An optimal dichotic-listening paradigm for the assessment of hemispheric dominance for speech processing.

Dichotic-listening paradigms are widely accepted as non-invasive tests of hemispheric dominance for language processing and represent a standard diagnostic tool for the assessment of developmental auditory and language disorders. Despite its popularity in research and clinical settings, dichotic paradigms show comparatively low reliability, significantly threatening the validity of conclusions drawn from the results. Thus, the aim of the present work was to design and evaluate a novel, highly reliable dichotic-listening paradigm for the assessment of hemispheric differences. Based on an extensive literature review, the paradigm was optimized to account for the main experimental variables which are known to systematically bias task performance or affect random error variance. The main design principle was to minimize the relevance of higher cognitive functions on task performance in order to obtain stimulus-driven laterality estimates. To this end, the key design features of the paradigm were the use of stop-consonant vowel (CV) syllables as stimulus material, a single stimulus pair per trial presentation mode, and a free recall (single) response instruction. Evaluating a verbal and manual response-format version of the paradigm in a sample of N = 50 healthy participants, we yielded test-retest intra-class correlations of rICC = .91 and .93 for the two response format versions. These excellent reliability estimates suggest that the optimal paradigm may offer an effective and efficient alternative to currently used paradigms both in research and diagnostic.

Introduction

Verbal dichotic-listening paradigms offer well-established behavioral tests for the assessment of latent hemispheric differences for language processing [1, 2] and are integral part of test procedures for the diagnosis of auditory processing disorders [3-5]. A significant advantage of dichotic compared with alternative paradigms (e.g., visual-half field techniques or neuroimaging approaches) is the simplicity of the testing procedure which can be easily understood and performed also by young children [e.g., 6, 7], elderly individuals [e.g., 8, 9], or patients with cognitive disabilities [e.g., 10, 11]. That is, in its basic form, pairs of verbal stimuli (e.g., words or syllables) are presented via headphones, with one of the stimuli presented to the left ear and the other one, simultaneously, to the right ear [12]. Instructed to report the one of the two stimuli which was heard best, participants typically report the right-ear stimulus more frequently, more accurately, and more rapidly than the left-ear stimulus. This right-ear advantage is widely accepted to indicate left hemispheric dominance for speech processing [2, 13, 14] and differences in the magnitude of this right-ear preferences have been related to interhemispheric auditory integration [15-17].

However, many different versions of dichotic-listening paradigms have been suggested [12, 18] and the test-retest reliability of most dichotic paradigms–even of those used for diagnostic purposes–are far from optimal [19, 20]. This shortcoming severely threatens the inferences that can be made using dichotic-listening measures, as the reliability of a test also sets the upper limit of its validity [21]. At the same time, however, substantial differences in reliability estimates between paradigms [19] suggest that certain design features systematically affect the consistency of an individual’s dichotic listening task performance. In fact, since the early conceptualization of dichotic listening more than six decades ago [22, 23], a plethora of studies has accumulated a significant amount of evidence on how features of the experimental set-up (e.g., stimulus order, stimulus material) and task instructions lead to systematic response biases [for review see ref. 18]. For example, it has been demonstrated that paradigms instructing participants to selectively attend to only one ear at a time are more difficult for participants to perform than paradigms allowing for a free selection [24]. Also, presenting multiple stimulus pairs per trial increases the working-memory load of the task by requiring the participant to keep the representation of multiple stimuli activated simultaneously, leading to a decreased right-ear advantage [25].

Ignoring these design variables or leaving them uncontrolled introduces error variance to the obtained measures, affecting both the reliability of the obtained laterality measures and the efficiency of the paradigm. In turn, however, considering these variables when designing a dichotic-listening task makes it possible to tailor a paradigm which is optimized for a given purpose. In the present paper, the intention is to design a dichotic listening paradigm optimized for the assessment of hemispheric dominance for speech and language processing. As theoretical framework, we assume that dichotic-listening performance can be best explained using a two-stage model [26-28]: an initial stage leading to a perceptual representation of the two competing stimuli in verbal short‐term memory [29, 30], and a second stage characterized by cognitive-control processes which may modulate the initial representation in line with task requirements, resulting in a response selection [31, 32]. It is further assumed that a “true” underlying right-ear preference exists as consequence of left-hemispheric specialization for speech and language processing [33, 34]. That is, in the initial stage, the right-ear stimulus can be expected to be more salient than the left-ear stimulus. An optimal paradigm to assess this underlying perceptual, “built-in” laterality, consequently needs to assure both an unbiased initial stimulus representation and a cognitively unaltered response selection during second stage processing [27]. Based on recent literature review [18], we here aim to create and evaluate such an optimal paradigm by following the eight design features listed in Table 1.

Table 1

Design features of the optimal dichotic-listening paradigm and arguments for their implementation.

#	Design feature	Argumentation
1	Stop consonant-vowel (CV) syllables as stimulus material	Proven to be valid test material [38]. CV show higher reliability than numeric or non-numeric words as stimulus material [19]
2	Pair only CV stimuli from the same voicing category	Same-voicing category pairs are more likely to fuse into one percept than mixed pairs; reduces relevance of attention and cognitive-control processes [28]
3	Includes binaural (diotic) trials	Allows to demonstrate stimulus appropriateness, i.e. whether participant is able to identify the used stimulus material [18]
4	Alternating trials of voiced and unvoiced stimulus pairs	Limits negative-priming effects [39] by preventing stimulus repetition between consecutive trials
5	Paradigm length of 120 dichotic trials	Previous studies indicate reliability estimates >.80 mostly for paradigms using 120 trials [18]
6	Single-stimulus pair per trial; single, immediate response	Minimizes working-memory load compared to multi-stimulus trials [25] and delayed response paradigms [36]
7	Free-recall instruction	Reduces task difficulty and relevance of cognitive-control processes compared to selective attention instructions [24]
8	All stimulus pairs are presented in both orientations (i.e., left-right ear and right-left orientation) in identical frequency	Averages otherwise uncontrolled biases across these trials (e.g., item difficulty effects; see [40, 41])

(a) for a detailed discussion refer to [18]

However, while many design features of dichotic paradigms have been studied systematically, it appears somewhat surprising that little is known about the effect of the response format on task performance [18]. When using a verbal-response format, the participant is typically instructed to repeat orally the stimulus she/he perceives after each trial, and the response is either ad-hoc scored by an experimenter or recorded for later decoding [25, 35]. Manual responses are typically collected via button press on a keyboard, special response box, or computer mouse [28, 36]. Both alternatives come with theoretical advantages and disadvantages [18]. For example, the verbal-response format carries the risk of increasing error variance in the recorded data as the response has to be decoded by the experimenter. As compared to a verbal version, using a manual response format changes the cognitive demands of the paradigm as it demands additional response mapping and selection processes including visual-motor coordination [37]. However, an empirical comparison of the two response formats is missing from the literature and it is to date unknown if the reliability of a dichotic paradigm is affected by the response format utilized.

