Physiological synchrony predicts observational threat learning in humans

Understanding how information about threats in the environment is shared and transmitted between individuals is crucial for explaining adaptive, survival-related behaviour in humans and other animals, and for developing treatments for phobias and other anxiety disorders. Research across species has shown that observing a conspecific’s, a ‘demonstrator’s,’ threat responses causes strong and persistent threat memories in the ‘observer’. Here, we examined if physiological synchrony between demonstrator and observer can serve to predict the strength of observationally acquired conditioned responses. We measured synchrony between demonstrators’ and observers’ phasic electrodermal signals during learning, which directly reflects autonomic nervous system activity. Prior interpersonal synchrony predicted the strength of the observer’s later skin conductance responses to threat predicting stimuli, in the absence of the demonstrator. Dynamic coupling between an observer’s and a demonstrator’s autonomic nervous system activity may reflect experience sharing processes facilitating the formation of observational threat associations.

General Interest -The paper should be of general interest, as physiological synchrony is a construct being research in psychology, kinesthesiolopgy, and biomedical engineering and some researchers are examining its clinical applications.
Quality of the paper -More evidence needs to be gathered before physiological synchrony can be considered a "biomarker" of "shared experience." Following the classification of psychophysiological relations outlined by Cacioppo and Tassinary (1990), physiological synchrony could only be called a biomarker of shared experience if (a) it were not related to other psychological constructs (i.e., physiological synchrony has a 1:1 relationship with shared experience) and (b) the relationship between physiological synchrony and shared experience is context-dependent (i.e., it is only expected in some contexts). Without arguing about the second requirement for the term "biomarker," the first requirement is not met: We know that physiological synchrony is related to many other psychological constructs (e.g., empathic accuracy, Levenson & Reuf, 1992;stress contagion, Waters, West, & Mendes, 2014;sociometric status, Kaplan, Burch, Bloom, & Edelberg, 1963). Covarying three potential confounding variables, as the authors did to establish specificity here, is not sufficient to establish specificity. So, it is a many-to-one relationship (i.e., viewing physiological synchrony as a physiological "outcome" of experience sharing; Cacioppo & Tassinary, 1990). This manuscript can make an equal contribution while accurately describing the relationship between physiological synchrony and shared experience.
Paper length -Some of the details provided on processing in the main manuscript lacked sufficient details for independent evaluation (see my first comment in the statistical section), so there was no real value to having them in the main manuscript; might as well move them to the supplement and provide complete details there.
Statistical Analyses -It would be ideal for the reader to have all the numeric information that the authors used as evidence to make any decision during data processing and analysis. For example, we learn that "SCRs were square-root transformed prior to analysis." (p. 8), which would make sense if the data were positively skewed, which the authors had to diagnose in some way. Please provide the skewness statistic or whatever evidence was used to determine that SCRs needed to be squareroot transformed. As another example, could you please provide a citation to support the specific low-and high-pass filters that you used or simply state what signals are being filtered out of the SCR waveform? An example of when these choices are supported is the citation to Haaker et al. (2017) to support the peak to peak amplitude metric of SCR. Similar to this last example, please provide the evidence (i.e., citation or statistical results) that supports all data processing and analysis decisions in the supplemental methods.
-It is important to use accepted terms when making inferences based on Bayes Factors. According to Raftery (1995), a Bayes Factor of BF_10 would not be showing "positive" evidence for the null until it were at least as low as 0.33 and not "strong" evidence for the null (i.e., no difference) until it were .05 or lower (Jarosz & Wiley, 2014). So, the conclusion that the "observers, on average, learned equally well in their first and second blocks as observers" (p. 10) is not aligned with the inference provided from the Bayes Factor. That Bayes Factor of BF_{10} = 1.43 suggests that there is no evidence either way or if anything, that there is anecdotal evidence for there being a difference between the blocks (i.e., the alternative hypothesis). Please cite whatever interpretational rubric that is used for the Bayes Factors at some point and ensure that the inferential language used for all tests sticks closely to the chosen rubric (e.g., in the test that immediately follows the one given as an example on page 10 or on page 14, for the lack of a trial effect for the relationship between synchrony and CS differentiation being described as "strong" when it should just be "positive").
