Abstract
Introduction: Dichotic listening (DL) is a test where auditory stimuli are presented simultaneously to each ear. DL has clinical applications in assessing APD, schizophrenia, and brain injury. It is sensitive to hemisphere differences and has been used to study language lateralization. The study aimed to develop, validate, and establish norms for a Quadri condition Dichotic Double Word Test (free recall, forced right, forced left, and switch attention) in the Kannada language (DDWT-K). Methods: This study was carried out in two phases. Phase 1 aimed to develop and validate the DDWT-K and phase 2 involved data collection. The developed novel dichotic test was administered on 100 normal-hearing adults in four conditions – free recall, forced right, forced left, and switch attention. Results: The results showed significant ear differences in free recall condition indicating right ear advantage. The switch attention condition showed slightly poorer performance when compared to both the forced attention condition indicating greater task difficulty. The test-retest reliability was above 0.7 for all the conditions indicating good reliability. Conclusions: The findings demonstrated the best preservation of the original dichotic effects and hence ready for clinical use.
Introduction
Dichotic listening (DL) is one of the auditory processes that involves receiving different auditory stimuli that are presented simultaneously to each ear [1]. Because the dichotic tests are rapid, simple, and non-invasive, they are frequently used in the assessment of hemispheric asymmetry and diagnosis of auditory processing disorders. Recently, with the addition of attention factors, dichotic tests have been used to assess cognitive function in both healthy adults and neurodegenerative disorders [2, 3]. This was achieved by meticulously focusing the attention on either the right or left ear which sets up a cognitive conflict between perceptual saliency (ear advantage) and top-down processing (focus instruction). The dichotic scores are enhanced when attention is focused on the dominant ear because of the congruency between top-down and bottom-up processing. Focusing attention on the non-dominant ear creates a conflicting situation where there is an incongruence between top-down and bottom-up processing resulting in the enhancement of dichotic scores in the non-dominant ear [2]. Several studies conducted using fMRI and dichotic forced attention paradigm concluded that free recall; forced right and forced left conditions can reflect three different cognitive processes, i.e., a lateralized perceptual process (NF), an attention process (FR), and an executive cognitive control process (FL) [4‒6].
Cognitive flexibility is the ability to modify or shift between cognitive sets in response to a change in the environment. Across the lifespan, increased cognitive flexibility has been linked to positive outcomes like improved reading comprehension in childhood [7], increased resilience to stressful events and life events in adulthood [8], and improved quality of life in older people [9]. Cognitive flexibility deficiencies are associated with several neurodevelopmental and neurodegenerative disorders, including Alzheimer’s disease [10] and autism spectrum disorder [11]. Rigorous investigation of this construct has proven tricky, despite the vast consequences of intact cognitive flexibility throughout development and into adulthood. Cognitive flexibility can be assessed using four different methods, namely working memory, response selection, response inhibition, and task switching [12]. The task-switching method has the greatest association which refers to the ability to shift one’s focus or concentration from one ear stimulus to another. Developing a dichotic version of such tests will aid in diagnosing specific language and auditory processing disorders, as well as guiding appropriate interventions and treatments.
DL tests in adults have failed to show ear advantage in adults which could be attributed to several factors such as response bias [13], attention [14], memory [15], and saturation of scores [16]. Moreover, there are very few standardized dichotic tests developed in Indian languages and most of them have been developed as a part of small student projects and, hence, have not been validated, with a lack of test-retest reliability and no adequate normative. Since these dichotic tests are highly related to language processing, there is a definite need for developing language-specific dichotic tests. The addition of cognitive components like attention, executive function, and cognitive flexibility will enhance the clinical utility of dichotic tests, especially in adults with neurodegenerative disorders. The Kannada language is a Dravidian language, native to the state of Karnataka, and 3.73% of the Indian population speak the language which makes Kannada the 8th most spoken language in the country. Hence, the study aimed to develop, validate and establish norms for a Quadri condition Dichotic Double Word Test (free recall, forced right, forced left, and switch attention) in the Kannada language.
Methodology
The study was carried out at the Department of Audiology and Speech-Language Pathology, Kasturba Medical College, Mangalore, from March 2023 to March 2024. The current study used a cross-sectional design and recruited 100 normal-hearing adults to construct normative data for the novel dichotic test. This research project was reviewed and approved by the Ethics Committee of Kasturba Medical College, Mangalore (IEC KMC MLR 03/2023/112). The study was carried out in two phases. Phase 1 included the development and validation of the DDWT-K and phase 2 included data collection and normative estimation.
