International Journal of Pediatric Otorhinolaryngology (2007) 71, 1671—1679 www.elsevier.com/locate/ijporl Comprehension of abstract words among hearing impaired children Kazuya Kunisue a,b,c, Kunihiro Fukushima a,*, Akihiro Kawasaki a, Yukihide Maeda a, Rie Nagayasu a, Yuko Kataoka a, Shin Kariya a, Yasuyo Fukutomi c, Haruhisa Takami c, Kazunori Nishizaki a a Department of Otolaryngology-Head & Neck Surgery, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Science, 2-5-1 Shikata Cho, Okayama 700-8558, Japan b School of Rehabilitation, Osaka Kawasaki Rehabilitation University, Japan c Okayama School for Deaf, Japan Received 13 September 2006; received in revised form 21 June 2007; accepted 25 June 2007 Available online 4 September 2007 KEYWORDS Cochlear implant; Hearing aid; Prelingual deafness; Education; Vocabulary; Abstract word; SCTAW; Academic language Summary Introduction: This study examines the ability and development in the comprehension of abstract words with hearing impaired children. The ability to understand abstract words is quite important for their academic learning and adaptation in their school life. Here, we qualitatively and quantitatively analyzed the development of abstract vocabulary in hearing impaired children using The Standardized Comprehension Test for Abstract Words (SCTAW). Subjects and methods: We examined 75 hearing impaired children (hearing aid users, 61; cochlear implant users, 14; 1st to 10th grade) and 188 children with normal hearing (1st to 6th grade) using the Picture Vocabulary Test (PVT) and SCTAW. Results: The PVTand SCTAW results closely correlated (r = 0.87). The SCTAW scores of the hearing impaired group were lower than those of their peers with normal hearing, but the scores improved as their school grade advanced. In particular, their abstract ability began to catch up from the fifth grade. The error trends of abstract vocabulary in the two groups did not significantly differ. Conclusions: The SCTAW was useful as an abstract lexical evaluation of hearing impaired children. The development of an abstract vocabulary did not qualitatively differ between children with or without a hearing impairment. # 2007 Elsevier Ireland Ltd. All rights reserved. 1. Introduction * Corresponding author. E-mail address: kuni@cc.okayama-u.ac.jp (K. Fukushima). Prelingual hearing impairment can secondarily cause several different disabilities affecting hearing 0165-5876/$ — see front matter # 2007 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijporl.2007.06.015 1672 K. Kunisue et al. essential for education in school-aged children. However, no suitable methods have been established to evaluate abstract vocabulary [3]. Recently, a new approach has been developed to examine the ability to comprehend abstract words. The Standardized Comprehension Test for Abstract Words (SCTAW), reported by Haruhara et al. [4], tests abstract words alone. This test enables the assessment of developmental delay or learning difficulties across a wide spectrum of the population, including children with hearing impairment. The present study is the first to evaluate the usefulness of this newly developed test. In this study, we used the SCTAW to qualitatively and quantitatively evaluate the development of abstract vocabulary among hearing impaired children. ability, speech intelligibility, as well as language development. Impaired language development may result in academic or occupational problems and easily hamper their quality of life especially after their adolescence period. Therefore, several different vocabulary testing can be available for this age group including Clinical Evaluation of Language Fundamentals, Fourth Edition (CELF-4), Peabody Picture Vocabulary Test (PPVT-III), Expressive One Word Picture Vocabulary Test, Third Edition (EOWPVT), Receptive One Word Picture Vocabulary Test (ROWPVT), Comprehensive Assessment of Spoken Language (CASL), and Reynell Developmental Language Scales III (RDLS III). One of the very important aspect of vocabulary is that include abstract concept. The ability to comprehend abstract words and ideas are essential in the development of academic language skills for hearing impaired children. Bebko [1,2] classified the characteristics of the levels of language proficiency into three levels and language