In summary, the aim of the present study was to develop and test a new dichotic-listening paradigm for the assessment of hemispheric dominance which considers the main design feature known to affect task performance and yields high test-retest reliability. To evaluate the retest reliability of this novel paradigm, participants had to complete the full paradigm four times: each two times using verbal and manual response format. This setup also allows evaluating possible effects of the response format on laterality estimates.

Material and methods

Participants

The final sample consisted of N = 50 participants (n = 30 female, 60%; age: mean ± s.d. was 25.0 ±4.2 years, range 19 to 39 years) recruited among students and employees of the Universities of Oslo (n = 17) and Bergen (n = 33), Norway. All participants were right-handed (as verified in Edinburgh Handedness Inventory, [42]) and had no history of neurological or psychiatric conditions (as verified in self-report). The participants were fluent speakers of Norwegian (42 native speakers, 8 non-native speakers) and high identification rate (mean: 98.3 ± 2.2%) of the used stimuli when presented diotically (see next section) assured that the used stimulus material was appropriate for all included participants. Hearing acuity was assessed in an audiometric screening (Oscilla USB-300, Inmedico, Lystrup, Denmark) for pure tones of 250, 500, 1000, 2000, and 3000 Hz. All included participants had an absolute interaural threshold difference smaller than 10 dB (range from -8 to 8 dB in favor of the right and left ear, respectively; mean = -0.6 ± 4.1) and absolute threshold of smaller than 25 dB (mean left ear: 4.7 ± 7.5 dB; right ear: 5.2 ± 7.2 dB) across the tested frequencies. For matter of completeness, the final sample excluded two participants: one who identified only 59 of the maximum 72 diotic stimuli correctly (81.9%), and one participant with an interaural acuity difference of 16 dB.

The study was approved by the ethical review board of the Department of Psychology, University of Oslo, and written informed consent was obtained from all participants.

Stimulus material and paradigm design

Six stop-consonant-vowel (CV) syllables served as stimulus material; three of voiced (/ba/, /da/, /ga/) and three of unvoiced articulation (/pa/, /ta/, /ka/). The syllables were natural recordings of a male Norwegian voice actor spoken in constant intonation and intensity. CV-syllables were preferred over alternatives (e.g., rhyming words) as they tend to show higher reliability than other stimulus material [19] and the validity of CV-based test has been repeatedly confirmed (e.g, using the Wada test [38], functional MRI [43], and in studies on surgical patients [44]; see point 1, Table 1). Furthermore, to increase the likelihood of perceptual fusion, syllables of the same voicing category were paired exclusively ([28], point 2, Table 1). That is, the dichotic stimulus pairs consisted of either the six possible pairwise combinations of each the three voiced syllables (e.g., /ba/-/da/, /da/-/ba/) or the three unvoiced syllables (e.g., /pa/-/ta/, /ta/-/pa/). The six diotic pairs with the same stimulus presented to both ears (e.g., /ba/-/ba/, /pa/-/pa/, etc.) were also included as control trials (see point 3, Table 1).

The paradigm started with a test block of 10 trials to familiarize the participants with the setup and response procedure. This was followed by three experimental blocks, each containing 46 stimulus pairs (40 dichotic and 6 diotic). Thus, one run of the experiment consists of a total of 120 dichotic (i.e., ten complete presentations of the full set of twelve stimulus pairs) and 18 diotic pairs (i.e., three times the full set of six). The order of trials within each block was pseudorandomized so that consecutive trials did not contain syllables from the same voicing category, preventing direct negative-priming effects ([39], point 4, Table1) and the order was the same for each participant. All stimulus pairs were presented in both orientations (i.e., left-right ear and right-left orientation) in identical frequency in one run of the paradigm (point 8, Table 1). The number of 120 dichotic trials was selected as a literature review has suggested that high reliability can be expected for 90 to 120 trials ([18], see point 5, Table 1).

Each trial was 4000 ms long. A preparation interval of 1000 ms was followed by the stimulus presentation (500 ms) and a response-collection interval of 2500 ms. The stimulus-onset asynchrony was fixed to 4000 ms. A fixation cross (+) was presented in the center of the PC monitor at trial onset but was briefly replaced by a circle (o) to confirm that a response was registered. One trial consisted of a single dichotic stimulus presentation and participants were instructed to report immediately after each stimulus presentation (point 6, Table 1) the one stimulus heard best (free-recall instruction; point 7, Table 1). Response collection was performed via a keyboard using the number pad, on which six response keys (numbers 1 to 6) were marked with the syllable names. There was no possibility to correct a response once given.

The paradigm was administered using both a verbal and a manual response format. The manual response required the participants to log the response themselves. In the verbal version, the participants repeated the best heard syllable aloud after each trial and the experimenter recorded the response. In order not to bias the experimenter’s perception of the participant’s response, the experimenter was blind to which syllables were presented on each trial. One run of the paradigm took approximately 12 min including instructions, and short breaks between the blocks. The paradigm is available as part of the OptDL Project on the OSF platform [45].

For further analyses, the relative difference between the number of correctly recalled left- (Lc) and right-ear (Rc) stimuli was determined as laterality index (i.e., LI  =  (Rc−Lc)/(Rc+Lc)) per person and test run.

