mediated by nonverbal reasoning ability. Within the CI cohort, there was significant variability in the presence and size of rhyme effect elicited in each participant. At an individual level, those children who demonstrated a rhyme effect tended to have better nonverbal reasoning skills than those who did not demonstrate a rhyme effect.Results from all four of the aforementioned studies contribute to a more complete understanding of how literacy develops in beginning readers with CIs. In the context of a classroom environment, the research findings suggest that children with CIs may show written language skill deficits, stemming in part from underlying phonological limitations.