-Given the high correlation among CRQA metrics (i.e., that results in them loading onto a single factor), it would be better to conduct the separate regressions on all four CRQA metrics at once as one multivariate regression. Only the multivariate statistics would need to be reported, and this would address the potential for inflated family-wise Type I error.
-Please map the CRQA metrics onto the psychophysiological constructs they are quantifying. What aspect of physiological synchrony is captured by DET and what aspect of physiology synchrony is captured by LAM, etc.? -Minor comment: Please describe how the continuous regressors were standardized, given the hierarchical nature of the dataset. Were they standardized across all observations, within participant, within dyad, or some multistep approach? Data -The authors did not provide the statistical code that they used to analyze the processed data that is available on the OSF repository. Rather, they state that this code is available by request. Especially given the emphasis on quantification in the physiological synchrony literature, publicly posting the statistical syntax would help readers independently evaluate the results and would allow future researchers to more easily build on this work by using the same quantitative approach. Furthermore, sharing the code seems to be a publication requirement of the Proceedings of the Royal Society B (i.e., on the submission form, it states, "It is a condition of publication that data, code and materials supporting your paper are made publicly available.").
-Sample size was determined using simulations based on observations in an earlier pilot study. Please provide the code on the OSF page or more information about the simulations conducted to calculate power.
-Also, please provide a Data Dictionary or Variable Coder for data_sync.txt Your manuscript has now been peer reviewed and the reviews have been assessed by an Associate Editor. The reviewers' comments (not including confidential comments to the Editor) and the comments from the Associate Editor are included at the end of this email for your reference. As you will see, the reviewers and the Editors have raised some concerns with your manuscript and we would like to invite you to revise your manuscript to address them.
We do not allow multiple rounds of revision so we urge you to make every effort to fully address all of the comments at this stage. If deemed necessary by the Associate Editor, your manuscript will be sent back to one or more of the original reviewers for assessment. If the original reviewers are not available we may invite new reviewers. Please note that we cannot guarantee eventual acceptance of your manuscript at this stage.
To submit your revision please log into http://mc.manuscriptcentral.com/prsb and enter your Author Centre, where you will find your manuscript title listed under "Manuscripts with Decisions." Under "Actions", click on "Create a Revision". Your manuscript number has been appended to denote a revision.
When submitting your revision please upload a file under "Response to Referees" -in the "File Upload" section. This should document, point by point, how you have responded to the reviewers' and Editors' comments, and the adjustments you have made to the manuscript. We require a copy of the manuscript with revisions made since the previous version marked as 'tracked changes' to be included in the 'response to referees' document.
Your main manuscript should be submitted as a text file (doc, txt, rtf or tex), not a PDF. Your figures should be submitted as separate files and not included within the main manuscript file.
When revising your manuscript you should also ensure that it adheres to our editorial policies (https://royalsociety.org/journals/ethics-policies/). You should pay particular attention to the following: Research ethics: If your study contains research on humans please ensure that you detail in the methods section whether you obtained ethical approval from your local research ethics committee and gained informed consent to participate from each of the participants.
Use of animals and field studies: If your study uses animals please include details in the methods section of any approval and licences given to carry out the study and include full details of how animal welfare standards were ensured. Field studies should be conducted in accordance with local legislation; please include details of the appropriate permission and licences that you obtained to carry out the field work.
Data accessibility and data citation: It is a condition of publication that you make available the data and research materials supporting the results in the article. Datasets should be deposited in an appropriate publicly available repository and details of the associated accession number, link or DOI to the datasets must be included in the Data Accessibility section of the article (https://royalsociety.org/journals/ethics-policies/data-sharing-mining/). Reference(s) to datasets should also be included in the reference list of the article with DOIs (where available).
In order to ensure effective and robust dissemination and appropriate credit to authors the dataset(s) used should also be fully cited and listed in the references.
If you wish to submit your data to Dryad (http://datadryad.org/) and have not already done so you can submit your data via this link http://datadryad.org/submit?journalID=RSPB&manu=(Document not available), which will take you to your unique entry in the Dryad repository.
If you have already submitted your data to dryad you can make any necessary revisions to your dataset by following the above link.
For more information please see our open data policy http://royalsocietypublishing.org/datasharing.
