Secondary Logo

Share this article on:

Spoken Word Recognition in Children With Cochlear Implants: A Five-Year Study on Speakers of a Tonal Language

Lee, Kathy Yuet Sheung; van Hasselt, Charles Andrew

Articles

Objectives: To study the effects of age at implantation and duration of implant use on the performance of spoken word recognition of pediatric cochlear implantees in a tonal language setting over a period of 5 years.

Design: Sixty-four children, given implants between the ages 1:01 and 14:09 (years:months), were divided into three age groups. They were tested on open-set word recognition ability at seven time intervals from before surgery to 5 years after surgery. Analyses of variance with repeated measurements were used to examine the effect of their age at implantation and the duration of implant use.

Results: Duration of implant experience was significant in spoken word recognition across the three age groups (p < 0.01). Children given implants below the age of 3 years caught up with the performance of the older children at 12 months after implantation. The difference in score reached statistical significance at 2 and at 3 years after surgery (p = 0.03, p = 0.00).

Conclusions: The performance of Cantonese-speaking children was similar to that of English-speaking children in that better outcomes were associated with longer implant experience as well as when implantation occurred at a younger age. The children implanted before the age of 3 and who had an implant experience of more than 2 years outperformed the children who were given implants after the age of 6 and also sustained these higher scores throughout 5 years of postimplant testing.

This study investigated the effects of age at implantation and duration of implant use on the performance of spoken word recognition on 64 Cantonese-speaking cochlear implant children over a period of five years. Continuous improvement was noted in all children in the five-year period irrespective of their age at implantation. Children implanted before the age of three improved at a slower rate before one full year of implant use. By two years of implant use, the performance of the young children outperformed the children who were implanted after the age of six, and also sustained these higher scores throughout five years of post-implant testing. The findings suggest that Cantonese- speaking children performed in a similar way to their English-speaking counterparts in that better word recognition outcomes were associated with longer implant experience and an earlier age of implantation.

The Institute of Human Communicative Research, The Chinese University of Hong Kong, SAR.

Address for correspondence: Kathy Yuet Sheung Lee, Ph.D., Division of Otorhinolaryngology, Department of Surgery, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, Hong Kong, SAR.

Received October 15, 2004; accepted March 29, 2005

A cochlear implant is an alternative for hearing-impaired people who have only limited benefit from conventional hearing aids. Earlier studies have focused on comparing the performance of cochlear implantees with hearing aid users. For prelingually deaf children with profound hearing loss, it is documented that cochlear implants are superior in assisting users to achieve better performance, both in terms of speech perception and language skills, including vocabulary knowledge and use of grammatical structures (Geers & Moog, 1994; Geers & Tobey; 1995; Miyamoto, Kirk, Todd, Robbins & Osberger, 1995; Tomblin, Spencer, Flock, Tyler & Gantz, 1999; Waltzman et al., 1997).

With cumulative data collected over time in children with implants, reports on more long-term outcome measures have become available. Implant experience was found to be a contributing factor to child speech perception performance. Continuous improvement in terms of overall speech perception, language, reading skills, and academic achievement was noted even after 3 years of implant use (Boothroyd & Boothroyd-Turner, 2002; Geers, 2004; Geers, Brenner, & Davidson, 2003; Kirk, Miyamoto, Lento, Ying, O’Neil, & Fears, 2002; Moeller, 2000).

Although age of implantation was found to be a determining factor in speech and language performance (Hammes, Novak, Rotz, Willis, Edmondson, & Thomas, 2002; Mondain et al., 1997; Waltzman et al., 1997), the claim was not supported in other studies (Gordon, Daya, Harrison, & Papsin, 2000; Tomblin, Spencer, Flock, Tyler, & Gantz, 1999). The different conclusions may be due to the different age ranges of the subjects included. Studies finding age at implantation as a significant contributor involved children given implants at ages younger than 5 years, whereas studies holding contrary conclusions involved subjects receiving implants at an older age (8.4 to 15.7 years).

The documented evidence is based primarily on English speakers. Reports on performance in Cantonese-speaking children are less numerous. Cheung, Lee, Chan, Tong & van Hasselt (2000) studied nine children with at least 3 months of implant experience and documented no major improvement in tone imitation and production. By using Cantonese lexical tone identification as the outcome measure, other studies included children given implants at between 2:06 to 8:08 (years:months), with a length of implant experience from 11 to 53 months (Ciocca, Francis, Aisha, & Wong, 2002; Lee, van Hasselt, Chiu & Cheung, 2002; Wong & Wong, 2004). Data have consistently shown that children with implants still presented great difficulties in accurate lexical tone identification irrespective of their age at implantation and duration of implant use.

