Assessing Binaural Auditory Processing in Native Tamil Speaking Children

Abstract

The objective of this study is to develop and validate the Binaural Fusion Test (BFT) in the Tamil language for children with learning disability. As many children with learning disability exhibit an auditory processing deficit there is a need for language specific assessment tool for these children. The BFT in Tamil comprises 100 words organized into four lists, with each list containing 25 words. A total of 84 children with normal auditory processing and 45 children diagnosed with learning difficulties were administrated the BFT in Tamil. Test-retest reliability was evaluated in a sample of 30 children who exhibited normal auditory processing. The results of BFT in children with normal auditory processing demonstrated satisfactory reliability by the Cronbach Alpha test. A statistically significant difference was seen in the mean scores of BFT in Tamil when comparing the groups. The present study offers novel normative data for the binaural fusion test in Tamil, which has not been previously documented in the existing literature. For children, the BFT in Tamil, which is a part of the CAPD test battery, can be used to find binaural interaction deficiencies.

Introduction

Children with learning difficulties may exhibit aberrant processing in several components of the auditory system, including the brainstem and auditory cortex [1]. Many children with learning disabilities exhibit symptoms of exhibit symptoms of auditory processing dysfunction, such as difficulties in understanding instructions, poor listening skills, and struggles in distinguishing between similar sounds. The deficit in the various processing of auditory skills such as auditory figure ground, binaural integration, and temporal processing can be assessed through the psychoacoustic test. The healthcare practitioners can identify the processing deficit and by performing language-specific auditory processing assessments.

Children with learning disability exhibit abnormalities in the binaural interaction processing. An important factor to consider when assessing children with learning disability is binaural interaction [2]. Binaural fusion or interaction is a cognitive process that involves merging auditory data delivered in a binaural manner. The binaural fusion procedure is crucial for human speech comprehension, as the reception of speech stimuli can vary between ears based on the direction of the signal source [3]. Binaural fusion is a highly sensitive technique utilized for the identification of irregularities in auditory processing [4]. In a binaural fusion test, the voice stimulus is presented to each ear independently, with low-frequency and high-frequency parts. In order to comprehend the entirety of a word, the individual must integrate both low-frequency and high-frequency information.

Matzke pioneered the use bi-syllabic, phonetically balanced syllables to develop the binaural fusion test. He states that the brainstem nuclei are responsible for the integration of two impulses, one originating from each ear [5]. When the cochlear nuclei and medial geniculate bodies effectively combine the filtered signals, subjects demonstrate higher scores in terms of intelligibility [5, 6]. The outcomes of auditory processing assessments for children with learning disability exhibit variability contingent upon the specific language utilized during the assessment and their level of proficiency [7]. Bilingual persons who have learned English as a second language demonstrate significantly reduced ability to recognize speech in English ^[8]. Linguistic variations exert a substantial influence on the outcomes of auditory processing tests in children with learning disabilities when a speech signal is employed as the stimulus. To reduce the impact of language on the evaluation process, it is crucial to conduct auditory processing tests that include voice stimuli in all native languages.

India ranks second in terms of the number of languages spoken, with a total of 780, as reported by the People’s Linguistic Survey of India. Tamil, a classical language from the Dravidian language family, is predominantly spoken in Tamil Nadu and neighbouring regions of India [8]. Kumar et al., state that the Tamil language comprises five vowel pairs, eighteen consonants, and an additional six consonants [9]. The Dichotic Word Test [10] and the Speech in Noise Test in Tamil for adults [11] are two assessments that can be employed to examine dichotic processing and auditory closure skills in Tamil. Indian languages like Malayalam [12] and Hindi [13] provide binaural fusion testing. Given that Tamil is the predominant language spoken by the majority of the population, it is imperative to conduct auditory processing tests that are specific to this language in Tamil Nadu. The objective of this study was to fill the gap in binaural fusion evaluations in the Tamil language by creating a specialized binaural fusion exam for children who have learning impairments. The aim of this study is to investigate the difference in scores on the binaural fusion test in Tamil between children with normal auditory processing ability and those who have been assessed for learning difficulties.

