Turn-Taking Transitions in Conversations Among Autistic–Autistic, Non-Autistic–Non-Autistic, and Mixed-Neurotype Adult Pairs

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for the opportunity to review this interesting paper on turn-taking transitions in pairs of speakers with varying neurotypes. This is an important topic and the results support neurotype mismatch as the important factor in interactions, rather than implying that autistic people have difficulties in the way they communicate, in line with the Double Empathy Problem. This is important for a range of reasons, including how we focus interventions to support communication, which need to focus on mutual understanding of communication styles across neurotypes, rather than making autistic people’s communication fit the neurotypical style.

The paper is overall clear and well-written but I have made some suggestions for revisions to develop it further.

 

Abstract

The abstract is clearly written and provides the key information needed to understand the study.

 

Introduction

The Introduction is overall clearly written and provides most of the key information needed to contextualise the study. Research aims and questions are clearly stated. I have a few suggestions for minor revisions listed below:

• Page 2, Lines 57-59 – this is a very short paragraph at present and I think it would create a smoother flow of information if this was closed up to the paragraph above.
• Page 2, Line 61 – it appears here that “neurotypical” is being contrasted with “autistic” as if these are opposite terms. There are many people who fall into neither of these categories, and I think that “non-autistic” may be a more accurate term if this is what is meant by the authors.
• Page 2, Lines 81-85 – this is also a very short paragraph that could be closed up to the paragraph above to support flow.
• Page 2, Line 81 – I think the word “a” is missing from this sentence e.g. “a study…..”
• Page 2, Lines 87 -90 – this statement seems in contradiction to the evidence provided in the section above about previous research, where studies about timing (as well as other aspects) of turn taking are discussed. This part may therefore need to be reworded for clarity. If the major novel contribution is in combining all of these aspects into one study, or studying all aspects with all three groups represented (or both) it would be good to state this more explicitly.
• Page 3, Lines 123-130 – are the studies you have cited in support of your hypotheses here about autistic or non-autistic participants? If they are based on non-autistic participants, it would be good to explicitly state whether the authors think autistic participants will value or benefit from the same characteristics of interaction – some of the research cited earlier in the paper suggests they may not.

 

Methods

Methods are overall clearly described, giving the reader a good sense of the participants, the procedure and the coding scheme, and I appreciate the inclusion of information about inter-rater reliability. I have some minor comments listed below to develop the Methods further:

• Characteristics of participants are presented clearly; it would be good to add a sentence or two stating whether characteristics are thought to be overall consistent across groups or whether the two groups differ in any aspects of their profiles.
• Most of the coding scheme is described clearly but I am less clear about the distinctions between simultaneous onset, simultaneous offset and simultaneous talk. I think it would be beneficial to the reader to clarify this a little more.
• Rapport is mentioned in the Methods but it is not clear how this was scored. It would be good to provide more details about this to help the reader understand the results.
• This is a minor point but it would be good to be specific in the Data Analysis section about which statistical software was used to generate results.

 

Results

The results are very clearly presented and easy to understand. Graphs are used to good effect. I have only one very minor proof-reading comment on this section:

• Page 11, Line 323 – I think a bracket is missing in this line somewhere.

 

Discussion

The Discussion is overall clear, and contextualises the results of the study well. I have some minor comments, listed below:

• Page 12, Lines 337-338 – is this previous literature based on non-autistic interactions, or also on interactions between autistic people? This is important for how the results are contextualized.
• Page 12, Lines 342-345 – The two explanations presented here seem contradictory to each other at my reading, and a little more explanation may be needed here. I think you might be suggesting that the autistic-autistic context is more comfortable for participants, but this very much implied at present and might benefit from being made more explicit if that is the argument being made.
• Page 13, Lines 415-418 – the sentence here is very long and not easy to read and understand. It might of benefit to split this into two shorter sentences to clarify the meaning.
• Page 14, Lines 443-445 – I think there are some grammatical errors in this sentence that make it hard to interpret. There are more gender-diverse people overall in the autistic group, it seems, so are you suggesting you might need more of these in the non-autistic group to make the groups comparable? Otherwise, I am not sure why gender diversity is being focused on here and not other characteristics.
• The interpretation of results around rapport is hard to judge at present without knowing how rapport is measured – adding this to the methods will help with the interpretation of these aspects of the discussion.