The emphasis on studying tones is probably related to the fact that the use of tone is one of the dimensions in which Cantonese differs from English. There are six distinctive tones in Cantonese differing according to the fundamental frequency patterns (Gandour, 1981; Matthews & Yip, 1994). Whereas Cantonese uses tones contrastively (Bauer & Benedict, 1997), English does not.

In addition to the frequently cited prosodic difference of tone, Cantonese and English also differ in segmental phonology. In terms of the number of contrastive phonemes and the possible combination of sounds permitted in the phonotactic rules, Cantonese has a relatively simpler system than English (Matthews & Yip, 1994). As shown in Table 1, there are fewer vowels, diphthongs, and consonants in Cantonese. The number of permissible consonants that can appear at the end of a syllable in English (N = 20) is higher than that in Cantonese (N = 6). While Cantonese has no consonant clusters, English permits up to a sequence of three and four consonants in the word-initial and final positions, respectively. The corresponding numbers of word-initial and word-final consonant clusters cited were up to 61 and 109 (Avery & Ehrlich, 1992; Brinton, 2000; Gimson, 1986; Gimson & Cruttenden, 1994).

TABLE 1

TABLE 1

The nonsignificant contribution of age at implantation and duration of implant use documented in Cantonese-speaking children may be related to the fact that prosodic perception of tone rather than segmental perception ability was measured. Long-term outcomes on segmental perception performance of Cantonese-speaking, implanted children were not available. Based on the fact that Cantonese has a much simpler phonological system than English, we believe earlier implantation and longer implant experience yielding a better outcome would be demonstrable in the Cantonese speakers.

The purpose of the present study was to investigate the effect of age at cochlear implantation and the length of implant experience on the open-set word recognition ability by Cantonese-speaking children over a 5-year period.

Back to Top | Article Outline

Methods

Subjects

Sixty-four prelingually deaf children, given implants at ages from 1:01 to 14:11 (mean = 6:02), were studied. All subjects were profoundly hearing impaired, with a unilateral cochlear implant in place for at least 1 year. According to the age at implantation, the children were divided into three groups: (1) under 3 years of age (N = 15); (2) from 3 to 6 years of age (N = 18); and (3) older than 6 years of age (N = 31).

The age range for the grouping of children reflects their educational status. Group 1 subjects were children who had not yet enrolled in formal education. Group 2 and 3 children had enrolled in preprimary centers and primary schools, respectively. All children used oral communication mode. Detailed demographic characteristics of the subjects are presented in Table 2.

TABLE 2

TABLE 2

Back to Top | Article Outline

Materials

Open-set word recognition tests were devised. The date of performance would determine which of the two versions of the test was used. The older version, developed in 1999, consists of 10 items of multisyllabic words. The more recent version has been used since 2002. There are five equivalent forms in the 2002 version. Twenty-five items, 12 monosyllabic and 13 disyllabic Cantonese words, are included in each equivalent form. Stimuli used in both versions are considered as highly familiar words to young children, though the 1999 version is regarded as easier because all stimuli were multisyllabic words. Stimulus lists are included in (Table 3) and (Table 4).

TABLE 3

TABLE 3

TABLE 4

TABLE 4

TABLE 4

TABLE 4

Back to Top | Article Outline

Procedures

The test was performed within 3 months before surgery as well as 0.5, 1, 2, 3, 4, and 5 years after implantation. The number of subjects who received the test at each interval was 64, 61, 61, 56, 43, 34, and 24, respectively. The varied figures across the test intervals were due to, first, the different length of implant use of the children, and second, the fact that not every child was available at every test interval. The examiner sat at the child’s better ear side, whereas the accompanying person, usually the mother, sat on the child’s other side. Test items, presented as words in isolation, were presented through monitored live voice at the level of 65 dB (A) HL, with a range of 5 dB (A) HL. Participants were then asked to repeat each stimulus perceived. Each item was presented once only with two exceptions: (1) at the request of the subject and (2) if the subject did not show any response after the stimulus was presented for 5 seconds. Under the two circumstances, the examiner presented the test item one more time. For children with limited verbal expressive ability, the accompanying person acted as a model to demonstrate that responses in the modes of gestures, finger pointing, written words, and drawings were also accepted. Children were encouraged to use other nonverbal means of communication throughout the testing procedure. Pen and paper were provided for making responses. Scoring, performed by the examiner on-site, was expressed as a percentage of words correctly perceived by the subjects. Homophones were also regarded as correct responses so long as the subject could indicate the referent meanings of the homophones expressed. Analyses of variance (ANOVAs) with repeated measurements were used to explore the effect of implant experience and age of implantation on word recognition scores.

