- Email: [email protected]
Lexical and age effects on word recognition in noise in normal-hearing children
International Journal of Pediatric Otorhinolaryngology 79 (2015) 2023–2027
Contents lists available at ScienceDirect
International Journal of Pediatric Otorhinolaryngology journal homepage: www.elsevier.com/locate/ijporl
Lexical and age effects on word recognition in noise in normal-hearing children Cuncun Ren a, Sha Liu a,*, Haihong Liu b, Ying Kong a, Xin Liu a, Shujing Li a a b
Beijing Tongren Hospital, Beijing Institute of Otolaryngology, Capital Medical University, Beijing, PR China Beijing Children Hospital, Capital Medical University, Beijing, PR China
A R T I C L E I N F O
A B S T R A C T
Article history: Received 25 June 2015 Received in revised form 19 August 2015 Accepted 21 August 2015 Available online 29 August 2015
Objectives: The purposes of the present study were (1) to examine the lexical and age effects on word recognition of normal-hearing (NH) children in noise, and (2) to compare the word-recognition performance in noise to that in quiet listening conditions. Methods: Participants were 213 NH children (age ranged between 3 and 6 years old). Eighty-nine and 124 of the participants were tested in noise and quiet listening conditions, respectively. The StandardChinese Lexical Neighborhood Test, which contains lists of words in four lexical categories (i.e., dissyllablic easy (DE), dissyllablic hard (DH), monosyllable easy (ME), and monosyllable hard (MH)) was used to evaluate the Mandarin Chinese word recognition in speech spectrum-shaped noise (SSN) with a signal-to-noise ratio (SNR) of 0 dB. A two-way repeated-measures analysis of variance was conducted to examine the lexical effects with syllable length and difficulty level as the main factors on word recognition in the quiet and noise listening conditions. The effects of age on word-recognition performance were examined using a regression model. Results: The word-recognition performance in noise was significantly poorer than that in quiet and the individual variations in performance in noise were much greater than those in quiet. Word recognition scores showed that the lexical effects were significant in the SSN. Children scored higher with dissyllabic words than with monosyllabic words; ‘‘easy’’ words scored higher than ‘‘hard’’ words in the noise condition. The scores of the NH children in the SSN (SNR = 0 dB) for the DE, DH, ME, and MH words were 85.4, 65.9, 71.7, and 46.2% correct, respectively. The word-recognition performance also increased with age in each lexical category for the NH children tested in noise. Conclusions: Both age and lexical characteristics of words had significant influences on the performance of Mandarin-Chinese word recognition in noise. The lexical effects were more obvious under noise listening conditions than in quiet. The word-recognition performance in noise increased with age in NH children of 3–6 years old and had not reached plateau at 6 years of age in the NH children. ß 2015 Published by Elsevier Ireland Ltd.
Keywords: Word recognition Noise Children Mandarin Chinese
1. Introduction In recent years, children with a hearing impairment can now be identified early as a result of the universal newborn hearing screening program. Hearing aids (HAs) and cochlear implants (CIs) have become the most efficient rehabilitative measures for children with a hearing impairment [1,2]. In China, the number of cochlear implantations has rapidly grown, reaching 32,000 users as of June, 2015. Both HAs and CIs have significantly improved
* Corresponding author at: Beijing Tongren Hospital, Beijing Institute of Otolaryngology, Capital Medical University, Beijing 100006, PR China. Tel.: +86 010 58265822. E-mail address: [email protected] (S. Liu). http://dx.doi.org/10.1016/j.ijporl.2015.08.034 0165-5876/ß 2015 Published by Elsevier Ireland Ltd.
speech perception in quiet environments for hearing-impaired children whose native language is Mandarin Chinese [3–5]. However, in everyday environments, background noise is inevitably present. The background noise exerts detrimental effects on speech recognition and may affect learning in children [6]. It is still a great challenge for users of either HAs or CIs to cope with the environment noise encountered in their everyday life [7]. While many clinical methods of assessment of speech perception in quiet have been developed and used for many years [3–12], few methods are available to evaluate speech perception ability of hearingimpaired children in noise [13]. In order to develop effective ways to evaluate speech perception in noise for hearing-impaired children, we need to have a good understanding of speech perception in noise for normal-hearing (NH) children. Many factors, such as sensory
2024
C. Ren et al. / International Journal of Pediatric Otorhinolaryngology 79 (2015) 2023–2027
factors, cognition, and linguistic experiences, may influence the ability of children’s speech perception in noise [14]. The speech perception ability in noise of NH children has been shown to develop gradually during 2–5 years of age in a nonlinear fashion (N = 174), exhibiting the largest stepwise increase between ages 3 and 4 years [15]. Nishi et al. [16] studied consonant recognition in noise in NH children between 4 and 9 years of age (N = 96). They indicated that 4- to 5-year-old children demonstrated poorer consonant perception abilities in noise compared to older children or adults. Leibold et al. [17] showed that NH children who were older than 11 years of age reached adult-like speech perception performance in the presence of speech-shaped noise (N = 90). Wilson et al. [18] examined word recognition in noise in children between 6 and 12 years old (N = 318). They found that wordrecognition performance in babble noise was stable between 9 and 12 years old, but still slightly poorer than the performance achieved by young adults over age 17. Studies mentioned above differ from each other in many aspects, such as age of the children, language, test material, type of noise, signal-to-noise ratio, and test type (open set or closed set). Little is known about word recognition in noise in children younger than 6 years old. Therefore, it is important to understand the speech perception ability in preschoolers (i.e., 3–6 years old) in noise conditions. The Lexical Neighborhood Test (LNT) developed by Kirk and her colleagues is widely used as an open-set word recognition assessment for pediatric listeners with cochlear implants [19,20]. The development of the LNT was theoretically driven by the Neighborhood Activation Model (NAM) [21]. In the NAM, a word stimulus input activates a set of similar acoustic-phonetic representations in memory. The process of selecting lexical words from a large number of potential candidates is dependent on the word frequency, the density and frequency of the neighbors. Lexically easy words are those with high word frequency and low neighborhood density, whereas lexically hard words are those with low word frequency and high neighborhood density. All items used in LNT were selected to be familiar to young children with relatively limited vocabularies. All stimuli were drawn from the Child Language Data Exchange System (CHILDES), which consists of transcripts of verbal exchanges between a child or children and a caregiver or another child in their everyday environment. Many researchers have also developed LNT materials for their native languages [22–24]. Our laboratory has established the Standard-Chinese version of the LNT [25] according to the same principles established by Kirk et al. [20]. Details of the word density and frequency were provided in the previous study [25]. We have administered the Standard-Chinese LNT in a quiet environment to children with NH and CIs. We have consistently found that lexical effects were present in quiet for NH children and for pediatric users of CIs, showing that recognition performance of easy words was better than hard words, and that of dissyllabic words was better than monosyllabic words [25–27]. These results in quiet offered support of the NAM in a different language (i.e., Mandarin Chinese) from the one that the NAM was initially developed (i.e., English). However, many NH children and a few children with CIs have reached ceiling performance with the Standard-Chinese LNT. Adding noise in the test would not only mimic everyday listening situations but also bring the performance down to overcome the ceiling effects. Thus far, no studies have tested the lexical effects in noise of young children using the LNT. The purposes of the present study were to determine whether the lexical effects exist in both quiet and noise testing conditions and to examine the age effects on word recognition of NH children in quiet and in noise. The results of word-recognition in noise will facilitate future usage of the LNT for Mandarin-speaking, hearingimpaired listeners in noise.
