Article Modification

8/3/2019 Article Modification

1/19

STUDY OF PERFORMANCE VARIATION OF SOME SEGMENTAL SPEECH

CONTRAST IN TELUGU ON CHILDREN WITH HEARING LOSS STUDYING IN

SPECIAL SCHOOLS AND REGULAR SCHOOLS

ABSTRACT

Introduction: Phoneme is the smallest linguistic unit of a language. Phoneme perception is a form of auditory perception in which the listener and speaker distinguish among the sound

contrasts in a language. Erber (1972) concluded that a group with losses of 70 to 95 dB HL had

difficulty with the auditory perception of place of articulation but not of manner or voicing. In

contrast, a group with losses above 95 dB HL had difficulty with the perception of all consonant

features.

Aim of the study: The present study is an attempt to understand the pattern of phonemic contrast

discrimination ability in native Telugu speaking children with hearing impairment studying inregular schools and special schools for children with hearing impairment as compared to

matched children with normal hearing.

Methodology: The present study examines the significant difference in the perception of selectedvowel and consonant contrasts between hearing impaired children (integrated and segregated

hearing impaired) as compared to normal hearing children. All the three groups were included

for study based on certain preset criteria explained in detail later in chapter on methodology. 16

children as subjects were selected for the each group (hearing impaired group). Minimal

contrasts in Telugu language were selected and audio recorded in the sound treated booth. The

recorded stimuli were presented through the loudspeaker to the participating subjects. The

participants were instructed to match the similar & contrast words. The obtained data werescored and analyzed with appropriate statistics.

Results and Discussion: The result showed that normal children performed better than children

with hearing loss (integrated and segregated hearing impaired) on phonemic contrast. Thevowel contrast perceptions were better than consonant contrast. Among the hearing impaired

children, the integrated hearing impaired children performed better than segregated hearingimpaired. Among the vowel perception mid vowel are perceived better than other contrast with

children with hearing impaired. The perception of consonant contrast were poorer than vowel

contrast, among the consonant contrast voicing were perceived better than other contrast with

the children with hearing impaired group.

Conclusion: The children with hearing impairment studying in regular school performed

better than the children with hearing impairment studying in special school showed that the

early intervention more appropriate intervention, possible the children with hearing impairment

studying in regular school could have obtained higher mean scores. Early intervention and

appropriate intervention will improve the perception of hearing impairment.


2/19

Introduction

Phoneme is the smallest linguistic unit of a language. Phoneme within a particularlanguage helps in differentiating one word from another. Warren (1976) defined phoneme as

merely a linguistic unit, which has no perceptual reality in speech perception. A minimal pair is

defined as two words of distinct meaning which exhibit different segments at one point but

identical segments at all other points (Trask, 1996). Auditory discrimination we mean an ability

to discriminate between the sounds at the articulatory, acoustics and other cognitive levels.

Nicolosi et al (1978) defined discrimination as the process of distinguishing among the speech

sounds or words by differentiating them as same or different.

Phoneme perception is a form of auditory perception in which the listener and speaker

distinguish among the sound contrasts in a language. Minimal pairs (Hyman, 1975) is a

traditional type of evidence used to ascertain the phonemic inventory of a language (Bloomfield,

1933), as well as the specific phonological relationship between two given speech sounds.

Vowels are produced without obstruction in the airflow. Perception of vowels is easier

because they are voiced and relatively high in intensity. Vowels are more accessible to auditory

analysis by virtue of their longer duration. Consonants are produced with the obstruction in the

airflow. They vary by the place of articulation, manner of articulation and voicing. Consonant

being less accessible to auditory analysis due to their brevity and relatively low intensity and

held briefly in auditory memory (Stevens, 2006).

(Kuhl et al. 1992) found that the segmentation is acquired during the first year of life. In

particular, vowel categories are established at around 6 months and consonant categories at

around 12 months (Werker et al, 1981).

Eimas et al (1971) studied the infants from one to four months of age in order to assess

their ability to differentiate stop consonant on the basis of voicing difference. Stimuli were /ba/

and /pa/ the result revealed that infants discriminated stimuli pairs best when they came from

differing adult phoneme categories.