Methodology

The research conducted was a cross-sectional study that utilized the purposive sampling technique to select participants. The binaural fusion test in Tamil was developed and validated through a six-phase process.

Phase I: Test stimulus development: Kumar et al. conducted a study that provided the most frequently occurring words in Tamil text books for children in grades 1 to 5. The list of 250 commonly used words from this study was taken for the familiarity test [9]. On a four-point scale, five adults who are native Tamil speakers were asked to rate the familiarity of Tamil words. The results indicate that out of a total of 250 words, 91 were rated as the most familiar, 52 as familiar, 98 as unfamiliar, and 8 as the least familiar. Four sets of 25 phonetically balanced words were compiled using the 143 most commonly recognized words. The Tamil script comprises 18 consonants, 6 additional consonants, 5 vowel pairs, and 2 diphthongs. Each of the lists contained a minimum of one representation of all the consonants, vowels, and diphthongs found in the Tamil language. The four-word list exhibited a repetition of 11 bi-syllable structures. For the purpose of recording the stimulus, a male adult who was a native Tamil speaker and whose speech was deemed intelligible by all participants according to the Speech Intelligibility Rating Scale was selected [14]. Following a period of sufficient practice, the speaker recited the words from the provided lists at a standard pace of speech within a sound-treated environment characterized by an ambient noise level below 30 dBA. The audio recording was conducted using the Praat voice analysis software, employing a condenser microphone positioned at a height of 10 cm relative to the speaker’s oral cavity. The recorded word lists underwent RMS loudness scaling at a level of 70 dBSPL using the Prat voice analysis software in order to achieve uniformity in the intensity of all words. The four wordlists, which were equalized in terms of loudness, were presented to a group of five native Tamil speakers using headphones. The participants were asked to rate the naturalness and clarity of the words on a 4-point Likert scale. A rating of 1 indicated that the words were not natural or clear, while a rating of 4 indicated that the words were very natural and clear. No words were assigned a rating of 1. Consequently, all the words that were captured were taken into account for the test stimulus.

Using Gold Wave digital audio editor software, we consistently maintained a five-second interval between each word in the four lists. We utilized a low-bandpass filter ranging from 500 to 700 Hz, with a rejection rate of 52 dB per octave. Additionally, we employed a high-bandpass filter ranging from 1800 to 2000 Hz with a rejection rate of 60 dB per octave. As a consequence, we produced eight .wav files. In two situations, we combined the high-pass and low-pass words from each list in stereo mode. Conditions 1 and 2 were characterized by the presence of a low pass in the right ear and a high pass in the left ear, as well as a high pass in the right ear and a low pass in the left ear, respectively. To validate the VU meter setting in the audiometer, we applied a 1 kHz calibration tone before the first test word in each of the mixed word lists under both conditions.

In Phase II, we conducted pilot testing by administering the binaural fusion test that we had designed in Tamil to a group of 10 young adults with typical hearing sensitivity. The purpose of this test was to confirm that the scores were equal across all four lists. The BFT was presented at 50 dBSL through a calibrated clinical audiometer (Otometrics Madsen Astera 2) using external A and external B routing channels. The participants were instructed to reiterate the heard words. A score of one was assigned to the accurate response, while a score of zero was assigned to the inaccurate response.

In Phase III, A cohort of 124 school going children from grade I to VI, between 7 to 12 years who were who were native Tamil speakers were selected. Informed consent was gained from the parents, whereas assent papers were received from the children. The children were assessed in order to ascertain the normalcy of their hearing sensitivity and middle-ear functioning. A total of 98 children satisfied the criteria for normal hearing and a tympanogram. The ‘Screening Checklist for Auditory Processing (SCAP)’ was introduced to the teachers which has a set of 12 binary-choice questions and the teachers were asked to rate their respective class student [15]. 84 children who exhibited usual development and successfully passes the SCAP screening were taken as the control group I. 45 children were chosen for group II who were studying at special schools catering to children with learning impairments. At their special school, psychologists had diagnosed all of these kids as children learning disability.