Phase 1: Development and Validation of DDWT-K
A total of 518 bisyllabic and trisyllabic common standard Kannada words [17] were selected out of which 450 words were chosen concerning the familiarity of the words. The familiarity check was done by 5 adult native Kannada speakers using a 5-point Likert scale, with 1 being the least familiar and 5 being the most familiar. The words from the final word list were then recorded with the help of a native male Kannada speaker (speakers who can read and write Kannada, whose past 3 generations have been residing in Karnataka). The speaker was instructed to keep the microphone at 6 cm from the mouth and was asked to produce a clear utterance with adequate intensity and intonation. The background noise present in the recorded stimuli was eliminated using the adaptive noise reduction method, and the intensity of all the recorded stimuli was group normalized using PRAAT software. The unwanted part from the recorded stimuli was cropped and saved in a .wav. Using the PISOLA (Pitch synchronous overlap and add) algorithm in PRAAT software, the duration of all the stimuli was adjusted between the range of 0.5–0.6 s by expanding or compressing the stimulus to normalize the duration. According to the literature, more than 50% compression is required to create a redundancy that distorts the stimulus [18]. One kHz pure tone (1 s) was created in Adobe Audition (3.0) which was used as an alerting tone for forced conditions. The recorded stimulus was cleaned and stored in WAV format. Dichotic stimuli were prepared using Adobe Audition version 3, where two stimuli were stored on the right and two on the left (Fig. 1, 2). The duration of each stimulus was 4 s (0.5-s stimulus, 1-s gap, 0.5-s stimulus, and 2-s gap). A total of 60 dichotic stimuli were prepared which were divided into four lists consisting of 15 stimuli in each list. Each stimulus had a total of 4 words, two presented simultaneously to each ear. The recorded words were then validated by 3 native Kannada speakers based on certain parameters like intensity, difficulty, duration, distortion, and appropriateness (Table 1). The words were rated from 1 to 5 where 1 being not intelligible and 5 being very intelligible. The words with an intelligibility rating of 3 and above are selected and are equally and randomly divided into each set.
Question rating . |
---|
Speech token is not difficult to perceive; strongly agree − 5 |
Speech tokens are not distorted; agree – 4 |
Stimulus intensity is adequate for perception; undecided – 3 |
Duration of the Speech token is adequate; disagree – 2 |
Naturalness of the stimulus is good; strongly disagree – 1 |
Question rating . |
---|
Speech token is not difficult to perceive; strongly agree − 5 |
Speech tokens are not distorted; agree – 4 |
Stimulus intensity is adequate for perception; undecided – 3 |
Duration of the Speech token is adequate; disagree – 2 |
Naturalness of the stimulus is good; strongly disagree – 1 |
Phase 2: Data Collection
A total of 100 participants were recruited for the study with their ages ranging from 18 to 50 years. These subjects were included in the study only if they had normal pure tone thresholds (pure tone average below 25 dBHL), speech identification score of 80% or more, Right-handed which was determined by Edinburgh handedness inventory, and native Kannada speakers who could read and write Kannada.
Procedure
The participants were given informed consent before the test. The subject’s handedness was determined by Edinburgh’s handedness inventory. The participants were then taken to sound treated testing room and seated comfortably. Pure tone audiometry was then carried out using the modified Hughson and Westlake method (Ascending and Descending Method). Detailed case histories related to hearing, communication, systemic disease, occupation, and family history were taken before PTA. Participants were also tested for middle ear dysfunction using tympanometry on the GSI TympStar middle ear analyzer using a 226 Hz probe tone (A type – normal tympanogram with admittance value of 0.3–1.66 cc and middle ear pressure −100 to +60 dapa). Participants with any history or present middle ear pathology were excluded from the study. After this, the newly developed DDWT-K was administered to each participant where the dichotic stimuli were presented through calibrated headphones. The participants were instructed to listen to the stimuli carefully and to recall the stimuli presented to both ears or to recall stimuli given to specific ears. The test-retest reliability was assessed by re-administering the test on 25 randomly selected subjects after 1 month of initial assessment.
Conditions
- 1.
Free recall: the subjects were asked to concentrate on the stimulus given to both ears and recall the stimulus presented in both ears.
- 2.
Forced right: the subjects were asked to recall whatever was heard from only the right ear.
- 3.
Forced left: the subjects were asked to recall whatever was heard from only the left ear.
- 4.
Switched attention: the subjects were asked to switch their attention between the right and left ear based on the beep tone heard.
Scoring
In the free recall condition, every correct recall of the word presented will be given a score of 1. Correct recall of all four stimuli presented to both ears will be scored as a double correct score of 1. Similarly, if the participant recalled both the stimuli presented in the left ear correctly and one stimulus in the right ear, then a score of 2 will be given to the left ear and a score of 1 will be given to the right ear, and vice versa. In the forced attention condition, correct recall of both stimuli presented to the attended ear will be given a score of 2.