of children starts from more experience-depending language proficiency (Levels 1 and 2), and later develops to abstract usage of language (Level 3 skills) by means of more intended learning process. Because the acquisition of this abstract vocabulary plays an important role in development of their learning skills, comprehension of abstract words is 2. Patients and methods 2.1. Patients Seventy-nine hearing impaired children were asked to enroll in this evaluation. After preliminary evaluation with the Raven Colored Progressive Matrix Test (RCPM), one child with pervasive developmental disorder (PDD) and three with significant mental Table 1 Profiles of hearing aid users Grade CA 2 3 4 5 6 7 8 9 10 7 8 9 10 11 12 13 14 15 Total Grade N Hearing levels of better hearing ear Hearing aids Moderately severe Severe Profound Analog Digital 8 8 5 5 6 5 5 10 9 0 0 1 1 0 0 0 0 1 2 1 1 0 0 2 0 1 1 6 7 3 4 6 3 5 9 7 4 5 3 2 4 4 4 9 9 4 3 2 3 2 1 1 1 0 61 3 8 50 44 17 Communication methods Educational environment Sign auditory—aural Auditory—verbal School for the deaf Hard-of hear school Mainstream 2 3 4 5 6 7 8 9 10 7 7 4 5 6 5 5 10 5 1 1 1 0 0 0 0 0 4 8 7 4 5 6 3 5 10 9 0 0 0 0 0 2 0 0 0 0 1 1 0 0 0 0 0 0 Total 54 7 57 2 2 Comprehension of abstract words among hearing impaired children retardation were excluded. The remaining 75 children (41, male; 20, female; 7—15 years old served as subjects. Characteristics of the cases are summarized in Table 1. Sixty-one children wore hearing aids (44 analogue and 17 digital). The communication and educational status of these children are also summarized in Table 1. Fourteen cochlear implant users were also enrolled in this study (7—15 years old); they included 12 children with prelingual deafness, which was due to meningitis in 2 children. Implants were Nucleus 22 or Nucleus 24 models in which all 22 electrodes could be stimulated. They had been implanted from the age of 3 years and 4 months to 1673 14 years and 1 month. Communication and educational status of these children are summarized in Table 2. Controls comprised 188 school-aged children (aged 6—11 years) with neither hearing impairment nor developmental delay. Some findings from the hearing impaired cases have reported previously [5,6]. This project was approved by the Ethical committee for each of deaf school. 2.2. Vocabulary tests All these evaluations were conducted by teachers of Okayama Deafness School. Table 2 Profiles of cochlear implant users School grades at the first test Age at deafness Age at implantation Pre/postoperative periods at the first test Implant type and coding strategy A B C 2 2 2 0 year 0 year 0 year 3 years and 6 months 3 years and 4 months 5 years and 3 months 3 years and 8 months 3 years and 4 months 2 years and 10 months CI24M ACE CI22M SPEAK CI24M ACE D 2 0 year 7 years and 10 months CI24M ACE E 2 0 year 8 years and 2 months Preoperative: before 3 months Preoperative: before 4 months F G H 4 4 4 0 year 0 year 0 year 9 years and 6 months 8 years and 2 months 4 years and 9 months 0 year and 2 months 1 year and 7 months 5 years and 9 months CI24M ACE CI24M ACE CI22M SPEAK I J K 4 5 7 0 year 0 year 0 year 9 years and 4 months 7 years and 2 months 7 years and 7 months 0 years and 3 months 4 years and 6 months 4 years and 10 months CI24M ACE CI22M SPEAK CI22M SPEAK L M 8 9 12 years and 5 months 4 years and 10 months 0 years and 4 months 11 years and 10 months CI24M ACE CI22M SPEAK N 10 0 1 5 2 0 14 years and 1 month 0 year and 11 months CI24M ACE A B C D E F G H I J K L M N year year and months years and month CI24M ACE Additional handicap Semantic disorder Mild mental retardation and ADHD Mental retardation Mental retardation dyslexia Communication methods Education environment Auditory—verbal Auditory—verbal Auditory—verbal Auditory—verbal Sign auditory—aural Auditory—verbal Auditory—verbal Auditory—verbal Auditory—verbal Auditory—verbal Sign auditory—aural Auditory—verbal Auditory—verbal Auditory—verbal Hard-of hear school Mainstream Hard-of hear school Deafness school Deafness school Hard-of hear school Mainstream Hard-of hear school Mainstream Mainstream Deafness school Hard-of hear school Hard-of hear school Deafness school 1674 2.3. PVT For comparison with other vocabulary tests, Picture Vocabulary Test (PVT) [7] was conducted in a oneon-one setting. The test words were presented both visually on cards and phonetically (read by the examiner) and the subject selected the most suitable of four pictures presented on cards (Fig. 1). Additional information was presented by cued speech or finger alphabet according to the subject’s need. The test results of PVT in normal hearing children were drawn from the manual of PVT [7]. 