Electronic supplementary material: All supplementary materials accompanying an accepted article will be treated as in their final form. They will be published alongside the paper on the journal website and posted on the online figshare repository. Files on figshare will be made available approximately one week before the accompanying article so that the supplementary material can be attributed a unique DOI. Please try to submit all supplementary material as a single file.
Online supplementary material will also carry the title and description provided during submission, so please ensure these are accurate and informative. Note that the Royal Society will not edit or typeset supplementary material and it will be hosted as provided. Please ensure that the supplementary material includes the paper details (authors, title, journal name, article DOI). Your article DOI will be 10.1098/rspb.[paper ID in form xxxx.xxxx e.g. 10.1098/rspb.2016.0049].
Please submit a copy of your revised paper within three weeks. If we do not hear from you within this time your manuscript will be rejected. If you are unable to meet this deadline please let us know as soon as possible, as we may be able to grant a short extension.
Thank you for submitting your manuscript to Proceedings B; we look forward to receiving your revision. If you have any questions at all, please do not hesitate to get in touch.
Best wishes, Dr Sasha Dall mailto: proceedingsb@royalsociety.org Associate Editor Board Member: 1 Comments to Author: We have now obtained two expert reviews of your manuscript, and I am happy to tell you that both the reviewers liked your paper. My own reading of your manuscript corroborates the reviews: this is an exciting piece of work that could become publishable in Proc B. However, although the reviews were generally positive, one of the reviewers raised concerns about the data processing, statistical analysis, and interpretations. This reviewer also requested that you would make the statistical code available in addition to the data. Addressing these points requires a significant revision of the text, but it should lead to an improved manuscript. The comments by the other reviewer are fewer, asking for the better organization of the result section and raising an additional point for discussion. They, too, would likely improve your paper.

Reviewer(s)' Comments to Author:
Referee: 1 Comments to the Author(s) This paper is well-written and the main ideas are appealing and accessible to the non-specialist. However, the details included in the Results section are sometimes presented in a way that is difficult to follow. I would suggest to put them (or, at least, part of them) in a tabular form, where they can be accessed if needed, avoiding cluttering the text. The Discussion section addresses a number of interesting issues. Since it is rather long, it would benefit having some subsections. Moreover, I miss some discussion regarding the possible impact on the conclusion that were drawn regarding the choice of parameters. For example, in pag. 9, it is said that "Optimal parameters for the CRQA analysis (...) were determined individually for each pair of signals...". Is this "tuning" potentially critical? Are there other parameters potentially sensitive to such "tuning"?
Referee: 2 Comments to the Author(s) Scientific Importance There are a number of things to like about this paper. Most notably, the design of the study was nice, as the authors used the concept of synchrony in an interesting way to try to "capture" observational threat learning. In comparison, most research on synchrony seeks to understand more about synchrony itself. Unfortunately, the language spoke beyond the findings. If the authors kept the interpretation of results to observational threat learning and the covariance of physiological synchrony with it, then the paper would be interesting for its data alone. However, the assertions made in this paper frequently go beyond the design, data, or results.
General Interest -The paper should be of general interest, as physiological synchrony is a construct being research in psychology, kinesthesiolopgy, and biomedical engineering and some researchers are examining its clinical applications.
Quality of the paper -More evidence needs to be gathered before physiological synchrony can be considered a "biomarker" of "shared experience." Following the classification of psychophysiological relations outlined by Cacioppo and Tassinary (1990), physiological synchrony could only be called a biomarker of shared experience if (a) it were not related to other psychological constructs (i.e., physiological synchrony has a 1:1 relationship with shared experience) and (b) the relationship between physiological synchrony and shared experience is context-dependent (i.e., it is only expected in some contexts). Without arguing about the second requirement for the term "biomarker," the first requirement is not met: We know that physiological synchrony is related to many other psychological constructs (e.g., empathic accuracy, Levenson & Reuf, 1992;stress contagion, Waters, West, & Mendes, 2014;sociometric status, Kaplan, Burch, Bloom, & Edelberg, 1963). Covarying three potential confounding variables, as the authors did to establish specificity here, is not sufficient to establish specificity. So, it is a many-to-one relationship (i.e., viewing physiological synchrony as a physiological "outcome" of experience sharing; Cacioppo & Tassinary, 1990). This manuscript can make an equal contribution while accurately describing the relationship between physiological synchrony and shared experience.