Back to Top | Article Outline

Results

Table 5 shows the descriptive statistics of word recognition scores of the three age groups over time. Considerable individual variations were noted, as shown in the high values of the standard deviations. The development of the children’s performance on word recognition is depicted in Figure 1.

TABLE 5

TABLE 5

Fig. 1. Mean word recognition score of the three age groups across the seven test intervals.

Fig. 1. Mean word recognition score of the three age groups across the seven test intervals.

A repeated-measures, univariate ANOVA was performed with word recognition score as the dependent variable. Age at implantation and experience of implant use were regarded as the independent variables. The main effect of implant experience was significant (F[6,322] = 31.54, p = 0.00), whereas that of age was not (F[2,322] = 2.95, p = 0.54). The interaction effect was significant (F[12,322] = 2.47, p = 0.04).

In view of the presence of interaction effect, which may confound the results obtained, individual one-way ANOVAs were used on the word recognition scores. Three separate repeated-measures ANOVAs were computed for each age group to test the effect of implant experience. The main effect was highly significant in each of the age groups (F[6,71] = 23.83, F[6, 95] = 10.12, F[6,156] = 7.54, respectively, p = 0.00). Post hoc Tamhane tests were applied to test the between group differences. As there were 21 pairs of comparison, the Bonferroni approach of an adjusted α value of 0.002 (0.05/21) was used to accommodate a potentially inflated type I error (Kutner, Nachtsheim, Neter & Li, 2005). Time interval comparisons yielding significant differences are listed in Table 6.

TABLE 6

TABLE 6

The effect of age at implantation was examined by using seven individual one-way ANOVAs at each interval of implant use. The effect was significant at the 6-month, 2-year, and 3-year postoperation time intervals (F[2, 58] = 3.31, p = 0.04; F[2,53] = 3.74, p = 0.03; F[2,40] = 5.74, p = 0.01). Post hoc Tamhanetests were subsequently performed at the three time intervals with the significance level set at the adjusted α of 0.017 (0.05/3). Significant group differences were noted between children given implants at younger than age 3 years and older than age 6 years at 2- and 3-year postoperative intervals. All other between-group differences were insignificant.

Back to Top | Article Outline

Discussion

Implant Experience

In general, a positive relation between the children’s implant experience and their word recognition performance was found. The mean scores and results of post hoc tests showed that the improvement in performance was particularly marked when compared with the preoperative condition where hearing aids had been used. The results obtained were consistent with those found in English-speaking and French-speaking children (Kirk, Miyamoto, Lento, Ying, O’Neil, & Fears, 2002; Mondain et al., 1997; Waltzman et al., 1997; Zwolan et al., 2004).

The benefits of cochlear implantation seem to apply to children over a diversified language background, including Cantonese. Further improvement in word recognition ability is expected, based on the fact that scores have been increasing continuously over the years. We believe a child’s word recognition score could serve as a good outcome indicator for long-term measurement, as it entails and captures the speech perception development with no ceiling effect being noted within our 5-year period.

Back to Top | Article Outline

Age at Implantation

The present study showed significantly better word recognition performance in children who had received their implant early (<3:00) than late (>6:00) when tested at 2 years and 3 years after surgery. In the first year, the early-implanted children showed slow progress, which may be related to their younger chronological age at the time of testing. The lower chronological age implies immature development in terms of their (1) language skills; (2) readiness and cooperation to take formal tests; and (3) ability to make consistent responses in the mapping process to attain the suitable level of stimulation. The effect of chronological age is also likely to be a factor contributing to the interaction between the length of implant use and age at implantation though this study is unlikely to be able to independently analyze it. Over time, as these early-implanted children began to build a sufficient language foundation together with improved test compliance, they surpassed the older children and had significantly higher scores at 2 and 3 years after surgery. The group of younger children sustained the highest score throughout the 5-year period of ongoing postoperation testing. These results suggest that children receiving implants before 3:00 have better outcomes than children implanted after 6:00 in the long run. Time for development of cognitive and basic language skills is, nevertheless, essential before the benefits in speech perception can be noted.