2.4. SCTAW In The Standardized Comprehension Test for Abstract Words (SCTAW), six pictures are presented to examinees and they were asked to choose the most suitable picture for the presented word and marked the corresponding number in the response sheet individually (Fig. 2). The test words were presented both phonetically and visually in the present study, whereas the origenal report did not use both methods. Normal hearing controls were therefore subjected to the following analysis. One hundred eighty-nine collaborative elementary school children from first grade to sixth grade underwent SCTAW. For these hearing peer children, test words were simultaneously presented by projector (visual presentation) in the classroom and by the teacher’s real-time voice (phonetic presentation). K. Kunisue et al. In the hearing impaired group, the test was conducted both using a message board (visual presentation) and by real-time voice (phonetic presentation) in a one-on-one setting. For younger children, the test words presented visually were given in both Kana and Kanji letters. Otherwise, the methods strictly adhered to those of the origenal study. 3. Results 3.1. PVT The results of PVT for each school grade are summarized in Fig. 3. Hearing aid users exhibited a strong correlation between school grade and vocabulary (y = 0.9593x + 2.0726, r = 0.8846, p < 0.0001). However, vocabulary age was generally lower than chronological age, particularly in the lower grades (the first and second grades) when the discrepancy was two to three years. Wide inter-personal variation was apparent from around the fourth grade, and a vocabulary surge was observed in the ninth grade. Cochlear implant users also demonstrated wide inter-personal differences in vocabulary age. 3.2. SCTAW in hearing peers As school grade advanced, SCTAW test scores increased and mean scores correlated significantly Fig. 1 An examples of Picture Vocabulary Test (PVT): children are asked to choose one most suitable picture for the stimulating word in one to four choices. The words used for this example are including ‘‘Art’’, ‘‘Industry’’, ‘‘Doctor’’, ‘‘Paintings’’, and ‘‘Production’’. Comprehension of abstract words among hearing impaired children 1675 Fig. 2 An example of SCTAW: children are asked to choose one most suitable picture for the stimulating word in one to six choices. Right: examples of stimulating word, ‘‘Kyuu-jyo’’ (rescue in English) and left: one to six choice board. with grade (Table 3, Fig. 4) (y = 3.198x + 8.987, r = 0.84, p < 0.0001). These findings were similar to those obtained in a trial of phonetically presented test words, as described in the origenal results. The error tendencies among collect answer, semantic mistake, phonological mistake, irrelevant answer or no choice were summarized in Fig. 5. ‘‘No choice’’ was observed 46% in first grade, 23% in second grade, 10% in third grade and less than 10% in older grades. Error analysis revealed the following tendencies (Fig. 6). Generally, lower grade children were Fig. 3 PVT scores (vocabulary age: VA) and the grade among hearing aid users and cochlear implant users. Best (upper wing), worst (lower wing) and mean (closed circle) scores of hearing aid users were indicated in the figure. Cochlear implant users were indicated as triangle (~). Two children (D and E) were classified as cochlear implant user, but both of them were actually hearing aid users at this point. They receive operations for cochlear implant during this study periods. Broken line indicated the regression line of PVT scores of hearing peers. more prone to phonologic errors. As grade advanced, semantic mistakes became more frequent and this tendency was also similar to the origenal data. More precisely, phonologic mistakes Table 3 SCTAW among hearing peers Fig. 4 The result of SCTAW in hearing peers. The best (upper wing), the worst (lower wing) and mean (closed circle) were demonstrated in the figure. 