Paper length -Some of the details provided on processing in the main manuscript lacked sufficient details for independent evaluation (see my first comment in the statistical section), so there was no real value to having them in the main manuscript; might as well move them to the supplement and provide complete details there.
Statistical Analyses -It would be ideal for the reader to have all the numeric information that the authors used as evidence to make any decision during data processing and analysis. For example, we learn that "SCRs were square-root transformed prior to analysis." (p. 8), which would make sense if the data were positively skewed, which the authors had to diagnose in some way. Please provide the skewness statistic or whatever evidence was used to determine that SCRs needed to be squareroot transformed. As another example, could you please provide a citation to support the specific low-and high-pass filters that you used or simply state what signals are being filtered out of the SCR waveform? An example of when these choices are supported is the citation to Haaker et al. (2017) to support the peak to peak amplitude metric of SCR. Similar to this last example, please provide the evidence (i.e., citation or statistical results) that supports all data processing and analysis decisions in the supplemental methods.
-It is important to use accepted terms when making inferences based on Bayes Factors. According to Raftery (1995), a Bayes Factor of BF_10 would not be showing "positive" evidence for the null until it were at least as low as 0.33 and not "strong" evidence for the null (i.e., no difference) until it were .05 or lower (Jarosz & Wiley, 2014). So, the conclusion that the "observers, on average, learned equally well in their first and second blocks as observers" (p. 10) is not aligned with the inference provided from the Bayes Factor. That Bayes Factor of BF_{10} = 1.43 suggests that there is no evidence either way or if anything, that there is anecdotal evidence for there being a difference between the blocks (i.e., the alternative hypothesis). Please cite whatever interpretational rubric that is used for the Bayes Factors at some point and ensure that the inferential language used for all tests sticks closely to the chosen rubric (e.g., in the test that immediately follows the one given as an example on page 10 or on page 14, for the lack of a trial effect for the relationship between synchrony and CS differentiation being described as "strong" when it should just be "positive").
-Given the high correlation among CRQA metrics (i.e., that results in them loading onto a single factor), it would be better to conduct the separate regressions on all four CRQA metrics at once as one multivariate regression. Only the multivariate statistics would need to be reported, and this would address the potential for inflated family-wise Type I error.
-Please map the CRQA metrics onto the psychophysiological constructs they are quantifying. What aspect of physiological synchrony is captured by DET and what aspect of physiology synchrony is captured by LAM, etc.? -Minor comment: Please describe how the continuous regressors were standardized, given the hierarchical nature of the dataset. Were they standardized across all observations, within participant, within dyad, or some multistep approach? Data -The authors did not provide the statistical code that they used to analyze the processed data that is available on the OSF repository. Rather, they state that this code is available by request. Especially given the emphasis on quantification in the physiological synchrony literature, publicly posting the statistical syntax would help readers independently evaluate the results and would allow future researchers to more easily build on this work by using the same quantitative approach. Furthermore, sharing the code seems to be a publication requirement of the Proceedings of the Royal Society B (i.e., on the submission form, it states, "It is a condition of publication that data, code and materials supporting your paper are made publicly available.").
-Sample size was determined using simulations based on observations in an earlier pilot study. Please provide the code on the OSF page or more information about the simulations conducted to calculate power.
-Also, please provide a Data Dictionary or Variable Coder for data_sync.txt It is a condition of publication that authors make their supporting data, code and materials available -either as supplementary material or hosted in an external repository. Please rate, if applicable, the supporting data on the following criteria.

Comments to the Author
The authors have taken into consideration my remarks in this revised version, hence, I recommend acceptance.

24-Apr-2020
Dear Dr Pärnamets I am pleased to inform you that your manuscript entitled "Physiological Synchrony Predicts Observational Threat Learning in Humans" has been accepted for publication in Proceedings B.
You can expect to receive a proof of your article from our Production office in due course, please check your spam filter if you do not receive it. PLEASE NOTE: you will be given the exact page length of your paper which may be different from the estimation from Editorial and you may be asked to reduce your paper if it goes over the 10 page limit.
If you are likely to be away from e-mail contact please let us know. Due to rapid publication and an extremely tight schedule, if comments are not received, we may publish the paper as it stands.