Having said earlier age at implantation is associated with better word recognition performance, the claim does not seem to hold in the present subject groups who were given implants from 1:01 to 6:00. The difference in score between children given implants before the age of 3 years and those who were given implants from 3:00 to 6:00 years of age did not reach a significant level despite the mean scores being higher in the younger group across six of the seven test intervals. The results run contrary to the research findings from English-speaking children given implants at age younger than 5 years (ranging from 1:06 to 5:00), when open-set test scores were significantly higher for children implanted earlier (Hammes, Novak, Rotz, Willis, Edmondson, & Thomas, 2002; Mondain et al., 1997; Waltzman et al., 1997).

One possible reason may be due to the small sample size in the present study. The other explanation may be that the word recognition task is easier for Cantonese than for English-speaking children. Whether Cantonese-speaking children received their implant younger than 3:00 or 6:00, they were able to cope with the task well, and thus no major difference in the performance between the two groups was shown.

The speculation that word recognition in Cantonese is easier than English is based on two observations. First, according to the Neighborhood Activation Model, a set of acoustic-phonetic patterns are activated on presentation of stimulus input. All patterns are activated regardless of whether or not they are real words in the lexicon (Luce & Pisoni, 1998). In the model, lexical density affects the ease of activation. One measure of lexical density is the number of words that can be generated from a target word by adding, deleting, or substituting one phoneme at a time. In Cantonese, the substitution of tone is also regarded as a lexical neighbor. Neighborhoods are highly correlated with phonotactics such that high and low phonotactics correspond to high and low density neighborhoods (Luce & Pisoni, 1998). As stated in the introduction, the phonotactic rules as well as the phonological system are simpler in Cantonese. Lower density neighborhood results when compared with English. The lower density neighborhood, in turn, speeds up the searching process and makes the word recognition task relatively easy for Cantonese-speaking children.

The other possible reason that Cantonese speakers would perform better than their English counterparts relates to the research design of the present study. Homophones, which exist in large quantities as lexical neighbors in Cantonese, were accepted as correct responses in the testing procedure. This would greatly reduce the load in the lexical decision process and result in a high accuracy rate.

Back to Top | Article Outline

Conclusions

In summary, results of the present study support the contention that Cantonese-speaking children performed in a similar way to their English-speaking counterparts in that better word recognition outcomes were associated with longer implant experience and an earlier age of implantation.

Continuous improvement in spoken word recognition performance was noted in all children in the 5-year period irrespective of their age at implantation. Children implanted before the age of 3 years improved at a slower rate before 1 full year of implant use. By 2 years of implant use, the performance of the young children had surpassed the performance of the older children, and they sustained the highest scores throughout the 5-year postoperation period.

The performance for children implanted before 3:00 did not showed a marked difference, with children receiving their implants from 3:00 to 6:00. The failure to show a significant difference may be related to the small sample size of the present study. Another possible explanation would be that Cantonese-speaking children are able to handle the word recognition task well so long as they receive sensory stimulation before the age of 6 years.

Back to Top | Article Outline

Acknowledgments

We thank all the children with cochlear implants and their parents in the Prince of Wales Hospital who kindly participated in the study. We are also very grateful to Dr. Theo Lorentz for his kind assistance in editing this paper.