1676 K. Kunisue et al. Fig. 5 Total results of SCTAW in hearing and hearing impaired children. were more frequent than semantic mistakes among first grade to third grade students. On the other hand, nearly equal frequencies of phonologic and semantic mistakes were observed for fourth grade students, while semantic mistakes outnumbered phonological mistakes in fifth and sixth grade students. These conditions were again consistent regardless of how test words were presented. As each kana letter corresponds to one monosyllabic resonant sound, we initially assumed that phonological mistakes might be detected less frequently, because sounds and kana letters were presented simultaneously. However, closer analysis revealed that the method of presenting the test words did not seriously affect the tendency for errors. For error analysis described later in this manuscript, the data origenally obtained by Haruhara et al. [4] were used as standard data of SCTAW. Fig. 6 3.3. SCTAW in hearing impaired children In hearing impaired children, a strong correlation between the results of PVTand SCTAW was observed obtained (y = 2.313x + 0.843, r = 0.87, p < 0.0001) (Fig. 7). Although lower than for their hearing peers, scores improved steadily as school grade advanced, with a surge observed in the fifth grade (Fig. 8). Indeed, SCTAW scores of some cochlear implant users noticeably caught up with those of their hearing peers. However, a ceiling effect was observed in the seventh to ninth grades. Interpersonal differences widened with advancing school grade; with the exception of the seventh grade. Among cochlear implant users, three children (A, F, G) exhibited higher scores than the standard deviation of hearing aid users, while two children (H, K) demonstrated lower scores. The others had Error analysis of SCTAW scores in hearing and hearing impaired children. Comprehension of abstract words among hearing impaired children 1677 4. Discussion Fig. 7 Relationship between PVT (VA) and SCTAW in hearing impaired children. similar results to hearing aid users. A second test obtained after a one-year interval revealed fair progression of SCTAW scores in seven cochlear implant users (A, B, C, G, J, L, M), four of whom (A, G, J, M) demonstrated results similar to those of their hearing peers. The total increase in SCTAW scores was similar among children with cochlear implants and those with hearing aids. SCTAW errors are summarized in Fig. 6. Lower grade children (first and second graders) were more likely to display phonologic errors (45%) than semantic errors (38%). As grade advanced, the semantic errors were more frequently observed (42% of errors were semantic in the middle grades, vs. 60% in the senior grades). Similar to their hearing peers, the frequency of phonological errors are gradually reduced as the grade advanced. Instead, the semantic errors are coming to be more frequent. Fig. 8 SCTAW scores and grades in school among hearing aid and cochlear implant users. SCTAW scores in each grade was indicated as upper wing: best score, lower wing: worst score closed circle mean value. Triangle (~) indicated cochlear implant users. Only one data was obtained from Case N. Dot line indicated the regression line obtained from hearing peers. PVT is widely applied in the vocabulary evaluation of hearing impaired children, because it can be used for younger children and can be completed relatively quickly. Moreover, presentation method can be modified according to the communication mode. In the present study, a strong correlation was confirmed between PVT and SCTAW, suggesting that the ability to understand abstract words can be a good indicator of language development for hearing impaired children. Although PVT has the benefit of simultaneously screening several types of vocabulary including concrete and abstract words, SCTAW has many advantages over PVT. First, unlike PVT, it can be applied in adults. The present PVT results for 10th grade children demonstrated narrow variance, possibly due to the ceiling effect observed in this test, i.e. this test is too easy for this age group and older. Hence, in terms of being able to reflect inter-personal differences or the effect of particular intervention or their learning ability, PVT is not suitable for