Procedure

Each participant completed the paradigm four times, each twice with verbal and manual response format, respectively. The order followed an AABB design, whereby the order was balanced (odd-even by order of recruitment) between individuals, so that n = 26 started with the verbal and n = 24 with the manual version. The setup and test procedure were parallelized at the two sites of data collection (Bergen and Oslo). The same audiometer system and headphone models (Sennheiser HD280) were used at both sites and testing took place in a quiet test room. However, differences in the exact equipment (i.e., computer build, keyboard) were unavoidable. While at both sites E-Prime (Psychology Software Tools, Sharpsburg, PA) was used to control the experiment, version 2.0 was available in Bergen and version 3.0 in Oslo. Nevertheless, we have no reason to believe that these equipment differences had a relevant effect on the present results. Firstly, the effects of interest in the present study are based on comparisons within participants. Secondly, exact timing was not critical as only accuracy and not reaction time data was used.

Statistical analysis

Test-retest reliability was assessed separately for the LI, Lc,and Rc as obtained from the verbal- and the manual-response format paradigm by using intra-class correlations (r). A two-way mixed-models aiming for absolute agreement and estimating the coefficient for a single measure (ICC(A,1) model according to [46]) was employed. That is, a two-way model was chosen as test-retest effects were expected (e.g., due to familiarization with the test procedure) so that the second measure might differ systematically from the first. The observations/participants were randomly assigned. while the measurements were determined by the research question (fixed), constituting a mixed design. As test length is a major determinant of reliability [21], we calculated reliability for the full length paradigm of 120 trials, as well as for the first 40 (i.e., blocks 1 vs. 2) and 80 trials (i.e., block 1 and 2 vs. block 3 and 4) to evaluate the test length effect for the present paradigm. Comparability of verbal and manual response format was evaluated (a) by intra-class correlations between the runs of the two versions, and (b) using a mixed-model analysis of variance (ANOVA). The ANOVA was a three-factorial design, including the repeated-measure factors Repetition (first vs. second run of the paradigm) and Response Format (verbal vs. manual), as well as the between participant factor Sex. The dependent variable was the LI. Statistical analyses were conducted in IBM SPSS Statistics (version 25.0, IBM Corp. Armonk, NY) and GPower (www.gpower.hhu.de, version 3.1) was used for test power calculation. Effect sizes were expressed as Cohen’s d or proportion explained variance (η). The data can be downloaded from the associated OSF platform [45].

Results

Mean laterality for the verbal response format was LI = 33.2 (± 22.9) at the first run of the paradigm and LI = 37.4 (± 22.3) upon retest. For the manual paradigm the mean was LI = 29.9 (± 22.7) at first testing and LI = 34.4 (± 22.8) for the retest. All four LI were significantly larger than 0 (all t(49) > 9.3; all p < .001, all Cohen’s d > 1.27). The mean differences between the two test runs were accordingly 4.4 (± 7.3) for the verbal response paradigm and 4.2 (±8.6) for the manual (both t(49) > 3.5, all p < .001, all d > 0.5). Mean values for the correct report of left- and right-ear stimuli are presented in Table 2.

Table 2

Mean values, standard deviation (s.d.), and reliability (r) a of correct recall of left- (Lc) and right-ear stimuli (Rc).

Response format	Run	Left ear			Right ear
		mean	s.d.	rICC (CI95%)	mean	sd	rICC (CI95%)
Manual	First	40.9	13.5	.93 (.82;97)	75.9	13.7	.93 (.82; 97)
	Retest	38.3	13.3		78.6	13.9
Verbal	First	39.1	13.5	.92 (.85;.96)	78.1	13.9	.88 (.73; .94)
	Retest	37.0	13.2		81.4	13.4

(a) Intra-class correlations determined as ICC(A,1)

The analysis of the full paradigm (i.e. 120 trials) yielded a reliability of r = .93 (95% confidence interval, CI95%: .82 to .97) and .91 (CI95%: .82 to .96) for the verbal and manual response paradigm, respectively (all p < .001). Fig 1 shows the scatterplots illustrating the correspondence between the first and the second measure of both paradigm versions. Considering a single block of 40 trials of the verbal-response version, the reliability was r = .65 (CI95%: .08 to .85), considering two blocks (i.e., 80 trials) it was r = .86 (CI95%: .76 to .92). For a single block in the manual paradigm the reliability was r = .64 (CI95%: .25 to .82); for two blocks it was r = .90 (CI95%: .82 to .94). Reliability estimates for correct report of left- and right-ear stimuli are provided in Table 2.

Fig 1

Scatterplots showing the test-retest correlation for the verbal- and manual-response format version of the optimal paradigm (full length, i.e. 120 dichotic trials).

r = intra-class correlation, using mixed model, considering absolute agreement, and a single measure.

Comparing the verbal- and manual-response version (120 trial paradigm), we found a r = .71 (CI95%: .54 to .82) of the LI of the first verbal test with the LI of the first manual test, and of r = .77 (CI95%: .63 to .86) with the second manual test. For the second run of the verbal test the correlations were r = .75 (CI95%: .53 to .86) with the first, and r = .83 (CI95%: .72 to .90) with the second manual test run.

A significant main effect of Repetition (F(1,48) = 30.7, p < .001) was found in the ANOVA, although with a small effect size (η = 0.01), indicative for a stronger LI at retest compared to first testing (across both response formats). The main effect of Response Format was not significant (F(1,48) = 2.28, p = .14, η<0.01). A sensitivity power analysis (at 5% alpha probability, and with r = .80 correlation of the repeated measures) suggests that medium to large effects (>3% variance explained) can be excluded with a power of .80. Neither the interaction of Repetition and Response Format (F(1,48) = 0.04, p = 0.84, η < 0.01) nor any of the main or interaction effects including the factor Sex (all F(1,48) <1, all p>.50, all η <0.01) were significant.