3/19

Erber (1972) concluded that a group with losses of 70 to 95 dB HL had difficulty with the

auditory perception of place of articulation but not of manner or voicing. In contrast, a group

with losses above 95 dB HL had difficulty with the perception of all consonant features.

Boothroyd (1984) studied the amplified speech is accessible to children with varying

degrees of sensorineural hearing loss. The values of hearing loss at which scores fell to 50%

(after correction for chance) were 75 dB HL for consonant place; 85 dB HL for initial consonant

voicing; 90 dB HL for initial consonant continuance; 100 dB HL for vowel place (front-back);

105 dB HL for talker sex; 115 dB HL for syllabic pattern; and in excess of 115 dB HL for vowel

height.Crandell et al (1998) The results from this investigation indicated the children with

sensorineural hearing loss obtained the poorest perception scores. Decreased teacher position

significantly improved speech-perception scores in all acoustical environments.

Review of literature shows that the performances of contrastive system are likely to be

different in hearing impaired children studying in special school and regular school. There is

scattered research on hearing impaired children studying in regular school, and there is no study

compares the performance of children with hearing impairment studying in regular school and

special school. Hence taking this into consideration the present study is an attempt to understand

the pattern of phonemic contrast discrimination ability in native Telugu speaking children with

hearing impairment studying in the regular schools and special schools for children with hearing

impairment as compared to matched children with normal hearing.

The present study aims to study the performance of phonemic contrast discrimination

ability in children with hearing loss studying in regular school and special school. The present

study was undertaken to test the following hypotheses: (1) There will be no significant difference

in the perception of selected vowel contrasts between hearing impaired children studying in

normal school and special school. (2) There will be no significant difference in the perception of

selected consonant contrasts between hearing impaired children studying in normal school and

special school.


4/19

Methodology

The development sequence of phonemic contrast discrimination is likely to be different indifferent language because of variation in phonemic system between different languages have

different phonemic system. Telugu is a language spoken as mother tongue in the state of Andhra

Pradesh in south India and belongs to indo Dravidian group of language. One of the

characteristic features of the Telugu language is that majority of Telugu words end in open

syllables. Also Telugu has numerous long consonant and clusters in the word final position. In

native Telugu language words clusters are not common in the word initial position.

Table. 1 given below explains details of subjects included for the study:

S. No Subjects &

group

Gender Age

range

Mean

age

Selection criteria

1. Normal

hearing

children

(group I)

Male 6-12

years

8.926

667

1. 20 dB HL or better puretone average (PTA).

2. No disorder with speech and hearing ability.

3. No history of hearing loss or middle ear

pathology.

4. Native speaker of Telugu.Female 6- 12

years

9.075

2. Integrated

H.I

(Studying in

regular

school as

Group II)

Male 6-12

yrs

8.925 1. Benefited with the hearing aid and were using

hearing aid consistently.

2. Hearing impaired studying in regular school.

3. No retro cochlear pathology.

4. Aided response should fall in speech banana.

5. Native speaker of telugu

Female 6-12

years

9.041

667

3. Segregated

H.I(studying in

special

school as

group III)

Male 6-12

years

8.958

333

1. Benefited with the hearing aid and were using

hearing aid consistently.2. Hearing impaired studying in special school.

3. No retro cochlear pathology.

4. Aided response should fall in speech banana.

5. Native speaker of telugu

Female 6-12

years

9.097

The number of subjects having normal hearing (24) and those with hearing impairment

studying in special school (group III) were restricted according to number of children with


5/19

hearing impairment studying at regular school (16) in group II. It was first intended to select 24

children with hearing impairment each in group II & group III children with normal hearing. It

could not be achieved as integrated children in regular school did not become available.

Stimuli:

The words in Telugu were considered to have minimum contrast based on vowel height,

vowel place, consonant place, consonant manner and consonant voicing. Most of the test items

were of two syllables length. Selection could not be based on the frequency & familiarity of the

words but was made on meaningfulness of words and the contrast they made with each contrast.

Three pairs of word were taken in each phonemic contrast.