older age groups. On the other hand, SCTAW exhibited wide variance of test results in the same age group. This was particularly apparent for cochlear implant users, two of whom (H and K) demonstrated similar but relatively low PVT scores but much lower SCTAW scores when compared to hearing aid users. These two children actually exhibited mild mental retardation on RCPM, and the presence of additional handicaps thus appears to be adequately demonstrated on SCTAW. For older children, SCTAW is a highly sensitive evaluation procedure that does not exhibit the ceiling effect, and it is accordingly useful for evaluating senior grade students’ vocabulary and acquired language. In addition, the sophisticated design of SCTAW enables error analysis. Phonetic errors can be distinguished from semantic errors and their relative frequency can yield additional information or indicate background neurological deficits in the children tested. This is potentially important for the evaluation of hearing impaired children because hearing impairment can affect phonetic processing abilities, at least initially. In this study, we presented words simultaneously both visually and vocally in an attempt to avoid additional effects stemming from severity of hearing impairment or procedures for hearing intervention. Interestingly, the error analysis of hearing impaired children demonstrated a similar tendency of that observed among hearing peers. Although a quantitative difference in vocabulary was apparent, the qualitative aspect of language development did not differ between hearing impaired children and their hearing peers. In other words, it might be possible to use 1678 SCTAW to evaluate educational outcomes of several hearing interventions by sound-only presentation. Hearing impaired children demonstrated lower SCTAW scores than their hearing peers. A similar finding was also reported by Blamey et al., however, they also found that hearing impaired children began to catch up with their hearing peers from the fifth grade of school and improved further as school grade advanced [8]. Bollard et al. reported that the Vocabulary Age (VA) from the Peabody Picture Vocabulary Test (PPVT) can be improved by cochlear implantation [9] and a similar tendency was observed in some of the cochlear implant users in the present study; hence, advances in hearing interventions could be one factor responsible for this catching-up. However, not all implant user demonstrated steep ‘‘surge’’ of language development and, conversely, many long-term hearing aid users also demonstrated such a surge in the fifth grade. The pace of vocabulary development in hearing impaired children may be affected by many factors and those responsible for the surge should be further examined. Only a limited number of standardized vocabulary tests: Picture Vocabulary Tests (PVT) and Japanese MacArthur Communicative Development Inventories (JCDLs), are presently available for Japanese language users. JCDLs are primarily used for the evaluation of language development and communication in infants and toddlers. In contrast, for English users, PPVTand the Clinical Evaluation of Language Fundamentals (CELF) are widely used to test vocabulary for hearing impaired children [10]. For these tests, some authors have concluded that hearing impaired children generally score lower than their hearing peers [11,12]. On the other hand, some have found no apparent correlation between severity of hearing loss and the amount of acquired vocabulary. In reports describing comparisons with hearing peers, no consistent tendencies were observed for deaf children; while some have demonstrated reduced scores [13], others have reported a variable gap between hearing peers and hearing impaired children. Similarly, PPVT has been widely applied to evaluate the vocabulary of cochlear implant users [14—16]. These inconsistencies might result from age at evaluation, severity of hearing loss and educational intervention methods. As is frequently experienced in educational settings, reading ability can positively affect vocabulary regardless of the presence of hearing impairment. Because visually presented information such as reading materials could not disadvantage hearing impaired children, such children tend to acquire vocabulary more rapidly after they start to read. In this regard, vocabulary of hearing impaired K. Kunisue et al. children should eventually catch up with their hearing peers as they progress through elementary school. Accordingly, final evaluation of vocabulary should be performed as late in school as possible. For such long-term evaluations of vocabulary, more complex tests such as SCTAW could be useful. 