2. Methods and materials 2.1. Subjects Participants in the present study were 213 native MandarinChinese-speaking, normal-hearing children with age ranging from 3 to 6 years old (mean SD, 4.96 1.05). Eighty-nine of the children participated in the open-set word recognition test in noise. Onehundred twenty-four of the participants took part in the open-set word recognition test in quiet conditions. Table 1 shows the number and mean age of subjects in each group of age and listening condition (i.e., quiet or noise). All participants had normal hearing (thresholds better than 20 dB HL at 500, 1000, 2000 Hz and 4000 Hz in both ears) measured with pure-tone audiometry. All participants reported no history of difficulty with speech or hearing. The use of human subjects were reviewed and approved by the Institutional Review Board of Beijing Tongren Hospital. 2.2. Materials The open-set word recognition was assessed by the StandardChinese version of the LNT that consisted of three dissyllabic easy word lists (DE), three dissyllabic hard word lists (DH), three monosyllabic easy word lists (ME), and three monosyllabic hard word lists (MH), with each list containing 20 items. The complete list of words in Chinese characters as well as in pinyin (the phonemic spelling) were shown in a previous study [27]. Word recognition test was performed in quiet and in noise. The speech spectrum-shaped noise (SSN) masker was used for the test in noise. The SSN was generated by the following procedures. First, the audio files of all word tokens were concatenated. The average spectrum level for the concatenated words was analyzed using a Fast Fourier Transform (FFT). The frequency and amplitude values obtained from the spectrum analysis were used to create a FFT filter mimicking the shape of the average spectrum. Finally, a white noise was processed through the FFT filter. The noise started 500 ms prior to the onset of individual speech tokens and ended 500 ms after the offset of the speech tokens. 2.3. Procedures All participants were tested in a quiet classroom where the background noise was below 35 dB (A). The participants were seated at the calibration point, 1 meter from the loudspeaker at 08 azimuth. For the quiet test conditions, the speech signals were presented at 70 dB SPL. For the noise conditions, the speech tokens were fixed at 70 dB SPL and the root-mean-square (RMS) values of the SSN were equated to those of the speech tokens so that a 0 dB signal-to-noise ratio (SNR) was achieved. Since the equivalency among the three lists of the same lexical category was established in a previous study [25], one list was randomly selected as test material in each of the four lexical categories (four lists total) for each participant. The order of the four lists as well as every word in each list was randomly assigned. Participants responded by verbally repeating each word they had heard and were encouraged to make their best guess. The results
Table 1 Number of subjects in each age and listening-condition group (mean age is shown in the brackets).
Noise Quiet Total
3 yrs
4 yrs
5 yrs
6 yrs
Total
24 (3.66) 31 (3.62) 55 (3.64)
19 (4.71) 36 (4.46) 55 (4.55)
25 (5.61) 33 (5.29) 58 (5.43)
21 (6.61) 24 (6.37) 45 (6.48)
89 (5.13) 124 (4.84) 213 (4.96)
C. Ren et al. / International Journal of Pediatric Otorhinolaryngology 79 (2015) 2023–2027
2025
were scored as percent correct of identified words. The test lasted for 15 min and was completed in one session for each participant. 2.4. Data analysis A two-way analysis of variance (two-way ANOVA) was performed to examine the two main factors of lexical effects on word recognition in noise, that is word difficulty (easy vs. hard) and syllable length (monosyllable vs. dissyllable). We also used the regression analyses to examine the effects of age on performance of word recognition in noise. The comparison of the results in noise and quiet was conducted using an independent sample t-test.
3. Results 3.1. Word recognition in quiet and noise condition The mean percent-correct scores of all the 124 children tested in quiet conditions were 97, 90, 93, and 82% for the four lexical categories (i.e., dissyllabic easy, dissyllabic hard, monosyllabic easy, and monosyllabic hard categories), respectively. The mean percent-correct scores of all the 89 children tested in noise at 0 dB SNR were 85, 66, 71, and 46% for the corresponding four lexical categories, respectively. The differences between the test scores of quiet and noise conditions were 11, 24, 22, and 36 percentage points. The gap of correct scores between noise and quiet was greatest of monosyllabic hard words, followed by dissyllabic hard words, monosyllabic easy words, and dissyllabic easy words (Fig. 1). Independent-Sample t-test was performed to examine the differences between the performance in noise and quiet listening conditions. Results showed that the word-recognition scores in noise were significantly lower than the corresponding scores in quiet for each age-matched group children (3 years old: t = 17.092, p < 0.0001; 4 years old: t = 10.932, p < 0.0001; 5 years old: t = 12.468, p < 0.0001; 6 years old: t = 14.488, p < 0.0001) (see Fig. 2).
Fig. 1. Group mean word-recognition scores in quiet and noise at 0 dB SNR. Each box represents a correct word-recognition score in noise or quiet (i.e., DE-N: scores using DE lists in noise, DE-Q: scores using DE lists in quiet, DH-N: scores using DH lists in noise, DH-Q: scores using DH lists in quiet, ME-N: scores using ME lists in noise, ME-Q: scores using ME lists in quiet, MH-N: scores using MH lists in noise, MH-Q: scores using MH lists in quiet).
Fig. 2. Group mean word-recognition scores in quiet and noise at 0 dB SNR. Each box represents a correct word-recognition score in noise or quiet (i.e., N: noise, Q: quiet).
3.2. Effects of lexical characteristics and age on word recognition Figs. 3 and 4 show the mean scores and standard deviation of word recognition for the 3- to 6-year-old children in noise and in quiet, respectively. The four sets of bars represents the four lexical categories. The best performance occurred on the dissyllabic easy list, followed by the monosyllabic easy list, dissyllabic hard list, and monosyllabic hard list. Two-way ANOVA was conducted to examine the lexical effects on word identification in noise and in quiet. The dependent variable, word-recognition percent correct score, was found to be normally distributed for both the easy/hard words and for the dissyllable/monosyllable words, as assessed by the Shapiro–Wilk test. There was homogeneity of variance for the test scores in both the easy/hard words and the dissyllable/ monosyllable words as assessed by Levene’s test for equality of error variances. The ANOVA revealed that both factors (easy vs.