If you have any queries regarding the production of your final article or the publication date please contact procb_proofs@royalsociety.org Open Access You are invited to opt for Open Access, making your freely available to all as soon as it is ready for publication under a CCBY licence. Our article processing charge for Open Access is £1700. Corresponding authors from member institutions (http://royalsocietypublishing.org/site/librarians/allmembers.xhtml) receive a 25% discount to these charges. For more information please visit http://royalsocietypublishing.org/open-access.
Your article has been estimated as being 10 pages long. Our Production Office will be able to confirm the exact length at proof stage.
Paper charges An e-mail request for payment of any related charges will be sent out after proof stage (within approximately 2-6 weeks). The preferred payment method is by credit card; however, other payment options are available Electronic supplementary material: All supplementary materials accompanying an accepted article will be treated as in their final form. They will be published alongside the paper on the journal website and posted on the online figshare repository. Files on figshare will be made available approximately one week before the accompanying article so that the supplementary material can be attributed a unique DOI.
Thank you for your fine contribution. On behalf of the Editors of the Proceedings B, we look forward to your continued contributions to the Journal. Referee #1: Comment: This paper is well-written and the main ideas are appealing and accessible to the non-specialist. However, the details included in the Results section are sometimes presented in a way that is difficult to follow. I would suggest to put them (or, at least, part of them) in a tabular form, where they can be accessed if needed, avoiding cluttering the text.
Response: Thank you for this comment. In an attempt to limit the burden on reader we have moved some details of our results to the Supplementary Results giving only summaries in the main text. Additionally, we now provide regression tables for all our analyses in the Supplementary Results.

Comment:
The Discussion section addresses a number of interesting issues. Since it is rather long, it would benefit having some subsections.

Response:
We have added subsections to the Discussion.
Comment: Moreover, I miss some discussion regarding the possible impact on the conclusion that were drawn regarding the choice of parameters. For example, in pag. 9, it is said that "Optimal parameters for the CRQA analysis (...) were determined individually for each pair of signals...". Is this "tuning" potentially critical? Are there other parameters potentially sensitive to such "tuning"?
Response: Selecting non-optimal sets of parameters would likely threaten the validity of the results. The parameters are tuned maximize mutual information between the two signals and to yield recurrence rates within the specified range: 2-4% which is the recommendation from the technical literature. Due to the nature of recurrence plots, virtually any recurrence rate can be yielded by changing the analysis parameters, which would inflate the occurrence of, for example, longer segments of recurrent points which are Referee #2: Comment: More evidence needs to be gathered before physiological synchrony can be considered a "biomarker" of "shared experience." Following the classification of psychophysiological relations outlined by Cacioppo and Tassinary (1990), physiological synchrony could only be called a biomarker of shared experience if (a) it were not related to other psychological constructs (i.e., physiological synchrony has a 1:1 relationship with shared experience) and (b) the relationship between physiological synchrony and shared experience is context-dependent (i.e., it is only expected in some contexts). Without arguing about the second requirement for the term "biomarker," the first requirement is not met: We know that physiological synchrony is related to many other psychological constructs (e.g., empathic accuracy, Levenson & Reuf, 1992;stress contagion, Waters, West, & Mendes, 2014;sociometric status, Kaplan, Burch, Bloom, & Edelberg, 1963). Covarying three potential confounding variables, as the authors did to establish specificity here, is not sufficient to establish specificity. So, it is a many-to-one relationship (i.e., viewing physiological synchrony as a physiological "outcome" of experience sharing; Cacioppo & Tassinary, 1990). This manuscript can make an equal contribution while accurately describing the relationship between physiological synchrony and shared experience.
Response: We agree with the Reviewer that the claim about "biomarker" was overstated. This claim was found in the abstract of the manuscript, and in the revised manuscript we have removed it.
Additionally, we went through the discussion to ensure that we did not make any similar overstated claims there. Our judgment is that we there provided a more measured account of our findings. Nevertheless, to guard against overinterpretation, we have added to the discussion where we discuss empathic sharing, where we now write: In sum, it possible, although not conclusive, that the experience sharing in our task also reflects empathic sharing of states between observers and demonstrators, although careful experimentation will be required to establish if synchrony during observational threat learning is specific to such processes or not.
Comment: Some of the details provided on processing in the main manuscript lacked sufficient details for independent evaluation (see my first comment in the statistical section), so there was no real value to having them in the main manuscript; might as well move them to the supplement and provide complete details there.