Discussion

Irrespective of response format the here introduced paradigm shows excellent retest reliability. With values of .91 and .93, the full-length paradigm of 120 trials yielded reliability estimates which are in the upper end of what is typically reported for dichotic-listening paradigms [19]. That is, in the past, test-retest correlations between a minimum of r  =  .63 ([47], using rhyming words as stimuli) and a maximum of r  =  .91 ([48]; using vowel-consonant-vowel stimuli) have been reported for paradigms of the same length. Even as a shortened version of 80 trials, the present paradigm reaches reliability estimates of .86 and .90, which can be considered “good” to “excellent” [49]. In fact, previous studies (even testing with 90 or 100 trials) commonly yield retest correlations below the lower 95%-confidence bound for the reliability estimate of the present 80-trial paradigm [cf. 19] suggesting that the difference is significant. For example, Speaks et al. [50]–using CV syllables as stimuli but demanding to report both stimuli presented per trial–report a retest correlation of r = .71 in a 90 trial paradigm. However, the confidence intervals for the 80-trials version are wider than for the full length paradigm, including reliabilities below .80 suggesting that the full length version should be preferred. Nevertheless, the above comparison of our results to those of various previous paradigms indicates superior reliability of the present paradigm irrespective of test length. Thus, it appears the design features implemented here indeed improve reliability estimates for LIs measured with dichotic listening.

According to the Spearman-Brown formula for test length adjustment, the number of trials is a major determinant of reliability [21] and, accordingly, we here found an increase in reliability from 40, via 80, to 120 trials. Thus, one might speculate that an additional increase in test length would yield further reliability improvement. However, the relationship between test length and reliability is not linear but rather shows diminishing returns with increasing reliability. Applying the Spearman-Brown formula in the present case, an increase to 160 or 200 trials can be predicted to only marginally increase the reliability. For example, considering the verbal version of the paradigm, reliability would be predicted to increase from .93 to .95 and .96, respectively. Furthermore, previous studies using paradigms longer than 120 trials do not report a substantial improvement or even see a reduction in the reliability estimates for higher trial numbers [50, 51]. Using an excessive number of trials also carries the risk of tiring the participants with adverse effects for test performance and reliability, especially in patient samples. Taken together, the here evaluated 120-trial paradigm seems to combine acceptable test length and excellent reliability estimates.

For matter of completeness, the above evaluation of the present paradigm in the context of previous paradigms, also has to consider that the reliability estimates of previous studies are usually reported as product-moment correlations. Product-moment correlations have two short-comings for estimating reliability [46]. Firstly, they do not account for the fact that the correlated measures represent the same “class” (i.e., the measures are repeated measures of the same variable collected with the same test). Secondly, product-moment correlations ignore mean differences between the two measures as the calculation of the correlation coefficient only considers the covariance of the variables. Addressing both issues, the here reported intra-class correlation for absolute agreement represent more appropriate but also slightly more conservative estimates of retest reliability than previous studies [49]. For illustration, regarding the manual paradigm, the product-moment correlation would be r = .95 as opposed to the r = .91 in the present study. Thus, when comparing the present with previously suggested paradigms, it has to be kept in mind that previous studies likely overestimate the “true” reliability of their paradigms.

The intra-class correlations between the measures of the two response format versions were, with values between .71 to .83 well below the reliability estimates. This might suggest that both paradigms measure slightly different aspects of hemispheric specialization. For example, regarding verbal responses, speech production as part of the response phase might modulate the perceptual laterality, whereas a manual response might not [18]. The oral response also has to be understood and registered by the experimenter, which might be associated with coding errors. Or, using the manual response format, the participant–in addition to identifying the syllables–also has to become familiarized with the response procedure and the response key setup, potentially making the task more difficult. At the same time, the main effect of Response Format was non-significant and small, while the test power was sufficient to exclude substantial mean LI differences between the paradigm versions with some confidence. Nevertheless, the reduced inter-correlations indicate that small differences between verbal and manual response format exist, and it is currently not possible to determine whether one of the two versions is better suited for measuring laterality.

As a limitation, it has to be acknowledge that the here presented reliability estimates are obtained on a sample of adult university students and employees with good hearing acuity and diotic identification rates. Thus, it remains to be established if individuals outside these inclusion criteria, especially clinical or aging populations, also provide reliable data, or whether adjustments to the testing procedure are required. For example, a recent study on an aging sample, rather than excluding individuals with hearing deficits, successfully adjusted the sound level of stimulus presentation to compensate for individual differences in the hearing threshold [9]. Future studies are also required to confirm validity of the paradigm. While past studies strongly suggest good validity of paradigms using CV stimulus material [43, 44] a formal test against more direct indicators of hemispheric dominance remains to be conducted.

In summary, we have designed a highly reliable dichotic-listening paradigm by (a) minimizing the relevance of higher cognitive functions on task performance and (b) optimizing crucial features of the stimulus presentation. Point (a) additionally has the beneficial side effect that individual and group differences in higher cognitive functions are less likely to affect the obtained laterality measures, allowing for a fair testing of clinical or developmental samples. Point (b) optimizes the number of trials required to balance the design, reducing the overall number of trials necessary to yield reliable measures (e.g., by reducing trials which are biased due to negative priming etc.), allowing for economic assessment of laterality, at least in experimental settings.

22 Apr 2020

PONE-D-20-06948

An optimal dichotic-listening paradigm for the assessment of hemispheric dominance for speech processing

PLOS ONE

Dear Dr Westerhausen,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

We would appreciate receiving your revised manuscript by Jun 06 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Simone Sulpizio

Academic Editor

PLOS ONE

Journal requirements:

When submitting your revision, we need you to address these additional requirements:

1.    Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at http://www.plosone.org/attachments/PLOSOne_formatting_sample_main_body.pdf and http://www.plosone.org/attachments/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for stating the following in the Acknowledgments Section of your manuscript:

"Funding was provided by the Department of Psychology, University of Oslo, Norway."

We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

"The author(s) received no specific funding for this work."

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Thank you for the opportunity to review this excellent manuscript. The authors have thoughtfully designed an evidence based dichotic listening test that seeks to minimize maximize reliability. Their evidence informed design is well researched and clearly supported by the text. The evaluation is an excellent research design and a significant contribution to the literature. I offer suggestions that I hope will be valuable.

Major issues:

1. The authors should include information about listeners’ absolute scores (right ear and left ear) in addition to laterality index. Because LI uses total correct responses (right plus left) as the denominator, for reliability purposes the number of test items is effectively reduced to whatever number the listener got correct [LI can then be modeled as a binomial]. Knowing the total correct scores are important for readers to be able to draw inferences about generalizability. (i.e., the laterality estimate may be reliable among people with high total scores but have much worse reliability among people with more errors, such as a clinical population.)