In Phase IV, the BFT in Tamil was conducted in a sample of 84 typically developing children who met the requirements of normal hearing sensitivity and had no risk of auditory processing deficiency. The purpose of this administration was to get age-related normative scores. The Tamil binaural fusion test was conducted at a sound level of 50 dBSL and connected to a calibrated clinical audiometer (Otometrics Madsen Astera 2) using external A and external B routing channels. The low-pass stimulus was directed towards the right ear, while the high-pass stimulus was directed towards the left ear. The children were instructed to reiterate the given words, and their oral reactions were documented using voice recording software on their mobile devices. A score of one was assigned to each accurate response, while a score of zero was assigned to the incorrect response.

Phase V: In order to assess the test-retest reliability of the binaural fusion test, a repeated procedure was conducted on a sample of 30 children selected at random from a total of 84 participants. The subsequent evaluation was conducted after a period of 7 to 15 days. The 30 children repeated the BFT in Tamil, within a consistent clinical setting, and their verbal responses were subsequently recorded and subjected to analysis.

Phase VI: The binaural fusion test was administered to a group of 45 children diagnosed with learning disabilities. These children were selected based on their normal hearing sensitivity and the absence of any comorbid conditions. The test was conducted using the same clinical setup as the children with normal auditory processing. The participants’ oral responses were documented and subjected to analysis.

Statistical Analysis

The statistical program for the social sciences (SPSS) version 24 was utilized to conduct the statistical analysis of the data. The collected data underwent analysis using descriptive statistics and normality tests, specifically the Kolmogorov-Smirnov and Shapiro-Wilk tests. The impact of the four lists was analysed using the Wilcoxon signed rank test, while the test-retest reliability was assessed using Cronbach’s alpha test. In order to ascertain the presence of a significant difference between the groups, the student t test was employed.

Results

The objective of this study is to create and authenticate the binaural fusion test in the Tamil language. The initial assessment conducted on a sample of 10 young adults yielded an average score of 25 across all four lists, suggesting that the four lists were comparable in terms of their suitability for testing purposes.

Binaural Fusion Test in Children with Normal Auditory Processing

The mean and standard deviation (SD) scores obtained for the BFT test were compared among a sample of 84 typically developing children with normal hearing sensitivity. The impact of age on usually developing children was examined in a study involving children aged between 7 and 12 years. Based on their chronological age, the participants were categorized into five subgroups: 7 to 7 years 11 months, 8 to 8 years 11 months, 9 to 9 years 11 months, 10 to 10 years 11 months, and 11 to 11 years 11 months. Table 1 presents data pertaining to the mean and standard deviation scores associated with age in children with normal auditory processing.

Table 1.

Mean scores for the 4 lists of BFT in children with normal auditory processing

	n	List 1 Mean (SD)	List 2 Mean (SD)	List 3 Mean (SD)	List 4 Mean (SD)
7 years - 7 years 11 months	18	20.11 (3.10)	20.66 (2.32)	21.22 (3.09)	21.77 (2.46)
8 years - 8 years 11 months	15	21.08 (1.97)	21.91 (1.67)	22.33 (1.43)	22 (1.70)
9 years - 9 years 11 months	15	21.4 (2.59)	22.4 (1.99)	22.93 (1.79)	22.93 (1.91)
10 years - 10 years 11 months	16	22.583 (1.51)	21.83 (2.08)	23.33 (0.78)	22.83 (1.59)
11 years - 11 years 11 months	20	22.96 (1.59)	22.61(1.50)	23.73 (1.04)	23.23 (1.27)

The Fig. 1 shows the comparison of the scores in children with normal auditory processing across the binaural fusion test four word lists. The test for normality Kolmogorov Smirnov and Shapiro Wilk revealed that the obtained data was not normally distributed. In order to compare the mean scores across the lists, a non-parametric Wilcoxon-signed rank test was conducted. List 1 had a significance level of z = -1.601, p = 0.109; List 2 significance level of z = -1.740, p = 0.82; List 3 significance level of z = -1.620, p = 0.1; List 4 significance level z = -1.60, p = 0.10. 