Fig. 3. Age effect on Mandarin Chinese word recognition in noise. Each set of four bars represents the four age groups (i.e., 3, 4, 5, and 6 years old). DE: Dissyllable Easy; DH: Dissyllable Hard; ME: Monosyllable Easy; MH: Monosyllable Hard. The horizontal bars indicate statistical significance between two age groups.
2026
C. Ren et al. / International Journal of Pediatric Otorhinolaryngology 79 (2015) 2023–2027
A regression analysis showed that age was significantly correlated with the mean word recognition scores in noise pooled across the four lexical categories (r = 0.586; p < 0.0001). The r2 value was 0.34, indicating that age could account for approximately one-third of the variance in the mean word recognition performance in noise. Fig. 4 shows the mean word-recognition scores as a function of age. The slope of the linear fit was 3.5 percentage points/year, suggesting that word-recognition performance in noise improved approximately 3.5 percentage points every year between 3 and 6 years old (Fig. 5). 4. Discussion 4.1. Lexical effect on word recognition in noise
Fig. 4. Age effect on Mandarin Chinese word recognition in quiet. Each set of four bars represents the four age groups (i.e., 3, 4, 5, and 6 years old). The horizontal bars indicate statistical significance between two age groups.
hard and dissyllable vs. monosyllable) had significant effects on the word-recognition performance in noise. In particular, the scores were significantly different between easy and hard word lists (F[1,88] = 410.180; p < 0.0001 in noise listening condition, F[1,123] = 198.304; p < 0.0001 in quiet listening condition) and between the disyllable and monosyllable lists (F[1,88] = 237.126; p < 0.0001 in noise listening condition, F[1,123] = 83.009; p < 0.0001 in quiet listening condition). The mean word-recognition scores in noise for the easy words were 20 percentage points higher than those for the hard words, and those for the dissyllabic words were 17 percentage points higher than those for the monosyllabic words. The mean word-recognition scores in quiet for the easy words were 9 percentage points higher than those for the hard words, and those for the dissyllabic words were 5 percentage points higher than those for the monosyllabic words. Interaction between the easy/hard words and dissyllabic/monosyllabic words on word recognition in noise was not significant (F[2,88] = 1.008, p = 0.316).
Lexical characteristics of the stimuli have been shown to have significant influences on spoken word recognition in quiet conditions [20,26–28]. The present study revealed that lexical characteristics have an even stronger effect on spoken word recognition in noise than in quiet listening conditions (see Fig. 1). Our results showed that the recognition scores were higher for the easy words than those for the hard words and that the recognition scores for the dissyllabic words were higher than those for the monosyllabic words. The lexical effect was weaker in the quiet listening conditions, perhaps due to the ceiling effects. We observed that the differences in word recognition scores between quiet and noise conditions were minimum for the dissyllabic easy words, and were maximum for the monosyllabic hard words, indicating that the lexical effects were stronger in the noisy listening conditions. We also found no interaction between the easy/hard words and dissyllabic/monosyllabic words on word recognition in noise. This finding is consistent with a previous study [29]. In that study, Krull et al. [29] tested word and sentence recognition in 34 NH children using words with different lexical characteristics in noise. They found that lexical effects on spoken word recognition were largest at poorer SNRs. It is suggested that lexical effects may be more evident at poorer SNRs because the degraded auditory signal limits access to the fine spectral details of speech. Children might have to rely more on the lexical properties of the stimuli to perform word-recognition tasks in noise. It is therefore speculated that, since the lexical effect is more obvious under difficult listening conditions, such as in the presence of background noise, children with hearing loss listening through a sensory aid (HA or CI) may exhibit strong reliance on the lexical properties to perform spoken word recognition. In a previous report, we showed that children with CIs (N = 230) demonstrated strong reliance on lexical properties (easy/hard and dissyllable/ monosyllable) of the stimuli when performing spoken word recognition [26]. Thus, the use of LNT in noise might be a valuable tool for clinical evaluation of spoken word recognition in the hearing impaired listeners. 4.2. Age effect on word recognition in noise
Fig. 5. Mean percent correct scores across the four categories (DE, DH, ME, and MH) in noise condition as a function of age. Each symbol represents one subject. The solid line is the linear fit of the data.
Speech perception in noisy environments by children is influenced by many factors, including age, cognition, and learning experience [14]. As a result, speech recognition performance of children with NH is poorer than that of adults. Bonino and colleagues [30] tested 5- to 10-year-old NH children (N = 16) and adults (18–30 years) (N = 10) to examine age-related changes in open-set word recognition in the presence of a continuous multitalker babble, a two-talker speech masker, or speech-shaped noise. Results showed that mature performance with multi-talker babble and speech-shaped noise was observed for 8–10-year-old, but not for those younger than 8 years old. All groups of children were significantly poorer at recognizing words than the adults in the
C. Ren et al. / International Journal of Pediatric Otorhinolaryngology 79 (2015) 2023–2027
two-talker masker. Wightman et al. [31] also reported the age effect for speech recognition in a single stream of competing speech for children up to 16 years of age (N = 46). There was a monotonic age effect, such that even the children in the oldest age group (13–16 years) demonstrated poorer performance than adults, indicating that informational masking effects are much larger in children than in adults. The present study examined the age effect on spoken word recognition ability in speech spectrumshaped noise for 3- to 6-year-old children. Results showed that the older children outperformed the younger children on word recognition tasks in noise environments and the oldest age group (6 years of age) had not reached plateau. Although these reports differed from each other in methodology, they all indicated that there were age-related trends in speech perception ability in noise. In sum, the present study examined the lexical and age effects on word-recognition in noise in NH children. Results showed that both lexical and age had significant effects on Mandarin-Chinese spoken word recognition in young children under noise conditions. The noise type used in the present study was speech spectrumshaped noise and the SNR used was set at 0 dB. The subjects in the present study were between 3 and 6 years old. Further studies will be necessary to examine word recognition performance in both NH children and children with HAs or CIs under various noise types (such as speech babble noise or speech spectrum-shaped noise) at several SNRs. Such studies will provide valuable information on masking release [31] and details on the developmental trajectory of word-recognition ability in noise in typically developing children and in children with hearing impairments. Acknowledgments This study was supported by grants from the National Natural Science Foundation of China (No. 81170916) and the Scientific Research Foundation for Capital Health Development (No. 20111017-05). Alexa Patton provided editorial assistance in the revision of the manuscript. References [1] J.K. Niparko, E.A. Tobey, D.J. Thal, L.S. Eisenberg, N.Y. Wang, et al., Spoken language development in children following cochlear implantation, JAMA 303 (15) (2010) 1498–1506. [2] B.S. Wilson, Getting a decent (but sparse) signal to the brain for users of cochlear implants, Hear Res. 322 (2015) 24–38. [3] L. Xu, X. Chen, H. Lu, N. Zhou, S. Wang, et al., Tone perception and production in pediatric cochlear implant users, Acta Otolaryngol. 131 (2011) 395–398. [4] Q. Liu, N. Zhou, B. Berger, D. Huang, L. Xu, Mandarin consonant contrast recognition among children with cochlear implants or hearing aids, and normal-hearing children, Otol. Neurotol. 34 (3) (2013) 471–476. [5] N. Zhou, J. Huang, X. Chen, L. Xu, Relationship between tone perception and production in prelingually-deafened children with cochlear implants, Otol. Neurotol. 34 (3) (2013) 499–506.