Back to Top | Article Outline

References

Avery, P. & Ehrlich, S. (1992). Teaching American English Pronunciation. Hong Kong: Oxford University Press.
Bauer, R.S. & Benedict, P.K. (1997). Modern Cantonese Phonology. Berlin: Mouton de Gruyter.
Boothroyd, A., & Boothroyd-Turner, D. (2002). Postimplantation audition and educational attainment in children with prelingually acquired profound deafness. Annals of Otology, Rhinology, & Laryngology, 111 (5), (Suppl. 189) 79–84.
Brinton, L. J. (2000). The Structure of Modern English: A Linguistic Introduction. Amsterdam/Philadelphia: John Benjamins.
Cheung, D. M. C., Lee, K. Y. S., Chan, B. Y. T., Tong, M. C. F., & van Hasselt, C. A. (2000). Paediatric cochlear implantation: perspectives in tone reception and production development in the Cantonese population. In Waltsman S. B., Cohen N. (Eds). Cochlear Implants. Article 11B, p. 209–210. New York: Thieme Medical Publishers.
Ciocca, V., Francis, A., Aisha, R., & Wong, L. (2002). The perception of Cantonese lexical tones by prelingually deaf cochlear implantees. Journal of the Acoustical Society of America, 111 (5), 2250–2256.
Gandour, J. (1981). Perceptual dimensions of tones: evidence from Cantonese. Journal of Chinese Linguistics, 9, 20–36.
Geers, A., & Moog, J. (1994). Spoken language results: vocabulary, syntax, and communication. Volta Review, 96(5), 131–148.
Geers, A., Brenner, C. & Davidson, L. (2003). Factors associated with development of speech perception skills in children implanted by age five. Ear and Hearing, 24 (1), 24S–35S.
Geers, A. E. (2004). Speech, language, and reading skills after early cochlear implantation. Archives of Otolaryngology – Head & Neck Surgery, 130 (5), 634–638.
Geers, A. E. & Tobey, E. (1995). Longitudinal comparison of the benefits of cochlear implants and tactile aids in a controlled educational setting. Annals of Otology, Rhinology, & Laryngology, 104 (Suppl. 166), 328–329.
Gimson, A. C. (1986). An Introduction to the Pronunciation of English. (3rd Ed). Hong Kong: ELBS.
Gimson, A. C. & Cruttenden, A. (1994). Introduction to the Pronunciation of English. (5th Ed). London: E. Arnold.
Gordon, K. A., Daya, H., Harrison, R. V., & Papsin, B. C. (2000). Factors contributing to limited open-set speech perception in children who use a cochlear implant. International Journal of Pediatric Otorhinolaryngology, 56, 101–111.
Hammes, D. M., Novak, M. A., Rotz, L. A., Willis, M., Edmondson, D. M., & Thomas, J. F. (2002). Early identification and cochlear implantation: critical factors for spoken language development. Annals of Otology, Rhinology, & Laryngology, 111, 74–78.
Kirk, I. K., Miyamoto, R. T., Lento, C. L., Ying, E., O’Neil, T., & Fears, B. (2002). Effect of age at implantation in young children. Annals of Otology, Rhinology, & Laryngology, 111, 69–73.
Kutner, M. H., Nachtsheim, C. J., Neter, J., & Li, W. (2005). Applied Linear Statistical Models. (5th Ed). Boston: McGraw Hill.
Lee, K. Y. S., van Hasselt, C. A., Chiu, S. N., & Cheung, D. M. C. (2002). Cantonese tone perception ability of cochlear implant children in comparison with normal-hearing children. International Journal of Pediatric Otolaryngology, 63(2), 137–147.
Luce, P. A. & Pisoni, D. B. (1998). Recognizing spoken words: the Neighborhood Activation Model. Ear and Hearing, 19(1), 1–36.
Matthews, S. & Yip, V. (1994). Cantonese: A Comprehensive Grammar. London: Routledge.
Miyamoto, R. T., Kirk, K. I., Todd, S. L., Robbins, A. M., & Osberger, M. J. (1995). Speech perception skills of children with multichannel cochlear implants or hearing aids. Annals of Otology, Rhinology, & Laryngology, 104 (Suppl. 166), 334–337.
Moeller, M. P. (2000). Early intervention and language development in children who are deaf and hard of hearing. Pediatrics, 106(3), 1–9.
Mondain, M., Sillon, M., Vieu, A., Lanvin, M., Reuillard-Artieres, F. Tobey, E., & Uziel, A. (1997). Speech perception skills and speech production intelligibility in French children with prelingual deafness and cochlear implants. Archives of Otolaryngology - Head & Neck Surgery, 123(2), 181–184.
Tomblin, J. B., Spencer, L., Flock, S., Tyler, R., & Gantz, B. (1999). A comparison of language achievement in children with cochlear implants and children using hearing aids. Journal of Speech, Language, and Hearing Research, 42, 497–511.
Waltzman, S., Cohen, N. L., Gomolin, R., Green, J., Shapiro, W., Brackett, D., & Zara, C. (1997). Perception and production results in children implanted between 2 and 5 years of age. In Honjo I., & Takahashi H. (Eds): Cochlear implant and related sciences update. Advances in Otorhinolaryngology, 52, 177–180. Basel: Karger.
Wong, A. O. & Wong, L. L. (2004). Tone perception of Cantonese-speaking prelingually hearing-impaired children with cochlear implants. Otolaryngology – Head & Neck Surgery, 130(6), 751–8.
Zwolan, T. A., Ashbaugh, C. M., Alarfaj, A., Kilney, P. R., Arts, H. A., El-Kashlan, H. K., & Telian, S. A. (2004). Pediatric cochlear implant patient performance as a function of age at implantation. Otology & Neurotology, 25(2), 112–120.
© 2005 Lippincott Williams & Wilkins, Inc.