Fig. 1.

Children with normal auditory processing comparison of BFT scores across lists

Test-Retest Reliability for Children with Normal Auditory Processing

The test-retest reliability of 30 children was assessed using the Cronbach’s alpha coefficient, for the four lists were 0.89, 0.74, 0.88, and 0.86 respectively. The Cronbach alpha correlation test α > 0.7 indicates that test-retest reliability for all four lists is reliable.

Binaural Fusion Test in Children with Learning Disability

The mean and standard deviation (SD) scores obtained for the BFT test were compared among a sample of 45 children with learning disability. The impact of age on children with learning disability was examined in a study involving children aged between 7 and 12 years. Table 2 presents data pertaining to the mean and standard deviation scores associated with age in children with learning disability.

Table 2.

Mean scores for the 4 lists of BFT in children with learning disability

	n	List 1 Mean (SD)	List 2 Mean (SD)	List 3 Mean (SD)	List 4 Mean (SD)
7 years - 7 years 11 months	8	16.11 (2.10)	16.66 (2.32)	16.22 (2.49)	16.77 (2.46)
8 years - 8 years 11 months	8	17.28 (2.17)	17.91 (2.37)	17.13 (2.53)	17.60 (2.27)
9 years - 9 years 11 months	9	17.04 (2.29)	17.41 (2.41)	17.93 (2.59)	17.65 (2.71)
10 years - 10 years 11 months	10	18.58 (2.05)	17.73 (2.08)	17.85 (2.65)	18.64 (2.48)
11 years - 11 years 11 months	10	18.77 (2.43)	18.75(2.60)	18.63 (2.16)	18.33 (2.38)

The Fig. 2 shows the comparison of the scores in children with learning disability across the four lists of binaural fusion test in Tamil.

Fig. 2.

Children with learning disability comparison of BFT scores across lists

Comparison between the Groups

The student t test was utilized to examine the statistically significant disparity in the average scores among the groups. The comparison between the groups across the lists is presented in Table 3, with t-values for each group. List 1 has a t-value of 17.924 (p < 0.001), List 2 has a t-value of 16.363 (p < 0.001), List 3 has a t-value of 17.468 (p < 0.00), and List 4 has a t-value of 18.410 (p < 0.000).

Table 3.

Comparison between the groups across the lists

	t(45)	p-value
List 1	17.924	0.000*
List 2	16.363	0.000*
List 3	17.468	0.000*
List 4	18.410	0.000*

*p<0.00 significance

Discussion

The central auditory processing assessment becomes challenging due to the linguistic foundations of these tests. Hence, it is imperative to devise language-specific assessments for central auditory processing in order to ascertain the fundamental processing impairment. Binaural interaction is an example of a core auditory processing skill, and the significance of binaural interaction in auditory processing has been documented [16]. Children with learning disability issues exhibit subpar performance in binaural interaction, as shown by the data. The absence of a conventional binaural fusion test in Tamil necessitated the development of a binaural fusion test in the language. However, it is important to note that this particular test has not yet been documented in the existing literature.