Author Response

Comment 1:

Comments and Suggestions for Authors

Thank you for the opportunity to review this interesting paper on turn-taking transitions in pairs of speakers with varying neurotypes. This is an important topic, and the results support neurotype mismatch as the important factor in interactions, rather than implying that autistic people have difficulties in the way they communicate, in line with the Double Empathy Problem. This is important for a range of reasons, including how we focus interventions to support communication, which need to focus on mutual understanding of communication styles across neurotypes, rather than making autistic people’s communication fit the neurotypical style.

The paper is clear and well-written overall, but I have made some suggestions for revisions to develop it further.

Response 1:

Thank you very much for taking the time to review this manuscript. We greatly appreciate your comments.

Comment 2:

Abstract: The abstract is clearly written and provides the key information needed to understand the study.

Response 2:

Thank you.

Comment 3:

Introduction: The Introduction is overall clearly written and provides most of the key information needed to contextualize the study. Research aims and questions are clearly stated. I have a few suggestions for minor revisions listed below:

Response 3:

Thank you. We have tried to address all comments and make the manuscript clearer in this section. Please see the respond for each section below.

Comment 3A:

Page 2, Lines 57-59 – this is a very short paragraph at present, and I think it would create a smoother flow of information if this was closed up to the paragraph above.

Response 3A:

We have revised the manuscript by merging this short paragraph with the preceding paragraph (please see line 59 in the revised version).

Comment 3B:

Page 2, Line 61 – It appears here that “neurotypical” is being contrasted with “autistic” as if these are opposite terms. There are many people who fall into neither of these categories, and I think that “non-autistic” may be a more accurate term if this is what is meant by the authors.

Response 3B:

Thank you for this helpful comment. We have revised the manuscript throughout to use “non-autistic” instead of “neurotypical” to avoid opposition with “autistic” and ensure consistent terminology across the paper.

Comment 3C:

Page 2, Lines 81-85 – this is also a very short paragraph that could be closed up to the paragraph above to support flow.

Response 3C:

We have revised the manuscript by merging this short paragraph with the preceding paragraph (please see line 87 in the revised version).

Comment 3D:

Page 2, Line 81 – I think the word “a” is missing from this sentence e.g. “a study…..”

Response 3D:

We corrected the sentence by adding the missing article (“a”) for clarity (please see line 88 in the revised version).

Comment 3E:

Page 2, Lines 87 -90 – this statement seems in contradiction to the evidence provided in the section above about previous research, where studies about timing (as well as other aspects) of turn taking are discussed. This part may therefore need to be reworded for clarity. If the major novel contribution is in combining all of these aspects into one study, or studying all aspects with all three groups represented (or both) it would be good to state this more explicitly.

Response 3E:

Thank you for this comment. We agree that our original wording was not clear. We have revised this section to clarify that previous studies have examined timing and turn-taking, but they have not included all three pair types together or all aspects in one study. We now state more clearly that the main contribution of our study is combining multiple measures of temporal coordination and including all three pair conditions in a single design (please see lines 95-103 in the revised version).

Comment 3F:

Page 3, Lines 123-130 – are the studies you have cited in support of your hypotheses here about autistic or non-autistic participants? If they are based on non-autistic participants, it would be good to explicitly state whether the authors think autistic participants will value or benefit from the same characteristics of interaction – some of the research cited earlier in the paper suggests they may not.

Response 3F:

Yes, the studies that we cited here (Rifai et al., 2022; Geelhand et al., 2024; Jameson & Bean, 2025) are about task performance in autistic-autistic,  non-autistic- non-autistic, and mixed neurotype pairs.

Comment 4:  

Methods are overall clearly described, giving the reader a good sense of the participants, the procedure and the coding scheme, and I appreciate the inclusion of information about inter-rater reliability. I have some minor comments listed below to develop the Methods further:

Response 4:

Thank you. We have tried to address all comments and make the manuscript clearer in this section. Please see the respond for each section below.