Response: We have addressed the Reviewer's comments below, and according to their suggestion moved most of the sections on physiological analysis and most of the statistical analysis section to the Supplemental Methods.
Comment: It would be ideal for the reader to have all the numeric information that the authors used as evidence to make any decision during data processing and analysis. For example, we learn that "SCRs were squareroot transformed prior to analysis." (p. 8), which would make sense if the data were positively skewed, which the authors had to diagnose in some way. Please provide the skewness statistic or whatever evidence was used to determine that SCRs needed to be square-root transformed. As another example, could you please provide a citation to support the specific low-and high-pass filters that you used or simply state what signals are being filtered out of the SCR waveform? An example of when these choices are supported is the citation to Haaker et al. (2017) to support the peak to peak amplitude metric of SCR. Similar to this last example, please provide the evidence (i.e., citation or statistical results) that supports all data processing and analysis decisions in the supplemental methods.
Response: We apologize that our analytic decisions were not made clear in the original manuscript. Our analysis followed our standard lab protocol (cf. Olsson et al., 2007Olsson et al., , 2016Haaker et al., 2017), which in turn conforms to standards widely adopted in the broader fear conditioning and psychophysiological literature (cf. LaBar et al., 1995;Lykken and Venables, 1971;Boucsein, 2012).
In the revised manuscript we now write: The raw signal from each participant was filtered offline in Ack-Knowledge with a low-pass filter (1Hz) to remove potential recording artefacts and then a high-pass filter (0.05Hz) to recover the phasic skin conductance responses by removing the tonic component of the signal (Boucsein, 2012). Using CS onset and shock delivery as event markers and following established protocols (Haaker et al., 2017), skin conductance responses (SCRs) were measured as the largest peak-to-peak amplitude difference in the phasic skin conductance signal in the 0.5 to 4.5 second window following stimulus onset. Responses below 0.02µS were scored as zero. Scoring was first done using AcqKnowledge's automated scoring algorithm, and then manually checked by an experimenter. SCRs were square-root transformed prior to analysis (LaBar et al., 1995).
Additionally, to ensure that our main findings are robust to transformations of the main outcome variables, we performed two additional analyses. We used a hurdle lognormal model (isntead of a Gaussian) to model the untransformed skin conductance responses. We replicated the first analysis for learning effects including the moderating effects of Role and Block. This is reported as Supplementary Table 2 and reproduced below as Table 1. We also replicated the analysis using the principle components of the CRQA metrics, reported as Supplementary Table 7 and reproduced below. If anything this analysis showed stronger model evidence (by a factor of 70!) for an effect of the first principle component on CS differentiation. We thank the Reviewer for providing us this opportunity to showcase the robustness of our findings.  Comment: It is important to use accepted terms when making inferences based on Bayes Factors. According to Raftery (1995), a Bayes Factor of BF 10 would not be showing "positive" evidence for the null until it were at least as low as 0.33 and not "strong" evidence for the null (i.e., no difference) until it were .05 or lower (Jarosz & Wiley, 2014). So, the conclusion that the "observers, on average, learned equally well in their first and second blocks as observers" (p. 10) is not aligned with the inference provided from the Bayes Factor. That Bayes Factor of BF 10 = 1.43 suggests that there is no evidence either way or if anything, that there is anecdotal evidence for there being a difference between the blocks (i.e., the alternative hypothesis). Please cite whatever interpretational rubric that is used for the Bayes Factors at some point and ensure that the inferential language used for all tests sticks closely to the chosen rubric (e.g., in the test that immediately follows the one given as an example on page 10 or on page 14, for the lack of a trial effect for the relationship between synchrony and CS differentiation being described as "strong" when it should just be "positive").
Response: This is an important point, we have gone through the manuscript to ensure that we do not overstate evidence and to carefully point out where Bayes Factors do not give any evidence (BF: 0.3 -3) against or for an effect. This improves the statistical reporting and we are thankful to the Reviewer for raising this concern.
Finally, we agree with the importance of giving an interpretational rubric, however, we did provide one in the original manuscript. In the Supplemental Methods, in the section concerning analysis we wrote: We rely on Bayes Factors to make inferences about effects. We interpret Bayes Factors above 10 to constitute strong evidence for an effect, and Bayes Factors between 3 and 10 to constitute weak evidence for an effect.