2027
[6] H.M. Yang, Y.J. Hsieh, J.L. Wu, Speech recognition performance under noisy conditions of children with hearing loss, Clin. Exp. Otorhinolaryngol. 5 (Suppl. 1) (2012) S73–S75. [7] J.D. Crew, J.J. Galvin III, D.M. Landsberger, Q.J. Fu, Contributions of electric and acoustic hearing to bimodal speech and music perception, PLOS ONE 10 (3) (2015) e0120279, 1–18. [8] J. Bench, A. Kowal, J. Bamford, B.K.B. The, (Bamford-Kowal-Bench) sentence lists for partially-hearing children, Br. J. Audiol. 13 (3) (1979) 108–112. [9] T.W. Tillman, R. Carhart, An Expanded Test for Speech Discrimination Utilizing CNC Monosyllabic Words, Northwestern University Auditory Test No. 6. SAM-TR66-55. Tech Rep SAM-TR, 1966, 1–12. [10] S.R. Silverman, I.J. Hirsh, Problems related to the use of speech in clinical audiometry, Ann. Otol. Rhinol. Laryngol. 64 (4) (1955) 1234–1244. [11] I.J. Hirsh, H. Davis, S.R. Silverman, E.G. Reynolds, E. Eldert, et al., Development of materials for speech audiometry, J. Speech Hear Disord. 17 (3) (1952) 321–337. [12] H.A. Haskins, A Phonetically Balanced Test of Speech Discrimination in Children, Northwestern University, Evanston, 1949. [13] S. Jerger, S. Lewis, J. Hawkins, J. Jerger, Pediatric speech intelligibility test. I. Generation of test materials, Int. J. Pediatr. Otorhinolaryngol. 2 (3) (1980) 217–230. [14] M. Talarico, G. Abdilla, M. Aliferis, I. Balazic, I. Giaprakis, et al., Effect of age and cognition on childhood speech in noise perception abilities, Audiol. Neurotol. 12 (1) (2007) 13–19. [15] Y. Zheng, K. Wang, D. Li, Z.L. Meng, Evaluation of speech perception in competing noise conditions for normally hearing children, Noise Health 15 (64) (2013) 178– 182. [16] K. Nishi, D.E. Lewis, B.M. Hoover, S. Choi, P.G. Stelmachowicz, Children’s recognition of American English consonants in noise, J. Acoust. Soc. Am. 127 (5) (2010) 3177–3188. [17] L.J. Leibold, E. Buss, Children’s identification of consonants in a speech-shaped noise or a two-talker masker, J. Speech Lang. Hear Res. 56 (4) (2013) 1144–1155. [18] R.H. Wilson, N.M. Farmer, A. Gandhi, E. Shelburne, J. Weaver, Normative data for the Words-in-Noise Test for 6- to 12-year-old children, J. Speech Lang. Hear Res. 53 (5) (2010) 1111–1121. [19] K.I. Kirk, L.S. Eisenberg, A.S. Martinez, M. Hay-McCucheon, The Lexical Neighborhood Test Test–Retest Reliability and Inter-list Equivalency, Indiana University, 1998. [20] K.I. Kirk, D.B. Pisoni, M.J. Osberger, Lexical effects on spoken word recognition by pediatric cochlear implant users, Ear Hear. 16 (5) (1995) 470–481. [21] M.S. Cluff, P.A. Luce, Similarity neighborhoods of spoken two-syllable words: retroactive effects on multiple activation, J. Exp. Psychol. Hum. Percept. Perform. 16 (3) (1990) 551–563. [22] K.C. Yuen, I.H. Ng, B.P. Luk, S.K. Chan, S.C. Chan, et al., The development of Cantonese Lexical Neighborhood Test: a pilot study, Int. J. Pediatr. Otorhinolaryngol. 72 (7) (2008) 1121–1129. [23] M.S. Vitevitch, E. Rodrı´guez, Neighborhood density effects in spoken word recognition in Spanish, J. Multiling. Commun. Disord. 3 (1) (2005) 64–73. [24] H.M. Yang, J.L. Wu, Y.H. Lin, Y.J. Sher, Development of Mandarin monosyllabic lexical neighborhood test (LNT), Cochlear Implants Int. 5 (Suppl. 1) (2004) 203–205. [25] C. Liu, S. Liu, N. Zhang, Y. Yang, Y. Kong, et al., Standard-Chinese Lexical Neighborhood Test in normal-hearing young children, Int. J. Pediatr. Otorhinolaryngol. 75 (6) (2011) 774–781. [26] H. Liu, S. Liu, S. Wang, C. Liu, Y. Kong, et al., Effects of lexical characteristics and demographic factors on Mandarin Chinese open-set word recognition in children with cochlear implants, Ear Hear. 34 (2) (2013) 221–228. [27] S. Wang, S. Liu, Y. Kong, H. Liu, J. Feng, et al., Psychometric properties of the StandardChinese lexical neighborhood test, Acta Otolaryngol. 134 (1) (2014) 66–72. [28] L.S. Eisenberg, A.S. Martinez, S.R. Holowecky, S. Pogorelsky, Recognition of lexically controlled words and sentences by children with normal hearing and children with cochlear implants, Ear Hear. 23 (5) (2002) 450–462. [29] V. Krull, S. Choi, K.I. Kirk, L. Prusick, B. French, Lexical effects on spoken-word recognition in children with normal hearing, Ear Hear. 31 (1) (2010) 102–114. [30] A.Y. Bonino, L.J. Leibold, E. Buss, Release from perceptual masking for children and adults, Ear Hear. 34 (1) (2013) 3–14. [31] F.L. Wightman, D.J. Kistler, Informational masking of speech in children: effects of ipsilateral and contralateral distracters, J. Acoust. Soc. Am. 118 (5) (2005) 3164–3176.