Comment 4A:

Characteristics of participants are presented clearly; it would be good to add a sentence or two stating whether characteristics are thought to be overall consistent across groups or whether the two groups differ in any aspects of their profiles.

Response 4A:

We have added a clarifying sentence in the Participants section to describe the extent to which participant characteristics were comparable across groups. Participants (autistic and non-autistics) were broadly similar in age distribution and showed overlap in educational background, while there were differences in gender diversity and racial/ethnic representation (please see lines 156-161 in the revised version).

Comment 4B:

Most of the coding scheme is described clearly, but I am less clear about the distinctions between simultaneous onset, simultaneous offset, and simultaneous talk. I think it would be beneficial to the reader to clarify this a little more.

Response 4B:

We have clarified the distinctions between simultaneous onset, simultaneous offset, and simultaneous talk by adding brief descriptions to each definition, specifying differences in timing (i.e., shared onset, shared offset, or sustained overlap; please see lines 240-245 in the revised version).

Comment 4C:

Rapport is mentioned in the Methods, but it is not clear how this was scored. It would be good to provide more details about this to help the reader understand the results.

Response 4C:

We have revised the Methods section (Recording Procedure) to clarify how rapport was measured,  including a description of the self-report questionnaire, the five dimensions assessed (ease, enjoyment, success, friendliness, and awkwardness; please see lines 206-210 in the revised version).

Comment 4D:

This is a minor point, but it would be good to be specific in the Data Analysis section about which statistical software was used to generate results.

Response 4D:

We have revised the Data Analysis section to specify that all analyses were conducted using R statistical software (R Core Team, 2025; please see line 267 in the revised version). 

Comment 5:

Results: The results are very clearly presented and easy to understand. Graphs are used to good effect. I have only one very minor proof-reading comment on this section: Page 11, Line 323 – I think a bracket is missing in this line somewhere.

Response 5:

We have corrected the missing bracket in the revised manuscript. Please see line 371 in the revised version.

Comment 6:

Discussion: The Discussion is clear overall and contextualizes the results of the study well. I have some minor comments, listed below:

Response 6:

Thank you. We have tried to address all comments and make the manuscript clearer in this section. Please see the respond for each section below.

Comment 6A:

Page 12, Lines 337-338 – is this previous literature based on non-autistic interactions, or also on interactions between autistic people? This is important for how the results are contextualized.

Response 6A:

The citations we previously included (Garrod & Pickering, 2015; Heldner & Edlund, 2010; Ruiter et al., 2006) are based only on non-autistic participants. However, we have added a citation (Wehrle et al., 2023) that considers both non-autistic–non-autistic and autistic–autistic conversations. We also clarified that these effects have been examined in both autistic–autistic and non-autistic–non-autistic pairs. Please see lines 386-387 in the revised version.

Comment 6B:

Page 12, Lines 342-345 – The two explanations presented here seem contradictory to each other at my reading, and a little more explanation may be needed here. I think you might be suggesting that the autistic-autistic context is more comfortable for participants, but this very much implied at present and might benefit from being made more explicit if that is the argument being made.

Response 6B:

We agree that the two explanations were not clear in the previous version. In the revised manuscript, we have clarified that higher coordination in autistic–autistic pairs may reflect interaction within a similar neurotype, which can create a more familiar and comfortable interactional context, thereby facilitating more rapid and synchronized timing. Please see lines 391-395 in the revised version.

Comment 6C:

Page 13, Lines 415-418 – the sentence here is very long and not easy to read and understand. It might of benefit to split this into two shorter sentences to clarify the meaning.

Response 6C:

In the revised manuscript, we split it into two shorter sentences and clarified the wording to improve clarity. Please see lines 464-471 in the revised version.

Comment 6D:

Page 14, Lines 443-445 – I think there are some grammatical errors in this sentence that make it hard to interpret. There is more gender-diverse people overall in the autistic group, it seems, so are you suggesting you might need more of these in the non-autistic group to make the groups comparable? Otherwise, I am not sure why gender diversity is being focused on here and no other characteristics.