Procedure of test stimuli:

In order to know the inter stimulus time interval needed between presentations of

each contrast. The material was first tested on 20 normal hearing children. The normal hearing

subjects were presented the test stimulus at the most comfortable level of hearing through live

voice presentation technique. The stimuli were presented through speech audiometer to the age

matched hearing children. Stimuli were presented in the A-B-X paradigm. That is it could be

ABA, ABB, where A & B are two words which could occur with the contrast X randomly as

ABA, ABB. The subjects were instructed to hear the three words. The average time taken by the

children to hear & perform the task of matching X with A or B and responded by naming the

number was noted down and stimuli were recorded in the same manner.

It was found that 5 seconds inter stimuli time gap and 2 seconds time gap between

contrasts was sufficient to perform the task by most children with normal hearing. The same

timing that is 5 seconds between stimulus and 2 seconds between contrasts was maintained

throughout the recorded stimulus. The words were selected in random order and recorded. The

order effects were taken care of during recording of the stimulus.


6/19

The following table 2 explaining category words, contrast & minimal pairs:

Category Contrast Minimal Pairs

Consonant voicing Bilabial voiced vs.

Bilabial voicelessa)

b)

c)

Labiodental voiced vs.Labiodental voiceless

a)

b)

c)

Velar voiced vs.

Velar voicelessa)

b)

c)

Place of Articulationfor consonant

Bilabial vs. Alveolar a)

b)

c)

Bilabial vs. Velar a)

b)

c)

Labiodental vs.

Alveopalatala)

b)

c)

Manner of

Articulation

Oral vs. Nasal a)

b)

c)

Stop vs. Fricative a)

b)

c)

Vowel Backness Front vs. Back a)

b)

c) Vowel Height High vs. Low a)

b)

c)

High vs. Mid a)

b)

c)


7/19

The speech materials were spoken by a female native Telugu speaker with general dialect

and recorded. Females are known to have higher fundamental frequencies compared to voice of

males and females voice is considered to be clearer and pleasant as compared to males.

Therefore female voice was preferred for presentation of test stimuli in the study. Live voice

speech presentation technique was used for administering the words to find psychometric

function of selected words A 24 years native Telugu speaker spoke the test words: she first

before administering test to normal hearing subjects practiced presentation of words such that the

VU meter needle of the audiometer ( GSI 61) will come to the same position as for a calibrating

tone of 1000Hz and 60 dB would bring it to approximately 5 seconds inter stimulus time gap was

given between test items: each item was presented please say and the tested word followed. Thesubjects repeated the test words only.

The words were recorded clearly and naturally. The words were recorded in the sound

treated room using praat software. Each contrast was recorded in A-B-X paradigm. The speaker

was given training to control the loudness and the appropriate inflection using speech audiometer

UV meter. The microphone was kept 2 inch away from the speakers mouth to avoid noise burst

during recording. Adobe Audition 3.0 software was used to remove the background noise from

the recorded stimuli. The inter time interval and between the contrasts time interval were kept at

5 seconds and 2 seconds respectively using this software.

Test administration:

The recorded stimuli were presented through Philips loud speaker connected to the laptop

(Lenovo Y3000, 500 series). The loudness was kept at most comfortable level of hearing for

each of the subject of the group. The participants were instructed to listen carefully and expected

to find similar word in that contrast (A-B-X paradigm.). The subject was expected to compare X

with A or B to indentify the similar contrast. The familiarizations on the task were given to each

subject using similar material but not test material. The practice responses were not included for

the analysis. The subjects responses were noted down by the experimenter in the record sheet

and the responses were analyzed. Scoring was done by assigning 1 point to correct response and

0 point to wrong response. Thus the maximum score including all items could only be 33.


8/19

Results & Discussion

Performance of normal hearing and children with hearing impairment on vowel contrast:

In the vowel contrast the children with normal hearing and children with hearing

impaired (integrated hearing impaired children and segregated hearing impaired children) were

instructed to discriminated between the contrast whether the perception of vowel contrast were

same in both the groups. They were evaluated on vowel contrast. The mean and standard

deviation evaluated for all the vowel contrast are presented in the table 3 and figure 1. The group

II & group III (integrated H.I & segregated H.I) showed difference in the performance on the

vowel contrast. Both the group performed poor in same contrast. Among those High vs low

contrast, both the group showed better performance than the other contrast. The mean values are

1.88, 1.605 respectively. Group I and group II showed poorer performance with front vs back

and high vs low contrast than high vs mid contrast but there is statistically significant difference

observed between the groups. Group II (Integrated H.I) performed better than the group III

(segregated H.I).