Contents lists available at ScienceDirect
International Journal of Pediatric Otorhinolaryngology journal homepage: www.elsevier.com/locate/ijporl
Lexical and age effects on word recognition in noise in normal-hearing children Cuncun Ren a, Sha Liu a,*, Haihong Liu b, Ying Kong a, Xin Liu a, Shujing Li a a b
Beijing Tongren Hospital, Beijing Institute of Otolaryngology, Capital Medical University, Beijing, PR China Beijing Children Hospital, Capital Medical University, Beijing, PR China
A R T I C L E I N F O
A B S T R A C T
Article history: Received 25 June 2015 Received in revised form 19 August 2015 Accepted 21 August 2015 Available online 29 August 2015
Objectives: The purposes of the present study were (1) to examine the lexical and age effects on word recognition of normal-hearing (NH) children in noise, and (2) to compare the word-recognition performance in noise to that in quiet listening conditions. Methods: Participants were 213 NH children (age ranged between 3 and 6 years old). Eighty-nine and 124 of the participants were tested in noise and quiet listening conditions, respectively. The StandardChinese Lexical Neighborhood Test, which contains lists of words in four lexical categories (i.e., dissyllablic easy (DE), dissyllablic hard (DH), monosyllable easy (ME), and monosyllable hard (MH)) was used to evaluate the Mandarin Chinese word recognition in speech spectrum-shaped noise (SSN) with a signal-to-noise ratio (SNR) of 0 dB. A two-way repeated-measures analysis of variance was conducted to examine the lexical effects with syllable length and difficulty level as the main factors on word recognition in the quiet and noise listening conditions. The effects of age on word-recognition performance were examined using a regression model. Results: The word-recognition performance in noise was significantly poorer than that in quiet and the individual variations in performance in noise were much greater than those in quiet. Word recognition scores showed that the lexical effects were significant in the SSN. Children scored higher with dissyllabic words than with monosyllabic words; ‘‘easy’’ words scored higher than ‘‘hard’’ words in the noise condition. The scores of the NH children in the SSN (SNR = 0 dB) for the DE, DH, ME, and MH words were 85.4, 65.9, 71.7, and 46.2% correct, respectively. The word-recognition performance also increased with age in each lexical category for the NH children tested in noise. Conclusions: Both age and lexical characteristics of words had significant influences on the performance of Mandarin-Chinese word recognition in noise. The lexical effects were more obvious under noise listening conditions than in quiet. The word-recognition performance in noise increased with age in NH children of 3–6 years old and had not reached plateau at 6 years of age in the NH children. ß 2015 Published by Elsevier Ireland Ltd.
Keywords: Word recognition Noise Children Mandarin Chinese
1. Introduction In recent years, children with a hearing impairment can now be identified early as a result of the universal newborn hearing screening program. Hearing aids (HAs) and cochlear implants (CIs) have become the most efficient rehabilitative measures for children with a hearing impairment [1,2]. In China, the number of cochlear implantations has rapidly grown, reaching 32,000 users as of June, 2015. Both HAs and CIs have significantly improved
* Corresponding author at: Beijing Tongren Hospital, Beijing Institute of Otolaryngology, Capital Medical University, Beijing 100006, PR China. Tel.: +86 010 58265822. E-mail address: [email protected] (S. Liu). http://dx.doi.org/10.1016/j.ijporl.2015.08.034 0165-5876/ß 2015 Published by Elsevier Ireland Ltd.
speech perception in quiet environments for hearing-impaired children whose native language is Mandarin Chinese [3–5]. However, in everyday environments, background noise is inevitably present. The background noise exerts detrimental effects on speech recognition and may affect learning in children [6]. It is still a great challenge for users of either HAs or CIs to cope with the environment noise encountered in their everyday life [7]. While many clinical methods of assessment of speech perception in quiet have been developed and used for many years [3–12], few methods are available to evaluate speech perception ability of hearingimpaired children in noise [13]. In order to develop effective ways to evaluate speech perception in noise for hearing-impaired children, we need to have a good understanding of speech perception in noise for normal-hearing (NH) children. Many factors, such as sensory
2024
C. Ren et al. / International Journal of Pediatric Otorhinolaryngology 79 (2015) 2023–2027
factors, cognition, and linguistic experiences, may influence the ability of children’s speech perception in noise [14]. The speech perception ability in noise of NH children has been shown to develop gradually during 2–5 years of age in a nonlinear fashion (N = 174), exhibiting the largest stepwise increase between ages 3 and 4 years [15]. Nishi et al. [16] studied consonant recognition in noise in NH children between 4 and 9 years of age (N = 96). They indicated that 4- to 5-year-old children demonstrated poorer consonant perception abilities in noise compared to older children or adults. Leibold et al. [17] showed that NH children who were older than 11 years of age reached adult-like speech perception performance in the presence of speech-shaped noise (N = 90). Wilson et al. [18] examined word recognition in noise in children between 6 and 12 years old (N = 318). They found that wordrecognition performance in babble noise was stable between 9 and 12 years old, but still slightly poorer than the performance achieved by young adults over age 17. Studies mentioned above differ from each other in many aspects, such as age of the children, language, test material, type of noise, signal-to-noise ratio, and test type (open set or closed set). Little is known about word recognition in noise in children younger than 6 years old. Therefore, it is important to understand the speech perception ability in preschoolers (i.e., 3–6 years old) in noise conditions. The Lexical Neighborhood Test (LNT) developed by Kirk and her colleagues is widely used as an open-set word recognition assessment for pediatric listeners with cochlear implants [19,20]. The development of the LNT was theoretically driven by the Neighborhood Activation Model (NAM) [21]. In the NAM, a word stimulus input activates a set of similar acoustic-phonetic representations in memory. The process of selecting lexical words from a large number of potential candidates is dependent on the word frequency, the density and frequency of the neighbors. Lexically easy words are those with high word frequency and low neighborhood density, whereas lexically hard words are those with low word frequency and high neighborhood density. All items used in LNT were selected to be familiar to young children with relatively limited vocabularies. All stimuli were drawn from the Child Language Data Exchange System (CHILDES), which consists of transcripts of verbal exchanges between a child or children and a caregiver or another child in their everyday environment. Many researchers have also developed LNT materials for their native languages [22–24]. Our laboratory has established the Standard-Chinese version of the LNT [25] according to the same principles established by Kirk et al. [20]. Details of the word density and frequency were provided in the previous study [25]. We have administered the Standard-Chinese LNT in a quiet environment to children with NH and CIs. We have consistently found that lexical effects were present in quiet for NH children and for pediatric users of CIs, showing that recognition performance of easy words was better than hard words, and that of dissyllabic words was better than monosyllabic words [25–27]. These results in quiet offered support of the NAM in a different language (i.e., Mandarin Chinese) from the one that the NAM was initially developed (i.e., English). However, many NH children and a few children with CIs have reached ceiling performance with the Standard-Chinese LNT. Adding noise in the test would not only mimic everyday listening situations but also bring the performance down to overcome the ceiling effects. Thus far, no studies have tested the lexical effects in noise of young children using the LNT. The purposes of the present study were to determine whether the lexical effects exist in both quiet and noise testing conditions and to examine the age effects on word recognition of NH children in quiet and in noise. The results of word-recognition in noise will facilitate future usage of the LNT for Mandarin-speaking, hearingimpaired listeners in noise.