Response 6D:  

We have revised the grammatical errors. Besides, in the revised manuscript, we clarified that our intention was to emphasize the importance of greater diversity across groups, including gender diversity, to improve comparability and generalizability, rather than focusing on one factor alone. Please see lines 492-498 in the revised version.

Comment 6E:

The interpretation of results around rapport is hard to judge at present without knowing how rapport is measured – adding this to the methods will help with the interpretation of these aspects of the discussion.

Response 6E:

We have revised the Methods section (Recording Procedure) to clarify how rapport was measured,  including a description of the self-report questionnaire, the five dimensions assessed (ease, enjoyment, success, friendliness, and awkwardness; please see lines 206-210 in the revised version).

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for the opportunity to review this manuscript.

 

Title, Abstract, and Introduction

Overall, the introduction addresses an important and timely topic and appropriately situates the study within the Double Empathy Theory framework. The manuscript also identifies a meaningful gap in the literature. However, several areas could be strengthened to improve clarity, coherence, and justification of the study.

The title generally aligns with the study focus. However, it appears to contain formatting inconsistencies. Please revise the title for clarity.

Lines 44 to 48
The introduction identifies the relevance of the Double Empathy Theory. However, the research gap could be articulated more explicitly. While you state that ‘relatively few studies have examined turn-exchange within the context of double-empathy frameworks,’ it would strengthen the argument to more clearly specify what exactly is missing.

Lines 57 to 59
The statement that neurotype mismatches may influence turn-taking dynamics is important, but it is presented somewhat briefly. Consider expanding this section to more explicitly connect the Double Empathy Theory to specific measurable variables in your study

Lines 86 to 89
The research gap is stated clearly in this part. (no research study that compares all three pair conditions…), but this key point appears relatively late in the introduction.

Lines 81 to 84
The transition beginning with ‘In contrast’ is unclear. It is not immediately evident what this section is contrasting with from prior studies in the previous sentences. Clarify the contrast and ensure consistency in how studies are grouped.

Lines 123 to 130
The link between turn-taking dynamics and task performance is relevant, but the transition from conversational variables to performance outcomes feels somewhat abrupt. Consider adding a brief bridging sentence to clarify why these outcomes are conceptually connected within your framework.

Minot issues:

Line 61
Use consistent terminology (e.g., neurotypical vs non-autistic). The manuscript currently uses both terms interchangeably, which may reduce clarity.

Ensure consistent formatting (e.g., removal of line-break artifacts such as autistic–au-tistic).

 

Methods

Lines 139 to 142
The inclusion of both formally diagnosed and self-diagnosed autistic participants is understandable. However, additional detail is needed regarding how self-diagnosis was operationalized.

Lines 158 to 159
There appears to be some inconsistency or potential confusion regarding participant counts. The total number of participants is reported as 32 autistic and 36 non-autistic individuals, but later the dyad composition is described differently. Please clarify how these totals map onto the dyad structure and whether any participants were excluded.

Lines 152 to 153
Participants were not informed of their dyad type. While this is appropriate, it would be helpful to briefly explain the rationale.

Lines 169 to 177
The tangram task is appropriate and well-described. However, consider briefly justifying why this task is suitable for examining turn-taking dynamics. For example, clarify how it elicits interactional variability relevant to your variables.

Lines 187 to 192
The criteria for utterance segmentation are clearly described. However, the use of coder judgment for the 200–300 ms range introduces subjectivity. Please clarify.

Lines 195 to 212
The coding categories are comprehensive while several categories appear conceptually overlapping. Please clarify how these categories are distinguished in practice to ensure they are mutually exclusive and reliably coded.

Lines 226 to 228
Statistical tests is appropriate, but additional detail would improve clarity.

Which specific models were used for each research question and any corrections for multiple comparisons if there was any.

Lines 230 to 231
Data were aggregated at the dyad level. Please justify this decision, particularly given that turn-taking events are nested within dyads.

Lines 240 to 242
Rapport is defined conceptually, but the method of measurement is unclear. Please clarify.