2. I request that the authors consider using Repeatability Coefficient (RC) rather than intraclass correlation (ICC) for evaluating test-retest reliability of the same response methods (see Vaz 2013 The Case for Using the Repeatability Coefficient When Calculating Test–Retest Reliability. PLOS ONE). In any case, I would like to see some presentation of data about change-on-retest (how much did individual’s scores change on average? What were the largest changes? Were the changes normally distributed?)

3. Should the authors choose to continue to use ICC to summarize test-retest reliability, I would like to see a justification that includes an explanation of why they have chosen a two-way random model with the test stimuli do not vary among subjects (a one-way random model would be indicated, I think, although this is not at all my area of expertise). I think it is fine to use ICC to compare agreement between the different response paradigms.

Minor issues:

4. The authors should use the term “systematic bias” more judiciously. They are seeking to reduce overall measurement error by minimizing both systematic and random measurement error. As an example, increasing test length serves to reduce random error not systematic bias. Limiting test length avoids the systematic bias on total percent correct scores that would be driven by fatigue.

5. The abstract says: “the paradigm was optimized to account for all variables known to systematically bias task performance.” Be clear with yourself and your readers that the paradigm seeks to reduce measurement error be controlling the stimulus factors know to affect variability. The authors cannot, and do not claim to, control listener health, fatigue, motivation, etc. even though these are variables are known to systematically bias dichotic listening task performance.

6. I think the authors should place more emphasis on the importance of balanced stimulus presentation between the ears (many diagnostic tests using words or digits as stimuli essentially have a list that goes to the right ear and a list to the left ear, so list effects are not controlled). This is a design feature that is a significant improvement on many existing protocols.

7. Similarly I would like to see the authors place more emphasis in the introduction on the comparison of verbal vs manual responses – this is novel and a significant contribution to the literature!

8. I suggest moving from the discussion section into the introduction any references that support the use of dichotic CVs for purpose of measuring hemispheric dominance (e.g., Line 286 to intro “past studies suggest good validity of paradigms using CV stimulus material against more direct indicators of hemispheric dominance”)

9. Pls add a brief limitations section. Acknowledge these results may only be generalizable to educated listeners with symmetrical hearing sensitivity and good diotic scores. The reliability must be replicated in clinical populations, younger/older people, etc.

10. Line 102: I am VERY curious about the person who only identified 81% of diotic stimuli – perhaps this is a person with an auditory processing problem? A person for whom we really want to know their dichotic ability? Pls include in discussion?

11. Section starting with Line 117: Can make this section more parallel with Table 1?

12. Consider adding to discussion – current diagnostic tests are generally short (20-4 pairs). The observation of reliability for a single block of trials is likely closer to the current reliability, but may be better because of the balancing, avoiding negative priming, using stimuli that are more likely to fuse perceptually,

13. Consider adding to discussion – results in line 210 – the effects of practice may be important. Both for reliability, but maybe also for learning (people with APD are unlikely to improve much from a single session of practice? Or people with apd start way low but get better with a little practice?)

14. Line 227-230: This is extremely important result! It suggests that the reliability increase is not due simply to the increase in number of items but rather due to the other factors that were considered in design. Highlight more! Mention it sooner?

Minor suggestions and copy edits:

15. Line 22: “Despite of its popularity” delete the word ‘of’

16. Line 62: “affect the consistency in which individuals perform a given dichotic listening task” wordy. Consider “affect the consistence of individual’s dichotic listening task performance”

17. Line 64: “studies has” to “studies have”

18. Line 43: clarify that the authors are talking about dichotic speech vs dichotic melody or emotion

19. Line 46: “dichotic compared with alternative paradigms” –what are the alternative paradigms that dichotic are better/simpler than? fMRI? Dichotic speech compared to music or emotional content?

20. Line 51-53: If you’re comparing to other paradigms (free-recall in which both are reported or directed attention) do you need to more explicitly acknowledge them?

21. Line 68: “paradigms instructing participants to selectively addend to only one ear at a a time are more difficult” Explain why this is important. There could be situations in which a more difficult test is desirable (e.g., to avoid ceiling effects). (I think you want a test that is driven by perception and difficult tasks are more prone to influence by cognitive factors – make it explicit.)

22. Line 71: Makes me think of tests by Wilson and Moncrieff (random dichotic digits test) and Cameron and Dillon (dichotic digits difference test) that use different approaches to measure the effects of working memory – wonder if that should be addressed in introduction?

23. Line 72: Suggest new paragraph starting with “Ignoring these design variables”

24. Line 73: suggest changing “systematic error variance to the obtained measures” to “measurement error”

25. Line 77: How do we know stimulus driven estimate is about language dominance? Pls explain.

26. Line 82: “All known moderator variables” dial it back – stimulus characteristics or test design variables

27. Line 90: FIRST COMPARISON OF MANUAL VERSUS VERBAL RESPONSE METHOD – THIS IS IMPORTANT!

28. Line 96: Pls start with inclusion/exclusion criteria (Uni Oslo and Bergen, fluent Norwegian, right handed, no known neuor/psych abnormalities, some minimum threshold of hearing sensitivity, symmetrical hearing sensitivity, able to identify stimulus materials diotically… did I get them all?) Then tell us about who presented (52 volunteers) and who was included in the final sample (50), and what criteria the 2 excluded volunteers did not meet. (Or, just leave them out of the conversation all together and just tell us what was true of all 50 participants).

29. Line 106: Pls report absolute hearing sensitivities (did the sample include only those with normal hearing or also those with hearing loss?)

30. Line 141: Was this the number row of a Qwerty keyboard? Were participants able to clearly see the response keys? What happened if they pressed the wrong key (e.g., a “q”)?

31. Lin 146: Blinding the rater was a good idea. Was it possible for the rater to get “off”? (That’s happened to me when scoring CVs on a paper sheet.) Was there any way for the rater to self-correct?

32. Line 151: The laterality index is the only score analyzed, so saying “for further analysis” seems odd. As mentioned before, I think the authors should provide ear specific scores as well.

