I. INTRODUCTION
Language has an important role in our life. By means of language we can express our ideas, feelings and needs. We can describe events, teach and learn, share experiences and ideas. Language permits knowledge to be passed from one generation to another (Bench, 1993). Because in our daily life we communicate orally, language is mostly thought synonymously with speech.
Normal hearing children learn their mother language without being aware of it but severe or profound hearing loss in early infancy significantly affects the development of spoken language of hearing impaired children and normal development of speech is often disrupted (Tüfekçioglu, 1996; Osberger and McGarr, 1982). The breakdown in the process of communication may result in difficulties in emotional and social behavior, beside these the child may encounter problems in educational or in vocational environments (Tüfekçioglu, 1996; Sanders, D., 1971).
Many children with severe or profound hearing loss acquire excellent speech if they are diagnosed early and fitted with suitable hearing aids; consistently use of hearing aids and if they are educated in a normal speaking environment (Abberton, 1988; Clark, 1986; Wood, 1984).
However, profoundly hearing-impaired children show a wide range of spoken language abilities, some have highly intelligible speech while others have unintelligible speech (Abberton, 1990; Monsen, 1983). This is due to hearing impaired children's speech production. While a hearing impaired child speaks, segmental and prosodic errors occurs, the intelligibility of his or her speech will be affected.
The most common segmental errors are:
Reason of poor articulation in the speech of hearing-impaired children is slow articulatory movement and slow articulation of syllables (Monsen, 1978; Osberger & McGarr, 1982).
Prosodic errors:
I. Poor control of fundamental frequency can result in:
II. Abnormal voice characteristics such as harshness, breathiness and nasality.
III. Abnormal uses of pauses: inappropriate place and uses longer pause duration at sentences.
IV. Poor respiratory control.
V. Speaking much slower than normal. (Osberger & Megane, 1982; Abberton, 1990)
Most profoundly hearing impaired children have residual hearing for frequencies below 1000 Hz., therefore they should be able to perceive prosodic information and use it (Engen, 1983). By the help of prosodic clues such as intonation, rhythm, stress and duration, they can develop their mother language and produce intelligible speech (Tüfekçioglu, 1989; John & Howarth, 1965; Engen, 1983).
Speech Intelligibility of Hearing-Impaired Children
The goal of the education of the hearing-impaired children is to developed good speaking abilities as the importance of speech intelligibility in the hearing world can't be denied (Monsen, 1983).
The degree of oral speech communication skills of the hearing impaired children can be tested by means of speech intelligibility (Mets, 1985).
The term speech intelligibility can be explained by how much of the child's speech can be understood by a listener (Osberger, 1982; Monsen, 1983).
The intelligibility of hearing-impaired talkers' speech can be measured in two ways:
Item Identification: Speech samples of hearing impaired children consist of words or sentences. Words are audio-recorded subsequently, randomized and played back to a panel of listeners. The listeners write down what they believe the talker has said. The score can be found by the percentage of words correctly understood (Markides, 1978).
Scaling: In this method a panel of listeners are asked to rate the speech intelligibility of each hearing impaired child on a six or five point scale: between normal to unintelligible, or very intelligible to would not attempt to use speech (Osberger, 1992).
Factors Affecting Intelligibility Measurements:
Speech intelligibility scores are affected by the speech material used, the characteristics of the listener, and the characteristics of the talkers.
The findings in the literature show that sentences intelligibility scores are higher than single word scores and spontaneous speech samples scores are higher than the read sample score. (Davis & Silverman, 1978; Markides, 1983; Conrad, 1979; McGarr, 1983).
The listeners are affected by the speech quality of hearing-impaired child, the visibility of the talker, listener's experience of hearing impaired child's speech and the repetition of the utterances to the listener (Monsen, 1983; Markides, 1982; John, 1976; Hudgins & Numbers, 1942).
There is an important relation ship between speech intelligibility and speech production skills of hearing impaired children. Past studies in this field shows that two factors were found to be strongly related to speech intelligibility and segmental production, and prosodic production (control of overall sentences duration). Hearing-impaired children with about the same number of segmental errors can exhibit wide ranges of speech intelligibility (Osberger, 1992). Most probably these differences are due to the influence of prosodic features of speech (Hudgins & Numbers, 1942; John, 1965). The direct relationship between intelligibility and prosodic features of speech can not be rejected.