 

Results

Overall, the results are clearly structured and aligned with the research questions. But several areas could be improved to enhance clarity.

Lines 249 to 253
The chi-square result is reported appropriately, but the effect size is very small (Cramér’s V = .044). Please consider briefly acknowledging the practical significance (or lack thereof) of this effect.

Lines 265 to 267
The use of a Gamma GLM with a log link is appropriate for skewed duration data. However, please clarify whether any model diagnostics were assessed.

Lines 309 to 317
The modeling approach for RQ5 is described, but it is somewhat unclear whether separate models were run for each outcome and predictor. Please clarify the modeling structure.

 

Lines 318 to 324
The sentence structure is somewhat unclear and difficult to follow (“when we compare mixed neurotype pairs to either…”). Consider revising for clarity and grammatical precision.

 

Discussion, limitation/future studies and importance of study

Lines 341 to 346
The interpretation that autistic–autistic pairs are interesting but somewhat speculative. Please consider softening the language or more clearly distinguishing between interpretation and empirical evidence.

Lines 350 to 357
The explanation of longer gap durations in mixed dyads is plausible but remains speculative. Consider explicitly acknowledging these as possible interpretations rather than definitive explanations.

 

Lines 358 to 362
The connection to the Double Empathy Theory is appropriate. This argument could be strengthened by more explicitly explaining how the observed timing differences reflect mutual misalignment rather than individual deficits.

Lines 381 to 387
The discussion appropriately acknowledges that simultaneous talk may reflect both coordination and technical artifacts. This is a strength. However, given this limitation, consider slightly tempering earlier interpretations that frame simultaneous talk primarily as a marker of coordination.

 

Lines 412 to 418
The sentence structure is somewhat unclear and difficult to interpret. Please revise for clarity, particularly the conclusion that overlap reflects shared expectations rather than deficits.

Lines 424 to 426
There appears to be an inconsistency with the Methods regarding demographic matching. Please clarify.  

Lines 439 to 445
This section could be more concise.

Author Response

Author's Reply to the Review Report (Reviewer 2):

Comment 1:

Thank you for the opportunity to review this manuscript.

Response 1:

Thank you very much for taking the time to review this manuscript. We greatly appreciate your comments.

Comment 2:

Title, Abstract, and Introduction

Overall, the introduction addresses an important and timely topic and appropriately situates the study within the Double Empathy Theory framework. The manuscript also identifies a meaningful gap in the literature. However, several areas could be strengthened to improve clarity, coherence, and justification of the study.

Response 2:

Thank you. We have tried to address all comments and make the manuscript clearer in this section. Please see the respond for each section below.

Comment 2A:

The title generally aligns with the study focus. However, it appears to contain formatting inconsistencies. Please revise the title for clarity.

Response 2A:

Thank you for this comment. We have revised the title to correct minor formatting inconsistencies. Specifically, we standardized capitalization in Non-Autistic–Non-Autistic and added a hyphen in Mixed-Neurotype for consistency. The new title is “ Turn-Taking Transitions in Conversations Among Autistic–Autistic, Non-Autistic–Non-Autistic, and Mixed-Neurotype Adult Pairs”. Please see lines 1-4 in the revised version.

Comment 2B:

Lines 44 to 48

The introduction identifies the relevance of the Double Empathy Theory. However, the research gap could be articulated more explicitly. While you state that ‘relatively few studies have examined turn-exchange within the context of double-empathy frameworks,’ it would strengthen the argument to more clearly specify what exactly is missing.

Response 2B:

We have revised this part to be clearer. Specifically, we now state what is missing in prior research, including the lack of studies examining specific turn-taking features across different neurotype pairings and within the Double Empathy framework. We added, “In particular, there is a lack of research examining how specific temporal features of turn-exchange (e.g., overlaps and duration of turn-exchange) differ across different neurotype pairings, and how these patterns may reflect mutual adaptation or misalignment in interaction.” Please see lines 46-49 in the revised version.