33. Line 158: “order was balanced between individuals” Does this refer to the response method order? How was this decided? Randomized? Alternating? Listener preference? All manual responses first for the first 24 subjects?

34. Line 162: delete “headphone models” from this list of exact equipment unless Sennheiser HD280 reported above is not accurate.

35. Line 168: In stimuli and procedure the authors failed to indicate the presentation level of the stimuli and whether the stimuli were presented in pseudo-random order or from a pre-ordered list or recording.

36. Line 172: two way mixed models – why not a one way model if the stimuli are the same each time

37. Line 247: “acceptable test length” – acceptable is a subjective assessment. Some listeners may not be able to tolerate 120 items, some clinicians may not be willing to spend 12 min to obtain a laterality estimate (that’s how long it takes me to do a comprehensive audiogram with air, bone, srt, and speech on a compliant adult with symmetric sensorineural hearing loss and normal word recognition – I’m unlikely to add a 12 minute protocol to my routine clinical practice)

Other Comments:

It is outside the scope of this paper, but I am curious if the authors have the data on individual responses to each item – just like there is a “short list” to estimate word recognition ability (stimuli organized by difficulty, if listeners get the first 10 items correct, it is very unlikely they have abnormally low ability, the test can be stopped) I can easily imagine a “short list” dichotic screening (using stimulus dominance information) that is built from the authors’ existing stimuli.

Reviewer #2: This manuscript is interesting and well written.

I ask only for some minor revisions:

Line 20 test or tests? please evaluate

Line 86 cognitive functions

Line 88 perhaps: the total paradigm? or better the entire paradigm

Line 89 allows to evaluate

Line 106 smaller than 10 dB

Lines 102,107 please delete However

Lines 103,108 please delete Furthermore

Line 151 perhaps: analyses

Lines 161, 163 at > in

Results (Lines 189-192): please comment on the relatively high differences in the mean lateralities found. It seems that DL has still some measurements problems.

Line 225: the same lenght.

I strongly agree with the last sentence of the manuscript (lines 285-287) and it would be nice if Authors would comment somewhat further the issue (e.g. concerning the potential use of DL in surgical settings to assess language laterality)

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Rocky Kairn Kelley, PhD MS/CCC-A

Reviewer #2: Yes: Alfredo Brancucci

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

28 Apr 2020

Response to reviewer comments, ms PONE-D-20-06948

We would like to thank the reviewers for thoughtful and very encouraging comments, as well as the editor for allow us to revise the manuscript. Please find a point-by-point response below. All changes to the manuscript are marked “red”.

Reviewer #1

Point 1. The authors should include information about listeners’ absolute scores (right ear and left ear) in addition to laterality index. Because LI uses total correct responses (right plus left) as the denominator, for reliability purposes the number of test items is effectively reduced to whatever number the listener got correct [LI can then be modeled as a binomial]. Knowing the total correct scores are important for readers to be able to draw inferences about generalizability. (i.e., the laterality estimate may be reliable among people with high total scores but have much worse reliability among people with more errors, such as a clinical population.)

Response: We have now calculated reliability separately for left and right ear correct responses and corresponding means, sd, and ICCs are provided in Table 2.

Point 2. I request that the authors consider using Repeatability Coefficient (RC) rather than intraclass correlation (ICC) for evaluating test-retest reliability of the same response methods (see Vaz 2013 The Case for Using the Repeatability Coefficient When Calculating Test–Retest Reliability. PLOS ONE). In any case, I would like to see some presentation of data about change-on-retest (how much did individual’s scores change on average? What were the largest changes? Were the changes normally distributed?)

Response: We were not familiar with the RC, but after consulting the reference paper, we feel that using the “absolute agreement” within the ICC framework accounts as well for both change in mean LI and covariation. However, as suggested, the average change between test 1 and 2 (retest) are now provided in the beginning of the result section. As you can see in the graph below [see uploaded docx version], the distribution in both cases is fairly normal. We are not sure, however, if adding these graphs to the manuscript is really necessary, especially since the scatterplots in Fig 1 nicely (in our opinion) illustrates the modest variability between test time points (i.e. change).

Point 3. Should the authors choose to continue to use ICC to summarize test-retest reliability, I would like to see a justification that includes an explanation of why they have chosen a two-way random model with the test stimuli do not vary among subjects (a one-way random model would be indicated, I think, although this is not at all my area of expertise). I think it is fine to use ICC to compare agreement between the different response paradigms.

Response: In fact we used a two-way mixed model (not two-way random model), but we have now specified this choice as follows (see Statistical Analysis section): “That is, a two-way model was chosen as test-retest effects were expected (e.g., due to familiarization with the test procedure) so that the second measure might differ systematically from the first. The observations/participants were randomly assigned while the measurements was determined by the research question (fixed), constituting a mixed design.”

Point 4. The authors should use the term “systematic bias” more judiciously. They are seeking to reduce overall measurement error by minimizing both systematic and random measurement error. As an example, increasing test length serves to reduce random error not systematic bias. Limiting test length avoids the systematic bias on total percent correct scores that would be driven by fatigue.

Response: Yes, arguably, we were not stringent in our description. Thus, we carefully reworded the text accordingly.

Point 5. The abstract says: “the paradigm was optimized to account for all variables known to systematically bias task performance.” Be clear with yourself and your readers that the paradigm seeks to reduce measurement error be controlling the stimulus factors know to affect variability. The authors cannot, and do not claim to, control listener health, fatigue, motivation, etc. even though these are variables are known to systematically bias dichotic listening task performance.

Response: Again, we carefully reworded the text accordingly.

Point 6. I think the authors should place more emphasis on the importance of balanced stimulus presentation between the ears (many diagnostic tests using words or digits as stimuli essentially have a list that goes to the right ear and a list to the left ear, so list effects are not controlled). This is a design feature that is a significant improvement on many existing protocols.

Response: Yes, we agree, this is an important design feature, maybe the most important. We are, however, hesitant in discussing this aspect in more detail here as this would simply reiterate the arguments thoroughly discussed in the referenced review paper (Westerhausen, 2019, Laterality). This review paper discusses this and many other general principles of how to design fair paradigm and is available as open access, i.e. readily available for everyone (https://doi.org/10.1080/1357650X.2019.1598426). In the present manuscript, we thus would rather prefer to focus on the reliability check of the paradigm which was designed accordingly. Table 1 was included to provide an overview of the main criteria we employed designing the paradigm, and the arguments are partially repeated in the method section..