Figure: 1 Denotes the performance of male & female participants in the three groups on vowel

contrast. (F vs L-Front vs Low, H vs L High vs Low, H vs L- High vs Mid)

The normal hearing children performed good as per mean and standard deviation.

Integrated hearing impaired children showed poor response as compared to normal hearing

children as per the mean and standard deviation value as presented in the table 3. The segregated

hearing impaired children performed poorer than normal hearing children and integrated hearing

impaired children based on the mean and standard deviation scores. The mean and standard

deviation are shown in the table 3. GLM 2- way repeated measures of ANOVA were used to


9/19

assess the perception of vowel contrast. The results indicate that there is a significant difference

between the group I, group II & group III (Normal, Integrated hearing impaired children &

segregated hearing impaired children). (F (2, 50) = 162.389, p = 0.50).

Table 3: the mean and standard deviation of Normal hearing children, children with hearing loss

studying in special school and regular school.

Consonant

Contrast

Normal Integrated H.I Segregated H.I

1.voicing: Mean S.D Mean S.D Mean S.D

Bilabial voiced vs

voiceless 2.88 0.448 1.69 0.479 1.06 0.574

labiodental voiced

vs

voiceless 2.96 0.204 1.37 0.885 1.44 0.814

velar voiced vs

voiceless 2.96 0.204 2.06 0.854 1.44 0.814

Place of

articulation:

Bilabial vs

Alveolar 3 0 1.19 0.834 0.56 0.727

Bilabial vs

Velar 2.88 0.448 1.87 0.885 0.69 0.602

labiodental vs

Alveolar Palatal 2.96 0.204 1.81 0.544 0.94 0.68

Manner of

articulation:

Oral vs Nasal 2.75 0.608 1.13 0.619 1.13 0.619

Stop vs Fricatives 3 0 1.5 0.518 1.31 0.479

Vowel Contrast

Front vs Back 2.79 0.588 1.25 0.775 1.375 0.619

High vs Low 2.875 0.338 1.505 0.632 1.19 0.75

High vs mid 3 0 1.88 0.806 1.605 0.516


10/19

The post hoc test (bonferroni) was also computed separately to find out the significant

between the groups. There is a significant difference observed between normal and integrated

hearing impaired children and the significant level p =


11/19

Consonant contrast in Normal and Hearing impaired group:

1. Voicing:

As can be observed from the table 3 the normal hearing children performed better with

the mean value. Integrated hearing impaired children showed poorer response as compared to

normal hearing children with poorer mean and standard deviation value as shown in the table 3.

The segregated hearing impaired children performed poorer than normal hearing children and

also than the integrated hearing impaired children with poorer mean and standard deviation

scores. Among the hearing impaired children i.e. integrated hearing impaired and segregated

hearing impaired they performed better in velar voiced versus velar voiceless and their mean

values are 2.06 and 1.44 respectively, other contrast such as bilabial voiced versus voiceless and

Labiodental voiced versus voiceless performed poorer then velar contrast.

Figure: 2 show Mean value of Normal, Integrated hearing impaired, segregated hearing

impaired children with gender difference on Consonant contrast (voicing).

(BL V vs VL- Bilabial voiced vs voiceless, VV vs VVL- Velar voiced versus voiceless, LDV vs

VL Labial dental voiced vs voiceless).

The results indicate that there is a significant difference between the group I, group II &

group III (F (2, 50) = 130.869 p =


12/19

groups is significant difference for consonant voicing contrast on voicing difference. (F (1, 50) =

35.304, p =


13/19

hearing impairment (integrated hearing impairment and segregated hearing impairment)

performed better with bilabial versus velar contrast and their mean values are 1.88 and .90

respectively, other contrast such as bilabial versus alveolar and Labiodental versus alveolar

palatal performed poorer than bilabial versus velar contrast the mean and standard deviation are

shown in table 3 and figure 3.