Comment 2C:

Lines 57 to 59
The statement that neurotype mismatches may influence turn-taking dynamics is important, but it is presented somewhat briefly. Consider expanding this section to more explicitly connect the Double Empathy Theory to specific measurable variables in your study.

Response 2C:

Thank you for this helpful suggestion. We have revised this section to more explicitly link the Double Empathy Theory to our measurable variables, to clarify how differences in interactional understanding are reflected in turn-taking patterns (please see lines 62-66 in the revised version).

Comment 2D:

Lines 86 to 89
The research gap is stated clearly in this part. (no research study that compares all three pair conditions…), but this key point appears relatively late in the introduction.

Response 2D:

Thank you for this comment. While this point appears toward the end of the Introduction, we believe it is important, based on the overall structure of the paper, to first review the relevant literature and then explicitly state the research gap and the present study. In the revised version, we have briefly introduced this point earlier in the section added in response to Comment 2B. However, we believe the current placement is appropriate for clearly highlighting the gap after sufficient context has been established. In addition, the “Present Study” section allows readers who wish to identify the research gap more quickly to easily find it.

Comment 2E:

Lines 81 to 84
The transition beginning with ‘In contrast’ is unclear. It is not immediately evident what this section is contrasting with from the prior studies in the previous sentences. Clarify the contrast and ensure consistency in how studies are grouped.

Response 2E:

Thank you for this comment. We agree that the contrast introduced by “In contrast” was not sufficiently clear. In the revised manuscript, we have clarified what this section: "In contrast to studies that compared matched dyads (same neurotype) with mixed dyads (different neurotypes), a study examining only matched dyads (same neurotype) reported that both autistic–autistic and non-autistic–non-autistic pairs preferred short-gap transitions (around 200 milliseconds) in turn-taking." Please see lines 87-88 in the revised version.

Comment 2F:

Lines 123 to 130
The link between turn-taking dynamics and task performance is relevant, but the transition from conversational variables to performance outcomes feels somewhat abrupt. Consider adding a brief bridging sentence to clarify why these outcomes are conceptually connected within your framework.

Response 2F:

We have added that smooth turn-taking may reflect engagement and alignment between speakers at a micro-level, which may support more successful communication at a broader, task-oriented level. This addition helps make explicit how conversational timing relates to accuracy, efficiency, and rapport. Please see lines 139-141 in the revised version.

Comment 2G:

Line 61
Use consistent terminology (e.g., neurotypical vs non-autistic). The manuscript currently uses both terms interchangeably, which may reduce clarity.

Response 2G:

Thank you for this helpful comment. We have revised the manuscript throughout to use “non-autistic” instead of “neurotypical” to ensure consistent terminology across the paper.

Comment 2H:

Ensure consistent formatting (e.g., removal of line-break artifacts such as autistic–au-tistic).

Response 2H:

Thank you for this comment. We have removed all line-break artifacts.

Comment 3A:

Methods

Lines 139 to 142
The inclusion of both formally diagnosed and self-diagnosed autistic participants is understandable. However, additional details are needed regarding how self-diagnosis was operationalized.

Response 3A:

We have added additional details including “All participants self-reported being autistic and then answered a follow-up question indicating whether they had been formally diagnosed or self-identified. All participants also completed the Comprehensive Autistic Trait Inventory (CATI). There was no significant difference in CATI scores between formally diagnosed and self-identified autistic individuals”. Please see lines 164-168 in the revised version.

Comment 3B:

Lines 158 to 159
There appears to be some inconsistency or potential confusion regarding participant counts. The total number of participants is reported as 32 autistic and 36 non-autistic individuals, but later the dyad composition is described differently. Please clarify how these totals map onto the dyad structure and whether any participants were excluded.

Response 3B:

In the Participants section, we specifically mention that we have 32 autistic and 36 non-autistic adults (sum = 68). Later, in the Recording Procedure section, we state that we have 10 autistic–autistic dyads, 11 non-autistic–non-autistic dyads, and 13 mixed dyads. So, in total, 20 + 22 + 26 = 68, which is consistent.

Comment 3C:

Lines 152 to 153
Participants were not informed of their dyad type. While this is appropriate, it would be helpful to briefly explain the rationale.