2. Methods and materials 2.1. Subjects Participants in the present study were 213 native MandarinChinese-speaking, normal-hearing children with age ranging from 3 to 6 years old (mean SD, 4.96 1.05). Eighty-nine of the children participated in the open-set word recognition test in noise. Onehundred twenty-four of the participants took part in the open-set word recognition test in quiet conditions. Table 1 shows the number and mean age of subjects in each group of age and listening condition (i.e., quiet or noise). All participants had normal hearing (thresholds better than 20 dB HL at 500, 1000, 2000 Hz and 4000 Hz in both ears) measured with pure-tone audiometry. All participants reported no history of difficulty with speech or hearing. The use of human subjects were reviewed and approved by the Institutional Review Board of Beijing Tongren Hospital. 2.2. Materials The open-set word recognition was assessed by the StandardChinese version of the LNT that consisted of three dissyllabic easy word lists (DE), three dissyllabic hard word lists (DH), three monosyllabic easy word lists (ME), and three monosyllabic hard word lists (MH), with each list containing 20 items. The complete list of words in Chinese characters as well as in pinyin (the phonemic spelling) were shown in a previous study [27]. Word recognition test was performed in quiet and in noise. The speech spectrum-shaped noise (SSN) masker was used for the test in noise. The SSN was generated by the following procedures. First, the audio files of all word tokens were concatenated. The average spectrum level for the concatenated words was analyzed using a Fast Fourier Transform (FFT). The frequency and amplitude values obtained from the spectrum analysis were used to create a FFT filter mimicking the shape of the average spectrum. Finally, a white noise was processed through the FFT filter. The noise started 500 ms prior to the onset of individual speech tokens and ended 500 ms after the offset of the speech tokens. 2.3. Procedures All participants were tested in a quiet classroom where the background noise was below 35 dB (A). The participants were seated at the calibration point, 1 meter from the loudspeaker at 08 azimuth. For the quiet test conditions, the speech signals were presented at 70 dB SPL. For the noise conditions, the speech tokens were fixed at 70 dB SPL and the root-mean-square (RMS) values of the SSN were equated to those of the speech tokens so that a 0 dB signal-to-noise ratio (SNR) was achieved. Since the equivalency among the three lists of the same lexical category was established in a previous study [25], one list was randomly selected as test material in each of the four lexical categories (four lists total) for each participant. The order of the four lists as well as every word in each list was randomly assigned. Participants responded by verbally repeating each word they had heard and were encouraged to make their best guess. The results
Table 1 Number of subjects in each age and listening-condition group (mean age is shown in the brackets).
Noise Quiet Total
3 yrs
4 yrs
5 yrs
6 yrs
Total
24 (3.66) 31 (3.62) 55 (3.64)
19 (4.71) 36 (4.46) 55 (4.55)
25 (5.61) 33 (5.29) 58 (5.43)
21 (6.61) 24 (6.37) 45 (6.48)
89 (5.13) 124 (4.84) 213 (4.96)
C. Ren et al. / International Journal of Pediatric Otorhinolaryngology 79 (2015) 2023–2027
2025
were scored as percent correct of identified words. The test lasted for 15 min and was completed in one session for each participant. 2.4. Data analysis A two-way analysis of variance (two-way ANOVA) was performed to examine the two main factors of lexical effects on word recognition in noise, that is word difficulty (easy vs. hard) and syllable length (monosyllable vs. dissyllable). We also used the regression analyses to examine the effects of age on performance of word recognition in noise. The comparison of the results in noise and quiet was conducted using an independent sample t-test.
3. Results 3.1. Word recognition in quiet and noise condition The mean percent-correct scores of all the 124 children tested in quiet conditions were 97, 90, 93, and 82% for the four lexical categories (i.e., dissyllabic easy, dissyllabic hard, monosyllabic easy, and monosyllabic hard categories), respectively. The mean percent-correct scores of all the 89 children tested in noise at 0 dB SNR were 85, 66, 71, and 46% for the corresponding four lexical categories, respectively. The differences between the test scores of quiet and noise conditions were 11, 24, 22, and 36 percentage points. The gap of correct scores between noise and quiet was greatest of monosyllabic hard words, followed by dissyllabic hard words, monosyllabic easy words, and dissyllabic easy words (Fig. 1). Independent-Sample t-test was performed to examine the differences between the performance in noise and quiet listening conditions. Results showed that the word-recognition scores in noise were significantly lower than the corresponding scores in quiet for each age-matched group children (3 years old: t = 17.092, p < 0.0001; 4 years old: t = 10.932, p < 0.0001; 5 years old: t = 12.468, p < 0.0001; 6 years old: t = 14.488, p < 0.0001) (see Fig. 2).
Fig. 1. Group mean word-recognition scores in quiet and noise at 0 dB SNR. Each box represents a correct word-recognition score in noise or quiet (i.e., DE-N: scores using DE lists in noise, DE-Q: scores using DE lists in quiet, DH-N: scores using DH lists in noise, DH-Q: scores using DH lists in quiet, ME-N: scores using ME lists in noise, ME-Q: scores using ME lists in quiet, MH-N: scores using MH lists in noise, MH-Q: scores using MH lists in quiet).
Fig. 2. Group mean word-recognition scores in quiet and noise at 0 dB SNR. Each box represents a correct word-recognition score in noise or quiet (i.e., N: noise, Q: quiet).
3.2. Effects of lexical characteristics and age on word recognition Figs. 3 and 4 show the mean scores and standard deviation of word recognition for the 3- to 6-year-old children in noise and in quiet, respectively. The four sets of bars represents the four lexical categories. The best performance occurred on the dissyllabic easy list, followed by the monosyllabic easy list, dissyllabic hard list, and monosyllabic hard list. Two-way ANOVA was conducted to examine the lexical effects on word identification in noise and in quiet. The dependent variable, word-recognition percent correct score, was found to be normally distributed for both the easy/hard words and for the dissyllable/monosyllable words, as assessed by the Shapiro–Wilk test. There was homogeneity of variance for the test scores in both the easy/hard words and the dissyllable/ monosyllable words as assessed by Levene’s test for equality of error variances. The ANOVA revealed that both factors (easy vs.