Figure: 3 show mean value of Normal, Integrated hearing impaired, segregated hearing impaired

children with gender difference on Consonant contrast (Place of articulation), (BL vs AL-

Bilabial vs Alveolar, BL vs VP- Bilabial vs velar, LDV vs APP- Labiodental vs Alveolar palatal)

GLM repeated Measure of ANOVA was computed. The results are presented in table 8.

The results indicate that there is a significant difference between the group I, group II & group

III (F (2, 50) = 163.48 p =


14/19

Table: 8 shows the results of 2- way repeated measures over group and gender as two factors.

Source

Type III

Sum ofSquares df

MeanSquare F Sig.

Intercept 505.355 1 505.355 1118.180 .000

Group 147.773 2 73.886 163.486 .000

Gender .147 1 .147 .325 .571

Group *

Gender

1.761 2 .880 1.948 .153

Error 22.597 50 .452

Barzaghi & Madureira (2005) found that similar found as the present findings. They

found that bilabial versus velar showed better performance on the perception of the consonant

place contrast. Bilabial versus alveolar and labiodental versus alveolar palatal showed poorer

performance than the bilabial versus velar contrast. Boothroyd (1984) reported that the voicing

and manner of articulation was perceived better than the place of articulation by children with

hearing impairment studying in special school. Voicing and manner of articulation would be

perceived better with the low frequency hearing. Ling (1966) reported that the use of low

frequency hearing increased in voiced versus voiceless contrast and the manner of articulation.

Place cues, as opposed to voicing or manner cues, have been shown consistently to be more

susceptible to the deleterious effects of noise and hearing loss (Miller and Nicely, 1955;

Rabinowitz et al., 1996 cited in Grant et al 1998.

3. Manner of Articulation:

As can be observed in the table 3 the normal hearing children performed better with themean value. Children with hearing impairment in regular schools showed poorer response as

compared to normal hearing children with poorer mean and standard deviation value. The

children with hearing impairment (segregated hearing impaired) performed poorer than normal

hearing children and also poorer than the integrated hearing impaired children with poorer mean

and standard deviation scores. Among the hearing impaired children i.e. integrated hearing

impaired and segregated hearing impaired they performed better stop versus fricatives and their

mean values are 1.50 and 1.31 respectively, other contrast such as oral versus nasal performed


15/19

poorer than stop versus fricatives contrast. Mean and standard deviation presented in table 3 and

figure 3.

Figure: 7 show mean and S.D values of Normal, Integrated hearing impaired, segregated hearing

impaired children with gender difference on Consonant contrast (Manner of articulation). (O vs

N- Oral vs Nasal, S vs F- Stop vs Fricative)

In order to find whether this observed difference in consonant manner contrast

between the groups of subjects and also seen in difference is statistically significant or not

GLM 2 way repeated Measure of ANOVA was computed. The results indicate that there is a

significant difference between the group I, group II & group III (F (2, 50) = 165.863 p =


16/19

Table: 9 shows the results of 2-way repeated measures of ANOVA over group and gender as

two factors.

Source

Type III Sum

of Squares df

Mean

Square F Sig.

Intercept 350.730 1 350.730 1634.472 .000

Group 71.183 2 35.592 165.863 .000

Gender .220 1 .220 1.025 .316

Group *

Gender

.141 2 .071 .329 .721

Error 10.729 50 .215

GLM repeated measures of ANOVA were computed. The analysis of interaction effect

within the contrast, among group, and gender of the subjects indicated that there is a significant

interaction effect within the contrasts on manner of articulation (F (1, 50) = 6.266, p =


17/19

concentration of energy in the low frequencies (500-1000 Hz), where as stop and fricative

has either a flat spectrum or on in which the higher frequency formants (above 4000 Hz).

Summary & Conclusion

The results of the vowel contrast revealed that the perception of vowel is easy when

compared with consonant because the vowels are voiced and relatively high in intensity and they

are produced without obstruction in the airflow. Vowels are more accessible to auditory analysis

by virtue of their longer duration and may hold in the auditory memory. Among the vowel

contrast the perception of mid-vowels are better than the other vowel because the mid vowel are

produced with higher intensity when we compare with other vowel contrast such as front versus

back vowel and high versus low vowel. This could be the possible reason that the children with

hearing impairment could perceive the high versus mid contrast then the other contrast in the

vowel perception. The children with hearing impairment studying in regular school showed

better performance than the children hearing impairment studying in special school.