Response 3C:

We have now clarified this point in the manuscript. We added that participants were not informed of their dyad type to prevent potential changes in behavior based on knowledge of their partner’s neurotype and to maintain more natural interaction patterns (please see lines 178-180 in the revised version).

Comment 3D:

Lines 169 to 177
The tangram task is appropriate and well-described. However, consider briefly justifying why this task is suitable for examining turn-taking dynamics. For example, clarify how it elicits interactional variability relevant to your variables.

Response 3D:

Thank you for this helpful suggestion. While we already clarified the rationale for using the tangram task by saying, "The tangram description task was chosen because participants produced rich language, including communicating collaboratively, developing shared strategies, and creating shared understandings to perform accurately. " But we also have added that "the task requires continuous back-and-forth coordination between participants, thereby naturally eliciting turn-taking dynamics relevant to our measures." Please see lines 199-201 in the revised version.

Comment 3E:

Lines 187 to 192
The criteria for utterance segmentation are clearly described. However, the use of coder judgment for the 200–300 ms range introduces subjectivity. Please clarify.

Response 3E:

We edited this passage to clarify that perceptual criteria were introduced to allow for contextual factors such as speech rate.  We believe that such factors account for ambiguities in the literature on physical durations of pauses or hesitations in connected casual speech (Please see lines 224-225 in the revised version).

 Comment 3F:

Lines 195 to 212
The coding categories are comprehensive while several categories appear conceptually overlapping. Please clarify how these categories are distinguished in practice to ensure they are mutually exclusive and reliably coded.

Response 3F:

We appreciate this comment. We have clarified in the revised manuscript that, although conversational categories were defined to be mutually exclusive and coders followed step-by-step decision rules, some categories rely on perceptual criteria and subtle distinctions. We have also highlighted the interrater reliability results, which were affected by overlapping between codes for these perceptually defined categories (please see lines 254-258 in the revised version).

Comment 3G:

Lines 226 to 228
Statistical tests are appropriate, but additional detail would improve clarity.

Which specific models were used for each research question and any corrections for multiple comparisons if there was any.

Response 3G:

We have added the software used and clarified which test was used for each research question (please see lines 264-268 in the revised version).

Comment 3H:

Lines 230 to 231
Data were aggregated at the dyad level. Please justify this decision, particularly given that turn-taking events are nested within dyads.

Response 3H:

We have clarified that aggregation at the dyad level was theoretically motivated by the Double Empathy Theory, which emphasizes interaction as a relational process between partners rather than an individual-level phenomenon (please see lines 269-271 in the revised version).

Comment 3I:

Lines 240 to 242
Rapport is defined conceptually, but the method of measurement is unclear. Please clarify.

Response 3I:

We have revised the Methods section (Recording Procedure) to clarify how rapport was measured,  including a description of the self-report questionnaire, the five dimensions assessed (ease, enjoyment, success, friendliness, and awkwardness; please see lines 206-210 in the revised version).

Comment 4:

Results: Overall, the results are clearly structured and aligned with the research questions. But several areas could be improved to enhance clarity.

Response 4:

Thank you. We have tried to address all comments and make the manuscript clearer in this section. Please see the respond for each section below.

Comment 4A:

Lines 249 to 253
The chi-square result is reported appropriately, but the effect size is very small (Cramér’s V = .044). Please consider briefly acknowledging the practical significance (or lack thereof) of this effect.

Response 4A:

We have now added a brief clarification in the manuscript to acknowledge that this effect reflects a weak association and should be interpreted with caution. Please see line 295 in the revised version.

Comment 4B:

Lines 265 to 267
The use of a Gamma GLM with a log link is appropriate for skewed duration data. However, please clarify whether any model diagnostics were assessed.

Response 4B:

We assessed model diagnostics for the Gamma GLM with a log link. Specifically, we examined the overall model fit to ensure that the assumptions of the Gamma distribution (e.g., Positive outcome variable, continuous data, right-skewed data, and an appropriate variance structure) were met. We have now added a brief statement in the manuscript to clarify this. Please see lines 310-312 in the revised version.