Data Analysis
Dichotic scores obtained from four conditions were documented in Excel files and then exported to IBM SPSS statistics (version 23). All the variables were checked for normality using the Altmann criteria and Shapiro-Wilks test. As the data were not normally distributed, nonparametric tests like Friedman’s test and post hoc Wilcoxon signed-rank test were used to infer a test of significance with a level of significance set at 5%. Normative data were constructed using the 5th and 95th percentile observations. With SPSS version 29 (SPSS Inc., Chicago, IL), intraclass correlation coefficient (ICC) estimations and their 95% confidence intervals were computed using absolute-agreement and two-way random-effects model. Koo and Li’s [19] guidelines were followed in interpreting the values. Based on these recommendations, poor, moderate, good, and excellent reliability are indicated by values less than 0.5, between 0.5 and 0.75, between 0.75 and 0.9, and greater than 0.90, respectively.
Results
A total of 100 participants were recruited for the study, consisting of 56 males and 44 females with their ages ranging from 18 to 50 years and a mean age of 32.5 years. All the data were first tested for normality using Shapiro-Wilk’s test, and the results showed that the data were not normally distributed. Hence, non-parametric statistics were applied to check for the test of significance. A pilot study done before the initiation of the main data collection phase showed no significant score difference between the lists (p = 0.98). Hence, lists 2 and 3 were used to measure dichotic scores in forced right and forced left conditions as using the same list may lead to a learning effect. The results are discussed separately for Free recall and attention conditions.
Free Recall Condition
Friedman’s test results indicated statistically significant differences between all the conditions (χ2 (2) = 193.5, p = 0.0001). Post hoc analysis using the Wilcoxon signed rank test revealed a statistically significant difference between right and left ear scores in free recall Condition (z = −3.58, p = 0.001) where right ear scores were significantly larger when compared to left ear scores indicating a right ear advantage (REA). The descriptive statistics and normative range are provided in Table 2. ICC was 0.768 with a 95% confidence interval of 0.61–0.90 indicating good test-retest reliability.
Conditions . | Scores . | Mean . | SD . | 5th percentile . | Median . | 95th percentile . |
---|---|---|---|---|---|---|
Free recall | Right | 25.36 | 4.300 | 13 | 24 | 29 |
Left | 20.48 | 5.116 | 2 | 8 | 21 | |
Double | 5.04 | 3.910 | 0 | 0 | 6 |
Conditions . | Scores . | Mean . | SD . | 5th percentile . | Median . | 95th percentile . |
---|---|---|---|---|---|---|
Free recall | Right | 25.36 | 4.300 | 13 | 24 | 29 |
Left | 20.48 | 5.116 | 2 | 8 | 21 | |
Double | 5.04 | 3.910 | 0 | 0 | 6 |
Attention Condition
Concerning the forced recall condition, Friedman’s test results indicated statistically significant differences between all the conditions (χ2 (4) = 153.5, p = 0.0001). Post hoc analysis using the Wilcoxon signed rank test revealed the results showed no statistically significant differences between the ear scores (right vs. left) (z = −0.55, p = 0.58). Also, there was a statistically significant difference between forced recall Right and switch attention condition where dichotic scores were greater in forced recall when compared to switch attention condition (SA versus FR – [z = −1.92, p = 0.04], SA vs. FL – [z = −1.35, p = 0.17]). Further analysis revealed a significant difference between free recall and forced attention conditions where forced recall scores were larger when compared to free recall scores (free recall right vs. FR – [z = −3.76, p = 0.001], free recall left vs. FL – [z = −4.26, p = 0.001]). The descriptive statistics and normative range are provided in Table 3. ICC was 0.78 with a 95% confident interval of 0.24–0.91 for forced right, 0.71 with a 95% confident interval of 0.52–0.92 for forced left, and 0.72 with a 95% confident interval of 0.48–0.90 for switch attention indicating good test-retest reliability.