Point 7. Similarly I would like to see the authors place more emphasis in the introduction on the comparison of verbal vs manual responses – this is novel and a significant contribution to the literature!

Response: Following the reviewers suggestion, we introduced a new paragraph in the introduction discussing the lack of such comparative studies.

Point 8. I suggest moving from the discussion section into the introduction any references that support the use of dichotic CVs for purpose of measuring hemispheric dominance (e.g., Line 286 to intro “past studies suggest good validity of paradigms using CV stimulus material against more direct indicators of hemispheric dominance”)

Response: As in response to point 6, we feel Table 1 covers this point sufficiently well considering the recent review paper (Westerhausen, 2019, Laterality) we are referencing here.

Point 9. Pls add a brief limitations section. Acknowledge these results may only be generalizable to educated listeners with symmetrical hearing sensitivity and good diotic scores. The reliability must be replicated in clinical populations, younger/older people, etc.

10. Line 102: I am VERY curious about the person who only identified 81% of diotic stimuli – perhaps this is a person with an auditory processing problem? A person for whom we really want to know their dichotic ability? Pls include in discussion?

Response: We added the limitation section as requested. However, we are not able to follow up this excluded individual, and any such speculation is certainly beyond what the data allows. The overall laterality measures of this individual were in a normal range (23-64%) and hearing acuity was also unremarkable. Thus, our first interpretation would be that this individual’s diotic results occurred by chance rather than due some underlying pathology. We found the exclusion necessary to make sure the sample is somewhat homogenous. However, including this individual also does not change the present rICC estimates. But we wholeheartedly agree, a reliability analysis in clinical sample should be done before applying it in diagnostics. From a basic research perspective, however, we feel the paradigm is a promising alternative to older established paradigms.

Point 11. Section starting with Line 117: Can make this section more parallel with Table 1?

Response: We have adjusted Table 1 by numbering and reordering the lines to parallel the order of the design features. We also introduced direct references to the table to make the relationship clearer.

Point 12. Consider adding to discussion – current diagnostic tests are generally short (20-4 pairs). The observation of reliability for a single block of trials is likely closer to the current reliability, but may be better because of the balancing, avoiding negative priming, using stimuli that are more likely to fuse perceptually

Response: We feel any such discussion should be based on a direct comparison in a clinical group. Also, the present paper primarily aims to design a paradigm to assess hemispheric dominance for speech/language processing (as outlined in the introduction and see title). Thus, we would prefer not to delve into this.

Point 13. Consider adding to discussion – results in line 210 – the effects of practice may be important. Both for reliability, but maybe also for learning (people with APD are unlikely to improve much from a single session of practice? Or people with apd start way low but get better with a little practice?)

Response: Great point! However, like the previous point, without any evidence we would prefer to not take up this discussion.

Point 14. Line 227-230: This is extremely important result! It suggests that the reliability increase is not due simply to the increase in number of items but rather due to the other factors that were considered in design. Highlight more! Mention it sooner?

Response: We extended the paragraph to highlight this more.

Minor suggestions and copy edits:

Point 15. Line 22: “Despite of its popularity” delete the word ‘of’

Point 16. Line 62: “affect the consistency in which individuals perform a given dichotic listening task” wordy. Consider “affect the consistence of individual’s dichotic listening task performance”

Response: Both corrected as suggested!

Point 17. Line 64: “studies has” to “studies have”

Response: As “a plethora” is singular, we would prefer to keep the following verb in singular too. But interpretations seem acceptable: https://www.merriam-webster.com/words-at-play/plethora-singular-or-plural

Point 18. Line 43: clarify that the authors are talking about dichotic speech vs dichotic melody or emotion

Point 19. Line 46: “dichotic compared with alternative paradigms” –what are the alternative paradigms that dichotic are better/simpler than? fMRI? Dichotic speech compared to music or emotional content?

Response: Specified a suggested.

Point 20. Line 51-53: If you’re comparing to other paradigms (free-recall in which both are reported or directed attention) do you need to more explicitly acknowledge them?

Response: We feel the paragraph makes sense as it is right now. But we are obviously explicitly referring to these more demanding paradigms in the next paragraph (now line 67 onwards).

Point 21. Line 68: “paradigms instructing participants to selectively addend to only one ear at a a time are more difficult” Explain why this is important. There could be situations in which a more difficult test is desirable (e.g., to avoid ceiling effects). (I think you want a test that is driven by perception and difficult tasks are more prone to influence by cognitive factors – make it explicit.)

Response: Absolutely! However, in the mentioned section we list examples for design features and their consequences for the paradigm. And we later specify that the primary aim is to assess hemispheric dominance for language (we now reworked the section to makes this clearer). And here, for sure, any “cognitive component” should be minimised (see next paragraph, lines 74 onwards). We now reworked abstract and introduction to more clearly emphasise that the primary purpose of the new paradigm is to assess hemispheric dominance.

Point 22. Line 71: Makes me think of tests by Wilson and Moncrieff (random dichotic digits test) and Cameron and Dillon (dichotic digits difference test) that use different approaches to measure the effects of working memory – wonder if that should be addressed in introduction?

Response: Rather feel that would distract from the main purpose. But in general it would be a great idea to have a systematic look at the paradigms used in APD and other speech impediment diagnostics under consideration of the background (Westerhausen, 2019, Laterality) paper. If the reviewer would be interested we attempt doing a joined critical review on what is done.

Point 23. Line 72: Suggest new paragraph starting with “Ignoring these design variables”

Point 24. Line 73: suggest changing “systematic error variance to the obtained measures” to “measurement error”

Response: Both done as suggested.

Point 25. Line 77: How do we know stimulus driven estimate is about language dominance? Pls explain.

Response: We have include a section into the introduction elaborating the assumed theoretical framework (line 74 onwards), i.e. our understanding of dichotic listening. A more detailed outline can be found in another recent reference (Westerhausen & Kompus, 2018, Scan J Psychol; https://doi.org/10.1111/sjop.12408).