Some of the hearing-impaired children omit or substitute consonants, neutralize vowels, and speak either monotonously or slower than normal and uses much more breaks than normal speaker. All these segmental and prosodic speech characteristics have an effect on the speech intelligibility (Hickerson, 1975; John, 1965; McGarr, 1983).
The aim of this research is to investigate the effect of some of the prosodic characteristics like sentence duration, pitch range, average pitch and hearing loss on speech intelligibility of the Turkish hearing-impaired children.
II. METOD
Subjects:
A group of 33 hearing female children were selected from the school population of the normal secondary school. All of these children had normal speech production.
Second group of 35 hearing impaired female children were all educated in a natural auditory-oral approach (Center for Hearing-Impaired School - I.Ç.E.M.). All had congenital/prelingual sensorineural hearing loss and had no handicap other than deafness. The mean of the puretone average for the 35 children was 103 dB (ISO) in the better ear. All of these students were fitted with personal hearing aids.
Materials for speech production
Ten simple sentences were constructed, taking into account the vocabulary and language abilities of these hearing impaired children considered and complex sentences did not used (John, 1976). The sentences were made up three to six words. Each sentences were uttered by the hearing and the hearing impaired children and recorded using the Laryngograph in a sound treated room.
Materials for speech intelligibility and listening jury:
For setting intelligibility scores 18 pictures were chosen. The pictures could be explained by simple sentences. Each child chose 15 pictures randomly. They produced a sentence for each picture and speech samples of these hearing-impaired children were recorded to Cp 230 Marantz tape recorder in a sound-treated room. 15 second pose was given between each sentences so the listeners could stop the tape and re-listen to the sentences three times. Then the listeners wrote down what they think to they heard.
Each child's ten sentences were listened by three listeners. The number of the words which were correctly understood by the listeners was divided to the number of the words which were produced by the hearing-impaired child and the intelligibility ratio of each child was found (Conrod, 1975; Markides, 1983; Carl, 1975; Monsen, 1978).
15 experienced listeners have participated in the present study. An experienced listener was defined as a person who had more than two years of experience in listening to the speech of the hearing impaired children (Mostand & Shugi, 1993). The listeners who were participated in this study have been working as teachers of hearing-impaired for more than three years.
Analyzing data
For comparing the sentences duration, pitch range and average pitch of hearing impaired children against normal hearing children: The duration mean of each children's sentences was found out by dividing total of 10 sentences duration to 10. Then duration means of 35 hearing-impaired children's sentences were compared against the duration means of 33 normal hearing children's sentences with "t" test (Table 1).
Each normal hearing and hearing-impaired child's pitch range and average pitch mean was found out by the help of Laringograph and compared with "t" test (Table 2,3).
Table 1. Compression of sentences duration of hearing and hearing-impaired children.
| n | mean (second) | SD | t-value | Level of significance | |
| Hearing impared | 35 | 3,126 | .564 | ||
| 8.654 | p=.0005 | ||||
| Normal Hearing | 33 | 2,212 | .231 |
Table 1 compares the sentences duration hearing-impaired children and normal hearing children. It shows that there is a statistically significant difference at 0.0005 level. This finding shows that hearing-impaired children's speech is slower than normal hearing children.
Table 2. Comparison of pitch range of hearing and hearing-impaired children.
| n | mean (Hz) | SD | t-value | Level of significance | |
| Hearing impared | 35 | 86.699 | 20.446 | ||
| 5.79 | p=.0005 | ||||
| Normal Hearing | 33 | 115.105 | 19.976 |
Table 2 compares the Pitch range (F0) of hearing-impaired children and hearing children. It shows that there is a statistically significant difference at 0.0005 level. This finding shows that hearing impaired children use narrower pitch range compared to hearing children.
Table 3. Comparison of average pitch of hearing and hearing-impaired children.
| n | mean (Hz) | SD | t-value | Level of significance | |
| Hearing impared | 35 | 06.114 | 47.475 | ||
| 2.017 | 01<po.025 | ||||
| Normal Hearing | 33 | 286.755 | 28.872 |
Table 3. Compares the average pitch of hearing-impaired children and hearing children. It shows that there is a statistically significant difference of 0.0005 level, that hearing-impaired children use higher pitch than normal.