Fig. 3. Age effect on Mandarin Chinese word recognition in noise. Each set of four bars represents the four age groups (i.e., 3, 4, 5, and 6 years old). DE: Dissyllable Easy; DH: Dissyllable Hard; ME: Monosyllable Easy; MH: Monosyllable Hard. The horizontal bars indicate statistical significance between two age groups.
2026
C. Ren et al. / International Journal of Pediatric Otorhinolaryngology 79 (2015) 2023–2027
A regression analysis showed that age was significantly correlated with the mean word recognition scores in noise pooled across the four lexical categories (r = 0.586; p < 0.0001). The r2 value was 0.34, indicating that age could account for approximately one-third of the variance in the mean word recognition performance in noise. Fig. 4 shows the mean word-recognition scores as a function of age. The slope of the linear fit was 3.5 percentage points/year, suggesting that word-recognition performance in noise improved approximately 3.5 percentage points every year between 3 and 6 years old (Fig. 5). 4. Discussion 4.1. Lexical effect on word recognition in noise
Fig. 4. Age effect on Mandarin Chinese word recognition in quiet. Each set of four bars represents the four age groups (i.e., 3, 4, 5, and 6 years old). The horizontal bars indicate statistical significance between two age groups.
hard and dissyllable vs. monosyllable) had significant effects on the word-recognition performance in noise. In particular, the scores were significantly different between easy and hard word lists (F[1,88] = 410.180; p < 0.0001 in noise listening condition, F[1,123] = 198.304; p < 0.0001 in quiet listening condition) and between the disyllable and monosyllable lists (F[1,88] = 237.126; p < 0.0001 in noise listening condition, F[1,123] = 83.009; p < 0.0001 in quiet listening condition). The mean word-recognition scores in noise for the easy words were 20 percentage points higher than those for the hard words, and those for the dissyllabic words were 17 percentage points higher than those for the monosyllabic words. The mean word-recognition scores in quiet for the easy words were 9 percentage points higher than those for the hard words, and those for the dissyllabic words were 5 percentage points higher than those for the monosyllabic words. Interaction between the easy/hard words and dissyllabic/monosyllabic words on word recognition in noise was not significant (F[2,88] = 1.008, p = 0.316).
Lexical characteristics of the stimuli have been shown to have significant influences on spoken word recognition in quiet conditions [20,26–28]. The present study revealed that lexical characteristics have an even stronger effect on spoken word recognition in noise than in quiet listening conditions (see Fig. 1). Our results showed that the recognition scores were higher for the easy words than those for the hard words and that the recognition scores for the dissyllabic words were higher than those for the monosyllabic words. The lexical effect was weaker in the quiet listening conditions, perhaps due to the ceiling effects. We observed that the differences in word recognition scores between quiet and noise conditions were minimum for the dissyllabic easy words, and were maximum for the monosyllabic hard words, indicating that the lexical effects were stronger in the noisy listening conditions. We also found no interaction between the easy/hard words and dissyllabic/monosyllabic words on word recognition in noise. This finding is consistent with a previous study [29]. In that study, Krull et al. [29] tested word and sentence recognition in 34 NH children using words with different lexical characteristics in noise. They found that lexical effects on spoken word recognition were largest at poorer SNRs. It is suggested that lexical effects may be more evident at poorer SNRs because the degraded auditory signal limits access to the fine spectral details of speech. Children might have to rely more on the lexical properties of the stimuli to perform word-recognition tasks in noise. It is therefore speculated that, since the lexical effect is more obvious under difficult listening conditions, such as in the presence of background noise, children with hearing loss listening through a sensory aid (HA or CI) may exhibit strong reliance on the lexical properties to perform spoken word recognition. In a previous report, we showed that children with CIs (N = 230) demonstrated strong reliance on lexical properties (easy/hard and dissyllable/ monosyllable) of the stimuli when performing spoken word recognition [26]. Thus, the use of LNT in noise might be a valuable tool for clinical evaluation of spoken word recognition in the hearing impaired listeners. 4.2. Age effect on word recognition in noise
Fig. 5. Mean percent correct scores across the four categories (DE, DH, ME, and MH) in noise condition as a function of age. Each symbol represents one subject. The solid line is the linear fit of the data.
Speech perception in noisy environments by children is influenced by many factors, including age, cognition, and learning experience [14]. As a result, speech recognition performance of children with NH is poorer than that of adults. Bonino and colleagues [30] tested 5- to 10-year-old NH children (N = 16) and adults (18–30 years) (N = 10) to examine age-related changes in open-set word recognition in the presence of a continuous multitalker babble, a two-talker speech masker, or speech-shaped noise. Results showed that mature performance with multi-talker babble and speech-shaped noise was observed for 8–10-year-old, but not for those younger than 8 years old. All groups of children were significantly poorer at recognizing words than the adults in the
C. Ren et al. / International Journal of Pediatric Otorhinolaryngology 79 (2015) 2023–2027
two-talker masker. Wightman et al. [31] also reported the age effect for speech recognition in a single stream of competing speech for children up to 16 years of age (N = 46). There was a monotonic age effect, such that even the children in the oldest age group (13–16 years) demonstrated poorer performance than adults, indicating that informational masking effects are much larger in children than in adults. The present study examined the age effect on spoken word recognition ability in speech spectrumshaped noise for 3- to 6-year-old children. Results showed that the older children outperformed the younger children on word recognition tasks in noise environments and the oldest age group (6 years of age) had not reached plateau. Although these reports differed from each other in methodology, they all indicated that there were age-related trends in speech perception ability in noise. In sum, the present study examined the lexical and age effects on word-recognition in noise in NH children. Results showed that both lexical and age had significant effects on Mandarin-Chinese spoken word recognition in young children under noise conditions. The noise type used in the present study was speech spectrumshaped noise and the SNR used was set at 0 dB. The subjects in the present study were between 3 and 6 years old. Further studies will be necessary to examine word recognition performance in both NH children and children with HAs or CIs under various noise types (such as speech babble noise or speech spectrum-shaped noise) at several SNRs. Such studies will provide valuable information on masking release [31] and details on the developmental trajectory of word-recognition ability in noise in typically developing children and in children with hearing impairments. Acknowledgments This study was supported by grants from the National Natural Science Foundation of China (No. 81170916) and the Scientific Research Foundation for Capital Health Development (No. 20111017-05). Alexa Patton provided editorial assistance in the revision of the manuscript. References [1] J.K. Niparko, E.A. Tobey, D.J. Thal, L.S. Eisenberg, N.Y. Wang, et al., Spoken language development in children following cochlear implantation, JAMA 303 (15) (2010) 1498–1506. [2] B.S. Wilson, Getting a decent (but sparse) signal to the brain for users of cochlear implants, Hear Res. 322 (2015) 24–38. [3] L. Xu, X. Chen, H. Lu, N. Zhou, S. Wang, et al., Tone perception and production in pediatric cochlear implant users, Acta Otolaryngol. 131 (2011) 395–398. [4] Q. Liu, N. Zhou, B. Berger, D. Huang, L. Xu, Mandarin consonant contrast recognition among children with cochlear implants or hearing aids, and normal-hearing children, Otol. Neurotol. 34 (3) (2013) 471–476. [5] N. Zhou, J. Huang, X. Chen, L. Xu, Relationship between tone perception and production in prelingually-deafened children with cochlear implants, Otol. Neurotol. 34 (3) (2013) 499–506.