Comment 4C:

Lines 309 to 317
The modeling approach for RQ5 is described, but it is somewhat unclear whether separate models were run for each outcome and predictor. Please clarify the modeling structure.

Response 4C:

We agree that the modeling structure requires clarification. Separate general linear models were conducted for each outcome variable (accuracy, efficiency, and rapport). Within each outcome, individual models were run for each conversational event (e.g., laughter, overlap, response, backchannel, etc.), with pair neurotype condition included as a factor. We have revised the manuscript to clarify this structure. Please see lines 355-359 in the revised version.

 Comment 4D:

Lines 318 to 324
The sentence structure is somewhat unclear and difficult to follow (“when we compare mixed neurotype pairs to either…”). Consider revising for clarity and grammatical precision.

Response 4D:

We have revised this paragraph to improve clarity and grammatical precision. Please see line 369 in the revised version.

Comment 5A:

Discussion, limitation/future studies and importance of study

Lines 341 to 346
The interpretation that autistic–autistic pairs are interesting but somewhat speculative. Please consider softening the language or more clearly distinguishing between interpretation and empirical evidence.

Response 5A:

We have revised this section by using more cautious phrasing (e.g., “may suggest,” “could be,” “potentially”) and by explicitly framing these explanations as possible interpretations rather than definitive conclusions. Please see lines 391-395 in the revised version.

Comment 5B:

Lines 350 to 357
The explanation of longer gap durations in mixed dyads is plausible but remains speculative. Consider explicitly acknowledging these as possible interpretations rather than definitive explanations.

Response 5B:

 We explicitly used verbs such as may and could to indicate that these are possible interpretations rather than definitive explanations. We have now also added these qualifiers in additional places where needed to further clarify this point. Please see lines 401-409 in the revised version.

Comment 5C:

Lines 358 to 362
The connection to the Double Empathy Theory is appropriate. This argument could be strengthened by more explicitly explaining how the observed timing differences reflect mutual misalignment rather than individual deficits.

Response 5C:

Thank you for this helpful suggestion. We have revised this section to state that longer gaps in mixed-neurotype pairs may arise from differences in how communicative timing cues are produced and interpreted across neurotypes, leading to reduced mutual alignment in expectations about turn timing, align with Double Empathy Theory. Please see lines 409-413 in the revised version.

Comment 5D:

Lines 381 to 387
The discussion appropriately acknowledges that simultaneous talk may reflect both coordination and technical artifacts. This is a strength. However, given this limitation, consider slightly tempering earlier interpretations that frame simultaneous talk primarily as a marker of coordination.

Response 5D:

Thank you for this helpful comment. We agree that simultaneous talk should not be interpreted solely as a marker of coordination. In response, we have revised the manuscript to frame simultaneous talk more cautiously as a conversational cue that may indicate coordination, rather than a direct measure of it (Please see lines 433-440 in the revised version).

Comment 5E:

Lines 412 to 418
The sentence structure is somewhat unclear and difficult to interpret. Please revise for clarity, particularly the conclusion that overlap reflects shared expectations rather than deficits.

Response 5E:

We have revised this section for clarity and improved sentence structure. Please see lines 469-471 in the revised version.

Comment 5F:

Lines 424 to 426
There appears to be an inconsistency with the Methods regarding demographic matching. Please clarify.

Response 5F:

There is no inconsistency with the Methods section. In the Methods, we explicitly state that participants were paired based on overlapping availability and at least one matching demographic factor (e.g., age group, gender, or ethnicity), rather than matching on all demographic factors simultaneously. We have now revised this section to clarify this point. Please see lines 478-479 in the revised version.

Comment 5G:

Lines 439 to 445
This section could be more concise.

Response 5G:

We have revised this section to improve conciseness and clarity. Please see lines 492-498 in the revised version.

 

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for the thoughtful revision of the manuscript. I have carefully reviewed the revised version and confirm that my previous comments have been appropriately addressed.  

At this stage, I have no further substantive concerns and believe the manuscript is suitable for publication.