Conditions . | Scores . | Mean . | SD . | 5th percentile . | Median . | 95th percentile . |
---|---|---|---|---|---|---|
Forced right | Right | 28.32 | 3.092 | 23 | 29 | 30 |
Left | 1.21 | 2.110 | ||||
Forced left | Right | 2.34 | 2.312 | |||
Left | 28.16 | 2.853 | 14 | 26 | 29 | |
Switch attention | - | 26.00 | 3.524 | 12 | 25 | 29 |
Conditions . | Scores . | Mean . | SD . | 5th percentile . | Median . | 95th percentile . |
---|---|---|---|---|---|---|
Forced right | Right | 28.32 | 3.092 | 23 | 29 | 30 |
Left | 1.21 | 2.110 | ||||
Forced left | Right | 2.34 | 2.312 | |||
Left | 28.16 | 2.853 | 14 | 26 | 29 | |
Switch attention | - | 26.00 | 3.524 | 12 | 25 | 29 |
Discussion
DL is a well-established neuropsychological tool to assess the hemispheric dominance effect. Because of its straightforward testing process and easily understood instructions, this non-invasive test is frequently used to evaluate individuals with cognitive problems, the elderly, and toddlers. However, the majority of dichotic tests were unable to detect a discernible ear advantage in adults, thus masking its implication in the assessment of patients with neurodegenerative disorders. Moreover, these dichotic tests are highly influenced by language proficiency posts a significant roadblock concerning their assessment and management in linguistically diverse countries like India. As per the Census of India, 2001, India is estimated to have 452 official languages [20] and the non-availability of test materials in native languages creates more challenges for audiologists. Hence, the study aimed to develop and validate a Quadri condition double dichotic word test in Kannada. The findings are discussed under two major subheadings.
Free Recall
The results showed a significant difference in dichotic scores between the right and left ear in healthy controls where greater scores were obtained in the right ear when compared to the left ear indicating a REA. The Kimura structural model provides the most plausible explanation for the ear advantage. According to this concept, linguistic stimuli sent to the left ear first reach the right hemisphere and subsequently, the left hemisphere through the corpus callosum, whereas stimuli supplied to the right ear have quicker access to the dominant left hemisphere through the contralateral route [21]. Several neurophysiological studies have shown larger activation in the left hemisphere when compared to the right for dichotic presentation validating this hypothesis [22]. Though there is some evidence in the literature that suggests that REA is seen only in childhood and vanishes as one reaches adulthood, this may not be universal as REA is heavily dependent on test construction and stimulus design. One such instance is the adult ceiling effect for dichotic digit and CV tests, which results from the saturation of dichotic scores in both ears. Numerous studies in the literature provide evidence that raising the difficulty (CV>Sentences>Spondess>Digits) and adjusting the interstimulus interval [16, 23] may decrease the ceiling effect and boost REA. The current study used double words per ear (double word test) and an ISI of 1.5 s which could be the reasons for obtaining a prominent REA even in adults.
Attention Condition
The forced attention condition was introduced to explore the effect of attention on language laterality where the subjects are instructed to focus their attention either on the right or left ear. These two attention conditions may reflect different cognitive processes and this hypothesis became more evident when there was a discrepancy between dichotic scores of forced right and left conditions in different disordered populations. Patients with Alzheimer’s disease [24], post-traumatic stress disorder [25], and schizophrenia [26] all showed selective deficits in the forced left condition while near-to-normal performance in the forced right condition. Hence, we can conclude that the top-down process has a differential influence on right and left ear attention.
The results showed no significant ear differences in forced right and forced left conditions. These results are consistent with a prior review of attention and cognitive control using the dichotic forced attention paradigm [2]. The results demonstrated that the REA was significantly increased when participants were instructed to focus their attention on the right ear (FR), while there was a left ear advantage when participants were instructed to focus their attention on the left ear (FL).
In the current study, we introduced a switch attention condition in an attempt to assess cognitive flexibility. Though there were no significant differences between the conditions, there was an interesting trend seen. The forced right and forced left scores were equal and the switch attention condition showed slightly poorer scores when compared to both the attention condition. Literature evidence suggests that switching attention is cognitively more challenging than focusing attention on a specific ear [27]. Previous neurophysiological research demonstrates that these attentional processes include both distinct and overlapping brain regions [28]. Hence, reduction in scores for switching attention when compared to forced attention indicates the influence of task difficulty on working memory [29].
Summary and Conclusion
The current study aimed the develop and validate a Quadri condition DDWT-K. The developed novel test was validated on 100 normal-hearing adults. Results showed significant ear differences in free recall Condition indicating REA. Test-retest reliability was above 0.7 for all the conditions indicating good reliability. The findings demonstrated the best preservation of the original dichotic effects under both forced and free recall scenarios, meaning that they may be applied in a clinical setting to assess individuals with neurodegenerative disorders like schizophrenia and Alzheimer’s disease.
Statement of Ethics
All participants provided written informed consent, and the Institutional Review Board approved all research procedures (IECKMCMLR-01/2023/26).
Conflict of Interest Statement
The authors declare no conflicts of interest.
Funding Sources
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Author Contributions
Mayur Bhat: writing – original draft, methodology, investigation, data curation, conceptualization, formal analysis, and project administration. Sweedal Chrystal Dsouza and C.S. Jyotsna: investigation and data curation. K. Rohith and M.A. Yashu: formal analysis and writing – original draft.
Data Availability Statement
The data that support the findings of this study are not publicly available due to their containing information that could compromise the privacy of research participants but are available from the corresponding author upon reasonable request.