2027
[6] H.M. Yang, Y.J. Hsieh, J.L. Wu, Speech recognition performance under noisy conditions of children with hearing loss, Clin. Exp. Otorhinolaryngol. 5 (Suppl. 1) (2012) S73–S75. [7] J.D. Crew, J.J. Galvin III, D.M. Landsberger, Q.J. Fu, Contributions of electric and acoustic hearing to bimodal speech and music perception, PLOS ONE 10 (3) (2015) e0120279, 1–18. [8] J. Bench, A. Kowal, J. Bamford, B.K.B. The, (Bamford-Kowal-Bench) sentence lists for partially-hearing children, Br. J. Audiol. 13 (3) (1979) 108–112. [9] T.W. Tillman, R. Carhart, An Expanded Test for Speech Discrimination Utilizing CNC Monosyllabic Words, Northwestern University Auditory Test No. 6. SAM-TR66-55. Tech Rep SAM-TR, 1966, 1–12. [10] S.R. Silverman, I.J. Hirsh, Problems related to the use of speech in clinical audiometry, Ann. Otol. Rhinol. Laryngol. 64 (4) (1955) 1234–1244. [11] I.J. Hirsh, H. Davis, S.R. Silverman, E.G. Reynolds, E. Eldert, et al., Development of materials for speech audiometry, J. Speech Hear Disord. 17 (3) (1952) 321–337. [12] H.A. Haskins, A Phonetically Balanced Test of Speech Discrimination in Children, Northwestern University, Evanston, 1949. [13] S. Jerger, S. Lewis, J. Hawkins, J. Jerger, Pediatric speech intelligibility test. I. Generation of test materials, Int. J. Pediatr. Otorhinolaryngol. 2 (3) (1980) 217–230. [14] M. Talarico, G. Abdilla, M. Aliferis, I. Balazic, I. Giaprakis, et al., Effect of age and cognition on childhood speech in noise perception abilities, Audiol. Neurotol. 12 (1) (2007) 13–19. [15] Y. Zheng, K. Wang, D. Li, Z.L. Meng, Evaluation of speech perception in competing noise conditions for normally hearing children, Noise Health 15 (64) (2013) 178– 182. [16] K. Nishi, D.E. Lewis, B.M. Hoover, S. Choi, P.G. Stelmachowicz, Children’s recognition of American English consonants in noise, J. Acoust. Soc. Am. 127 (5) (2010) 3177–3188. [17] L.J. Leibold, E. Buss, Children’s identification of consonants in a speech-shaped noise or a two-talker masker, J. Speech Lang. Hear Res. 56 (4) (2013) 1144–1155. [18] R.H. Wilson, N.M. Farmer, A. Gandhi, E. Shelburne, J. Weaver, Normative data for the Words-in-Noise Test for 6- to 12-year-old children, J. Speech Lang. Hear Res. 53 (5) (2010) 1111–1121. [19] K.I. Kirk, L.S. Eisenberg, A.S. Martinez, M. Hay-McCucheon, The Lexical Neighborhood Test Test–Retest Reliability and Inter-list Equivalency, Indiana University, 1998. [20] K.I. Kirk, D.B. Pisoni, M.J. Osberger, Lexical effects on spoken word recognition by pediatric cochlear implant users, Ear Hear. 16 (5) (1995) 470–481. [21] M.S. Cluff, P.A. Luce, Similarity neighborhoods of spoken two-syllable words: retroactive effects on multiple activation, J. Exp. Psychol. Hum. Percept. Perform. 16 (3) (1990) 551–563. [22] K.C. Yuen, I.H. Ng, B.P. Luk, S.K. Chan, S.C. Chan, et al., The development of Cantonese Lexical Neighborhood Test: a pilot study, Int. J. Pediatr. Otorhinolaryngol. 72 (7) (2008) 1121–1129. [23] M.S. Vitevitch, E. Rodrı´guez, Neighborhood density effects in spoken word recognition in Spanish, J. Multiling. Commun. Disord. 3 (1) (2005) 64–73. [24] H.M. Yang, J.L. Wu, Y.H. Lin, Y.J. Sher, Development of Mandarin monosyllabic lexical neighborhood test (LNT), Cochlear Implants Int. 5 (Suppl. 1) (2004) 203–205. [25] C. Liu, S. Liu, N. Zhang, Y. Yang, Y. Kong, et al., Standard-Chinese Lexical Neighborhood Test in normal-hearing young children, Int. J. Pediatr. Otorhinolaryngol. 75 (6) (2011) 774–781. [26] H. Liu, S. Liu, S. Wang, C. Liu, Y. Kong, et al., Effects of lexical characteristics and demographic factors on Mandarin Chinese open-set word recognition in children with cochlear implants, Ear Hear. 34 (2) (2013) 221–228. [27] S. Wang, S. Liu, Y. Kong, H. Liu, J. Feng, et al., Psychometric properties of the StandardChinese lexical neighborhood test, Acta Otolaryngol. 134 (1) (2014) 66–72. [28] L.S. Eisenberg, A.S. Martinez, S.R. Holowecky, S. Pogorelsky, Recognition of lexically controlled words and sentences by children with normal hearing and children with cochlear implants, Ear Hear. 23 (5) (2002) 450–462. [29] V. Krull, S. Choi, K.I. Kirk, L. Prusick, B. French, Lexical effects on spoken-word recognition in children with normal hearing, Ear Hear. 31 (1) (2010) 102–114. [30] A.Y. Bonino, L.J. Leibold, E. Buss, Release from perceptual masking for children and adults, Ear Hear. 34 (1) (2013) 3–14. [31] F.L. Wightman, D.J. Kistler, Informational masking of speech in children: effects of ipsilateral and contralateral distracters, J. Acoust. Soc. Am. 118 (5) (2005) 3164–3176.