This study presents normative data pertaining to the test administered to children aged 7 to 12 years. To develop a speech perception test, it is essential to ensure that the material is phonetically balanced. In order to achieve a comprehensive phonetically balanced list, it is important to take into account the frequency of sound occurrences within the specific language under consideration. According to the age and subject matter that people articulate, the frequency of occurrence varies [17]. In the present work, the frequency of each word in the textbooks of children in grades I to V to generate a phonetically balanced wordlist for the development of a binaural fusion test. When developing auditory processing test material, it is important to take into account the frequency of occurrence. This study concentrated on the frequency of occurrence and specifically chose the words that were listed as the most frequently occurring words in the Tamil text books in the study by Kumar et al. [9]. All 18 consonants, 6 additional consonants, 5 vowels, and 2 diphthongs found in Tamil were included in each of the compiled lists. Therefore, the four lists were constructed in order to achieve phonetic balance in the word lists. The familiarity of the words was also assessed in order to generate the word list.

Auditory processing tests should include the capability to accurately identify children who exhibit typical processing abilities. In this particular scenario, the mean score obtained from all four lists of the binaural fusion test in Tamil ranged from 21 to 22 out of 25. This implies that youngsters who possess typical auditory processing abilities are capable of achieving high results on this assessment. Furthermore, the four lists failed to meet the maximum threshold for complete, accurate identification. This suggests that the test is presenting a state of competition that is necessary for evaluating core auditory processing. This exam, when administered to children with processing problems, will be beneficial in distinguishing them from typically auditory-processing youngsters due to its reduced redundancy. It is imperative to ensure the test-retest reliability of any recently created auditory processing assessment. The test-retest reliability analysis was conducted on the binaural fusion test in Tamil, revealing a significant connection between the two measures.

The mean scores of the BFT for children identified with learning impairments varied between 16 and 17 across all age categories. The scores achieved by children with learning disabilities were comparatively lower than the mean scores attained by children exhibiting typical auditory processing abilities. The observed disparity across the groups indicates that the created BFT possesses the capability to distinguish between children exhibiting processing impairments and those displaying typical auditory processing abilities. The findings of this study align with previous research indicating that the central auditory processing tasks had the capability to distinguish between children who are usually developing and those who have learning difficulties [18–20].

Currently, there is no existing test specifically designed for binaural fusion in the Tamil language. The present investigation aims to evaluate the processing of binaural interactions in children with learning disabilities who are native Tamil speakers. In the context of the CAPD test battery, the BFT in Tamil can serve as a valuable tool for distinguishing between children who exhibit auditory processing deficits. The language roots of these assessments make the central auditory processing assessment difficult [21]. Therefore, it is crucial to develop examinations that are specific to each language in order to determine the underlying issue in central auditory processing [10, 22]. Binaural interaction is a fundamental auditory processing skill, and its importance in auditory processing has been well documented [5, 20, 23]. The results show that children with learning disabilities have below-average proficiency in binaural interaction. Due to the lack of a traditional binaural fusion test in Tamil, it was necessary to create a binaural fusion test specifically for the language. Nevertheless, it is crucial to acknowledge that this specific test has not been recorded in the current body of research.

Conclusion

This research offers distinctive normative data for the binaural fusion test conducted in the Tamil language. The Tamil version of the binaural fusion test (BFT) demonstrates strong consistency when administered multiple times to children. Additionally, it effectively differentiates between children with auditory processing impairment and those with normal auditory processing. Therefore, the binaural fusion test now being developed in Tamil can be used in the test battery to detect central auditory processing dysfunction in Tamil-speaking children.

Limitations

The data analysis did not take into account the subgroups of learning disability. Gender-based performance disparities were not taken into account during the analysis.

Author Contribution

The conceptualization process involved equal contributions from all authors. The contribution to the process of data collection was made by Ms. Praveena J. The writing contribution was done by Ms. Praveena J, Dr. Selvarajan G and Dr. Savitha V. H. All of the study’s authors participated in the editing and reviewing process. The final manuscript has received approval from all authors.

Declarations

Ethical Approval

The scientific committee and the ethical committee connected to the institute both conceptualized and approved the project. According to the authors, reference number 2205/IEC/2020 demonstrates that all procedures used in this study abide by the ethical standards established by the relevant institution.

Competing Interests

The authors did not have any conflicts of interest. The study doesn’t have any financial support to disclose.