Intelligibility Analysis:
Spearman rank correlation and Pearson correlation analysis were used to find out the relationship between the intelligibility and sentences duration, pitch range and hearing loss. Correlation analysis is the technique used most often because it is the most practical approach but it does not give causal relationship (Osberger, 1992).
For Spearmen rank correlation test, 330 sentences which were produced by normal hearing children were used to find out mean of normal children's sentences' duration, pitch range and average pitch. 35 hearing impaired children were ranked according to this criteria (Downief&Weath, 1970; Gravetter&Wallnau, 1988).
Table 4. Relation between intelligibility and sentences duration, pitch range and average pitch
| n | r s | p | r | p | ||
| Sentence Duration | 35 | .782 | .0001 | -791 | * | |
| Pitch Range | 35 | .612 | .0004 | 525 | * | |
| Average Pitch | 35 | .0065 | .97 | 313 | * |
In table 4 the results of Spearman Rank Correlation and Pearson correlation are given: There is statistically significant high negative correlation between p=.05 sentences duration and speech intelligibility (r=-79.5 and Sr=782 p=.0001). We can say that the children whose speech is faster or whose speech rate nearer to normal speech rate has more intelligible speech.
There is statistically significant high relationship between pitch range and intelligibility according to Spearman Rank Correlation (rs. 617 p=.0004) but there is moderate positive relation between pitch range and intelligibility according to Pearson Correlation (r=.525 p=.05). We can say that the hearing impaired child who has same or nearer to normal hearing children's pitch range has better intelligibility. Some hearing impaired children use much wider pitch range than normal because they could not control their speech organ properly, so the result of Pearson correlation can be effected. Because of using wider pitch range.
There is no statististicaly significant relation between average pitch and intelligibility in both correlation (r=313 p=.05, rs .0065 p=.97).
Table 5. Relation between intelligibility and average hearing loss of hearing-impaired children
| N | r | z(n-2) | P |
| 35 | 3.01 | 3.01 | =.05 |
There is statisticaly significant moderate relation between average hearing loss and intelligibility. Most of the present studies find high relation ship between average hearing loss and intelligibility (Markides, 1985; Osberger, 1982). In this study we could not find high relation ship, this is due to the fact that all the subjects were diagnosed early, hearing aids were fitted and educated in a auditory oral approach and in natural language settings.
III. DISCUSSION
The present results show that there is statistically significant differences between Turkish normal hearing and hearing-impaired children's speech as hearing-impaired children speak slower, use narrower pitch range and higher pitch than normal speakers. The speech intelligibility could be affected by these speech characteristics of hearing-impaired children.
When we look at the relation ship between intelligibility and prosodic feature, there is a high relation ship between speech duration and intelligibility. We could say that the hearing-impaired children whose speech rate nearer to normal or who speaks near to normal rate produce more intelligible speech. At same time there is a high relation ship between intelligibility and pitch range. Intelligibility of hearing-impaired children improve when they produce pitch range nearer to normal pitch range.
There is moderate relation ship found in this study between intelligibility and average hearing loss.
Most of the profoundly hearing-impaired children have some residual hearing. By the help of residual hearing, hearing-impaired children can use acoustic information of prosody to make linguistic distinction and can acquire their mother language (Engen, 1983). So hearing-impaired children must be taught to combine the segmental sounds into connected speech and to use appropriate prosodic features in their oral speech (Bench, 1993).
The hearing-impaired children must be thought to combine speech sounds into connected speech in natural communication environments. Because the speech sounds have little meaning individually, we can get acoustic information of prosody by means of connected speech.
Certainly, further studies are needed to be done on speech production features which would affect the intelligibility of Turkish hearing-impaired speakers.
Reference
ABBERTON, E., HAZAN, V. and FOURCIN, A.: "The development of contrastiveness in profoundly deaf children's speech," Clinical Linguistic and Phonetica, Vol. 4, No. 3, 1990, p. 209-220.
BENCH, R. John.: Communication Skills in Hearing-Impaired Children, Whurr Publishers, London, 1992.
BROWN, W.S. Jr., Goldberg, Donald M. "An Acoustic study of the intelligible offences of hearing-impaired children speakers", Folia Phoniatr, 1990, Vol. 42, p. 230-238.
CARL, W.: "Measurement of aural speech perception and natural speech production of the hearing-impaired," Measurement Procedures in Speech, Hearing and Language, Ed. Singh, S. University Park Press., Baltimore, 1975.