Point 26. Line 82: “All known moderator variables” dial it back – stimulus characteristics or test design variables

Response: “Dialed down”

Point 27. Line 90: FIRST COMPARISON OF MANUAL VERSUS VERBAL RESPONSE METHOD – THIS IS IMPORTANT!

Response: Thanks for sharing our enthusiasm for this work. As indicated in response to point 7, a new paragraph was added to the introduction appreciating the importance of the response format comparison.

Point 28. Line 96: Pls start with inclusion/exclusion criteria (Uni Oslo and Bergen, fluent Norwegian, right handed, no known neuor/psych abnormalities, some minimum threshold of hearing sensitivity, symmetrical hearing sensitivity, able to identify stimulus materials diotically… did I get them all?) Then tell us about who presented (52 volunteers) and who was included in the final sample (50), and what criteria the 2 excluded volunteers did not meet. (Or, just leave them out of the conversation all together and just tell us what was true of all 50 participants).

Point 29. Line 106: Pls report absolute hearing sensitivities (did the sample include only those with normal hearing or also those with hearing loss?)

Response: we restructured the section accordingly and absolute hearing info was provided. We prefer to report all participants recruited and tested for matter of completeness.

point 30. Line 141: Was this the number row of a Qwerty keyboard? Were participants able to clearly see the response keys? What happened if they pressed the wrong key (e.g., a “q”)?

Response: As written in text it was the number pad (not the row) and there was no possibility to correct the response (also to avoid long thinking about the response, which might alter things) as now indicated in text. The buttons were clearly marked.

Point 31. Lin 146: Blinding the rater was a good idea. Was it possible for the rater to get “off”? (That’s happened to me when scoring CVs on a paper sheet.) Was there any way for the rater to self-correct?

Response: No possibility to correct either, as the same response set up was used as for the manual response. Possible “rater errors” we considered as “noise” in the system but considering the high number of correct answers (see Table 2), cannot have been frequent.

Point 32. Line 151: The laterality index is the only score analyzed, so saying “for further analysis” seems odd. As mentioned before, I think the authors should provide ear specific scores as well.

Response: we now provide the analysis of the correct left and right ear score

Point 33. Line 158: “order was balanced between individuals” Does this refer to the response method order? How was this decided? Randomized? Alternating? Listener preference? All manual responses first for the first 24 subjects?

Response: Yes, response format; and “odd-even” split to get the same amount of individuals starting with verbal and manual response

Point 34. Line 162: delete “headphone models” from this list of exact equipment unless Sennheiser HD280 reported above is not accurate.

Response: Yes, same model at each site. Thanks for spotting!

Point 35. Line 168: In stimuli and procedure the authors failed to indicate the presentation level of the stimuli and whether the stimuli were presented in pseudo-random order or from a pre-ordered list or recording.

Response: information now provided (line 158)

Point 36. Line 172: two way mixed models – why not a one way model if the stimuli are the same each time

Response: Please refer to point 3

point 37. Line 247: “acceptable test length” – acceptable is a subjective assessment. Some listeners may not be able to tolerate 120 items, some clinicians may not be willing to spend 12 min to obtain a laterality estimate (that’s how long it takes me to do a comprehensive audiogram with air, bone, srt, and speech on a compliant adult with symmetric sensorineural hearing loss and normal word recognition – I’m unlikely to add a 12 minute protocol to my routine clinical practice)

Response: We were more thinking of experiments. Thanks for pointing this out, we specified this comment now to read “at least in experimental settings”. But for the assessment of LIs, our data suggests that 80 trials, so about 8 mins, seem to be the minimum (in healthy indivdiuals)

Other Comments:

It is outside the scope of this paper, but I am curious if the authors have the data on individual responses to each item – just like there is a “short list” to estimate word recognition ability (stimuli organized by difficulty, if listeners get the first 10 items correct, it is very unlikely they have abnormally low ability, the test can be stopped) I can easily imagine a “short list” dichotic screening (using stimulus dominance information) that is built from the authors’ existing stimuli.

Response: We are happy to share the excel files of the trial by trial response if you would like to explore the data deeper. Although the experiment only consists of 12 different stimulus pairs, and if one wants to avoid negative priming one might be able to reduce it to 8 pairs. One pair for each voicing combination and presented in the left-right and right-left order.

Reviewer #2

Line 20 test or tests? please evaluate

Response: “Tests”, thanks for pointing out the inconsistency

Line 86 cognitive functions

Response: Sentence removed in revision

Line 88 perhaps: the total paradigm? or better the entire paradigm

Line 89 allows to evaluate

Line 106 smaller than 10 dB

Lines 102,107 please delete However

Lines 103,108 please delete Furthermore

Line 151 perhaps: analyses

Lines 161, 163 at > in

Response: corrected as suggested

Results (Lines 189-192): please comment on the relatively high differences in the mean lateralities found. It seems that DL has still some measurements problems.

Response: This translates to the small main effect of Repetition we find in the ANOVA, which we accordingly discuss.

Line 225: the same lenght.

Response: corrected, thanks again

I strongly agree with the last sentence of the manuscript (lines 285-287) and it would be nice if Authors would comment somewhat further the issue issue (e.g. concerning the potential use of DL in surgical settings to assess language laterality)

Response: Thanks! If one should use the paradigm for planning surgery would be more question of validity than reliability, which we now discuss in the “limitations” paragraph. However, review one finds the paradigm too long for being of use in audiological diagnostics.

Submitted filename: Response to reviewer comments.docx

Click here for additional data file.

1 Jun 2020

An optimal dichotic-listening paradigm for the assessment of hemispheric dominance for speech processing

PONE-D-20-06948R1

Dear Dr. Westerhausen,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Simone Sulpizio

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This is an excellent manuscript and a valuable new test paradigm for improved measurement of dichotic listening. The authors have met or exceeded all standards required for publication.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

3 Jun 2020

PONE-D-20-06948R1

An optimal dichotic-listening paradigm for the assessment of hemispheric dominance for speech processing

Dear Dr. Westerhausen:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Simone Sulpizio

Academic Editor

PLOS ONE