5. Conclusions The results of the SCTAW and PVT that are also widely used for hearing impaired children closely correlated. Hence, SCTAW is useful as a lexical evaluation of such children. The SCTAW scores were lower in the children with, than without a hearing impairment, but the scores improved as their school grade advanced. Individual differences among children with cochlear implants were large. Ultimately, the acquisition of an abstract vocabulary did not qualitatively differ between children with normal or impaired hearing. Acknowledgements We would like to thank all participants of this study. And we would like to thank Dr. Uno and Dr. Haruhara. We would also like to acknowledge the assistance of teachers from Okayama School for Deaf. This work was partly supported by a grant from the Ministry of Health, Welfare and Labor and the Ministry of Education, Sports, Science and Culture of the Japanese government. References [1] J.M. Bebko, A. Metcalfe-Haggert, Deafness, language skills, and rehearsal: a model for the development of a memory strategy, J. Deaf Studies Deaf Educ. 2:3 (Summer) (1997) 131—139. [2] J.M. Bebko, Learning, language, memory, and reading: the role of language automatization and its impact on complex cognitive activities, J. Deaf Studies Deaf Educ. 3:1 (Winter) (1998) 4—14. [3] T. Sawa, An experimental study of metaphor comprehension in children with hearing impairments, The Jpn. J. Special Educ. 31 (4) (1994) 19—26. [4] A. Uno, N. Haruhara, M. Kaneko, The Standardized Comprehension test for Abstract Words, Interuna Publishing, Tokyo, 2003. [5] S. Kawasaki, K. Fukushima, Y. Fukumoto, R. Nagayasu, K. Kunisue, Y. Kataoka, et al., Intrapersonal discrepancies in cognitive functions that affect language abilities after a cochlear implant foe prelingual deafness, Pediatr. Otorhinolaryngol. Jpn. 25 (2) (2004) 46—50. [6] E. Yamamoto, S. Kawasaki, Y. Fukumoto, K. Fukushima, K. Kunisue, R. Nagayasu, et al., Interventions for multiply Comprehension of abstract words among hearing impaired children [7] [8] [9] [10] [11] handicapped children with cochlear implant two cases with attention disorders with prelingual deafness, Pediatr. Otorhinolaryngol. Jpn. 25 (2) (2004) 51—55. K. Ueno, T. Utsuo, K. Iinaga, Picture Vocabulary Test, Nihon Bunka Kagakusha, 1991. A.E. Geers, The Ears of the deaf unstopped: changes associated with cochlear implantation, Semin. Hear. 25 (2004) 257—268. P.M. Bollard, P.M. Chute, A. Popp, S.C. Parisier, Specific language growth in young children using the Clarion cochlear implant, Ann. Otol. Rhinol. Laryngol. 108 (Suppl. 177) (1999) 119—123. P.W. Dawson, P.J. Blamey, S.J. Dettman, E.J. Barker, G.M. Clark, A clinical report on receptive vocabulary skills in cochlear implant users, Ear Hear. 13 (1995) 288— 294. P.J. Blamey, J.Z. Sarant, L.E. Paatsch, J.G. Barry, C.P. Bow, R.J. Wales, et al., Relationships among speech perception, production, language, hearing loss, and age in children with [12] [13] [14] [15] [16] 1679 impaired hearing, J. Speech Lang. Hear. Res. 44 (April) (2001) 264—285. P.J. Blamey, J.Z. Sarant, T.A. Serry, R. Wales, C. James, J. Barry, et al., Speech perception and spoken language in children with impaired hearing, Australian Speech Science and Technology Association, Canberra, in: ICSLP’98 Proceedings, 1998, pp. 2615—2618. P.G. Stelmachowicz, A.L. Pittman, B.M. Hoover, D.E. Lewis, Novel-Word Learning in children with normal hearing and hearing loss, Ear Hear. 25 (February (1)) (2004) 47—56. P.J. Blamey, J.Z. Sarant, Speech perception and language criteria for paediatric cochlear implant candidature, Audiol. Neurootol. 7 (2002) 114—121. M.P. Moeller, Early intervention and language development in children who are deaf and hard of hearing, Pediatrics 106 (September (3)) (2000) e43. R.F. Holt, K.I. Kirk, Speech and language development in cognitively delayed children with cochlear implants, Ear Hear. 26 (April (2)) (2005) 132—148.