CLARK, Morag.: "Introduction: For which population is an auditory approach suitable?" The Volta Review: Vol. 88, N. 5, 1986.
CONRAD, R.: The Deaf School Child. Harper and Row Ltd. London, 1979.
CONRAD, R.: "Speech quality of deaf children: Disorders of auditory function," II. Proceeding of British Society of Audiology, 2nd. Conference, University of Southampton, 1975, 16-18 July.
DAVIS, H. and SILVERMAN, S.R.: Hearing and Deafness, Holt, Rinehart and Winston, 1978, U.S.A.
DOWNIE, N.M. & WEATH, R.W.: Basic Statistical Method, Harper and Row Pub., New York, 1970.
ENGEN, T. at all.: "Discrimination of intonation by hearing-impaired children", Applied Psycholinguistics, 1983, Vol. 4, p. 149-160.
GRAVETTER, F.J. & WALLNAU, L.B.: Statistics for the Behavioral Sciences, West Publishing Co., San Francisco, 1988.
HUDGINS, C.V. and NUMBERS F.C.: "An investigation of the intelligibility of the numbers, F.C. speech of the deaf," Genetic Psychology Monographs. Vol. 25, 1942, p. 286-392.
KING, A. & PARKER, A.: "The relevance of prosodic features to speech work with hearing-impaired children", Ed. JONES, F.M, Language Disability in Children, MTP Press Limited, 1980.
JOHN, J.E.J., GEMMILL, N., KITZENGER, M. and SYKES, N.: "Some Factors Affecting the Intelligibility of Deaf Children's Speech," Ed: Stephens, S., D., G., Disorders of Auditory Funtion, Academic Press. London, 1976.
JOHN, J.E.J. and HOWARTH, J.N.: "The effect of time distortion on the intelligibility of deaf children's speech," Language and Speech. Vol. 8, 1965, p. 127-134.
MARKIDES, Andreas.: "Type of pure tone audiogram configuration and rated speech intelligibility," J. British Ass. Teachers of the Deaf, Vol. 9, N. 2, 1985, p. 33-36.
MARKIDES, Andreas.: The Speech of Hearing-Impaired Children, Manchester University Press, 1983.
McGARR, Nancy, S.: "The intelligibility of deaf speech to experienced and inexperienced listeners," Journal of Speech and Hearing Research. Vol. 26, 1983, p. 451-458.
METZ, D.E. at al.: "Acoustic dimensions of hearing-impaired speakers' intelligibility," Journal of Speech and Hearing Research. Vol. 328, 1985, p. 345-355.
MONSEN, R.B.: "The oral speech intelligibility of hearing-impaired talkers," Journal of Speech and Hearing Disorders. Vol.: 48, 1983, p. 286-296.
MONSEN, R.B.: "Toward measuring how well hearing impaired children speak," Journal of Speech and Hearing Research. Vol. 21, 1978, p. 197-219.
MOST, T. and SHURGI, M.: "The effect of listeners' experience on the evaluation of intonation contours produced by hearing-impaired children," Ear & Hearing. Vol. 14, 1993, p. 112-117.
OSBERGER, M.J. and MCGARR, N.S.: "Speech production characteristics of the hearing-impaired," Speech and Language, Vol. 8, 1982, p. 222-283.
OSBERGER, M.J.: "Speech intelligibility in the hearing impaired: Research and clinical implications", Ed:
KENT, R.D., Intelligibility in Speech Disorders, John Benjamins Pub. Co., 1992.
SANDERS, D.A.: Aural Rehabilitation, Prentice-Hall, Inc., New Jersey 1971.
TÜFEKÇIOGLU, Umran.: "Comparative Results of the Spoken Language Analyses of Hearing-Impaired Children Attending Two Different Programs in Turkey," Verch, Klaus, (ed.), Hard of Hearing, Deafened, Born Deaf. International Organization of the Rehabilitation Institutions for Hearing Impaired Persons, reg. (IOR), Germany, 1996.
TÜFEKÇIOGLU, Umran.: "Farkli egitim ortamlarinda sözel iletisim egitimi gören isitme engelli ögrencilerin quondam dillerinin karsilastirilmasi." Unpublished Ph.D teases . A.Ü. Eskisehir, 1989.
WOOD, D.J.: "Aspects of the Linguistic Competence of Deaf Children," British Journal of Audiology, 1984, Vol. 18, p. 23-30.
 |
 |