The results of the present study revealed that there is a significant difference between the

normal hearing children and children with hearing impairment studying in regular school and

special school in the perception of consonant contrast. Further the results suggest that the

children with hearing impairment studying in normal school performed better than the children

with hearing impairment studying in special school and poorer than the children with normal

hearing children. The children with hearing impairment studying in special school performed

poorer than the children with normal hearing and children with hearing impairment studying in

regular school. Therefore the 2nd

hypothesis is rejected.

The following conclusions can be drawn from the results of the present study:

The perception of vowel contrast was found to be better than the consonant contrast. Among

the vowel contrast the high versus mid vowel were perceived better than the perception of front

versus back vowel and high versus low vowel. The children with hearing impairment studying in

regular school performed better than the children with hearing impairment in special school.

Among the consonant contrast the voicing contrast perceived better than the other contrast such

as place of articulation, manner of articulation.


18/19

References

Arakawa, T & Erber, N.P .(1976). Comparison of specialhearing aid with a conventional

hearing aid for detection and discrimination of high frequency phonemes. Independent

Study, spring.

Barzaghi, L,. Madureira, S. (2005). Production and perception of stop consonants by a hearing-

impaired Brazilian subject, unpublished article.

Bloomfield, L. (1933).Language.New York: H. Holt and Company.

Boothroyd, A. (1984). Auditory perception of speech contrasts by Subjects with sensorineural

hearing loss,Journal of Speech and Hearing Research, Volume 27, 134-144.

Crandell, C.C., Siebein, G.W., Martin, A., Gold, M.A., Hassell, M.J., Lee, H.W., Abbott, P.,

Herr, C.R., & Lehde, M. (1998). Speech perception of normal-hearing and hearing-impaired

children in classrooms. The Journal of the Acoustical Society of America, 103, 5, 3063-3070.

Eimas, P. D., Siqueland, E. R., Jusczyk, P., & Vigorito, J. (1971). Speech perception in infants.

Science, 171 (968), 3036.

Erber, N.P., (1972), Auditory, visual, and Auditory-visual Recognition of consonants by children

with normal and impaired hearing.Journal of Speech and Hearing Research, 15, 413-422.

Grant, K.W., Walden, B.E., & Seitz, P.F. (1998). Auditory-visual speech recognition by hearing-

impaired subjects: Consonant recognition, sentence recognition, and auditory-visual

integration,Journal of Acoustic Society of America, 103 (5), 2677-2690.

Hack, Z. C., & Erber, N. P. (1982). Auditory, visual, and auditory- visual perception of vowels

by hearing-impaired children.Journal of Speech and Hearing Research, 25, 100-107.


19/19

Halle, M. (1957). Acoustic properties of stop consonants,Journal of acoustic society of America,

29, 245-256.

Hyman, L.M. (1975).Phonology: Theory and analysis. New York: Holt, Rinehart & Winston.

Ling. D., (1966). The use of low frequency residual hearing in profound deaf children, masters

thesis, Mcgill University.

Kuhl, P. K., Williams, K. A., Lacerda, F., Stevens, K. N., & Lindblad, B. (1992). Linguistic

experience alters phonetic perception in infants by 6 months of age. Science, 255, 606-608.

Miller, G. A., & Nicely, P. E. (1955). Analysis of perceptual confusions among some English

consonants.Journal of acoustic Society of America, 27, 338-352 (1955).

Nicolosi, L., Harryman, & Kreschech, J.S. (1978). Terminology of communication disorders,

Baltimore, MD Williams and Wilkins.

Stevens, K.N. (2006). Features in Speech Perception and Lexical Access in the Handbook of

Speech Perception (2nd

Eds), Malden, Pisoni & Remez, Blackwell Publishing Ltd.

Trask RL. (1996).A dictionary of phonetics and phonology: London: Routledge.

Werker, J.F., Gilbert, J.H.V., Humphrey, K., & Tess, R.C. (1981). Development aspects of cross

language speech perception, child development, 52, 349-355.

Warren, R.H. (1976).Auditory illusions and perceptual processes. In lass, N.J. (Ed) contempory

issue in experimental phonetics, 389-417.

Documents

Article Modification