We remain committed to this original rubric, which corresponds to that of Jeffreys (1961), but other readers may of course make other interpretations based on the numerical information given. We believe this is one of the strengths of the Bayesian approach generally.
Comment: Given the high correlation among CRQA metrics (i.e., that results in them loading onto a single factor), it would be better to conduct the separate regressions on all four CRQA metrics at once as one multivariate regression. Only the multivariate statistics would need to be reported, and this would address the potential for inflated family-wise Type I error.
Response: We agree with the Reviewer's sentiment, however, the reason for conducting the principle component regression was precisely to deal with the high correlations between CRQA metrics. It is well known that regressing multiple correlated variables can cause masking. Indeed, following the Reviewer's suggestion 1 we ran a regression model testing the interaction between CS status and all of the CRQA metrics (see Table ). As could be expected, the posterior estimates of all four relationships widen and with them the evidence for an effect becomes anecdotal or indeed negative. We believe this further supports our original analysis choice of using principle components regression to investigate our effects, as is supported by prior literature (Mønster et al., 2016). In the revised manuscript we report  Response: We were puzzled by this comment, since we believe we did produce such a mapping in the original manuscript. In the methods section we gave a technical introduction to all metrics: From each resulting cross-recurrence plot various metrics can be computed that capture the dynamics of the system being analyzed (Marwan et al., 2007;Shockley, 2005;Coco and Dale, 2014). Here we computed four metrics: DETerminism, LAMinarity, maximum line (maxL) and relative Entropy (rENTR). DET represents the relative amount of recurrent points forming diagonal segments, as such DET measures the predictability of the time-series as they evolve over time. LAM is analogous to DET but instead represents recurrent points forming vertical line segments, which can be thought of capturing relative stability in the system. maxL is length of the longest diagonal sequence of recurrent points, capturing the maximal strength of coupling between the two time series. rENTR calculates the Shannon entropy of the histogram of the deterministic (diagonal) sequences and indexes the complexity of the relationship between the time series.
In the discussion we additionally wrote: Synchrony, as measured through CRQA, reflects similarity in the electrodermal activity trajectories of the observer and demonstrator during the learning phase. We analyzed four common used metrics derived using the cross-recurrence plots from each learning phase recorded in our experiment (see Fig 1). These metrics capture salient patterns in how the patterns of similarity between observers and demonstrators evolve. We found particularly strong evidence for determinism (DET) and laminarity (LAM) as predictors of later conditioned responses. Determinism implies a stronger coupling between the trajectories of the two signals, as indicated by a larger proportion of the recurrent time points form diagonal lines in the cross-recurrence plot. Laminarity suggests sustained, smooth periods in the signal's mutual evolution, as indicated by vertical segments in the cross-recurrence plots. Across all our analyses, the more synchronized demonstrators and observers were in their electrodermal activity during the observational learning phase, the stronger the observer's CS differentiation was during the testing phase.
Comment: Minor comment: Please describe how the continuous regressors were standardized, given the hierarchical nature of the dataset. Were they standardized across all observations, within participant, within dyad, or some multistep approach?
Response: All continuous regressors were standardized across all observations, so that 0 always reflects the population average as does the intercept of all the statistical models and, consequently, that our population level estimates ("fixed-effects") are deviations from this average. We now clarify this in the methods section on analysis, in the Supplemental Materials where we state the procedure for standardizing variables.
Comment: The authors did not provide the statistical code that they used to analyze the processed data that is available on the OSF repository. Rather, they state that this code is available by request. Especially given the emphasis on quantification in the physiological synchrony literature, publicly posting the statistical syntax would help readers independently evaluate the results and would allow future researchers to more easily build on this work by using the same quantitative approach. Furthermore, sharing the code seems to be a publication requirement of the Proceedings of the Royal Society B (i.e., on the submission form, it states, "It is a condition of publication that data, code and materials supporting your paper are made publicly available.").
Response: We absolutely agree with the Reviewer here and it was an omission on our part not having done so to begin with. We have now uploaded code to the OSF repository.
Comment: Sample size was determined using simulations based on observations in an earlier pilot study. Please provide the code on the OSF page or more information about the simulations conducted to calculate power.

Response:
We have now shared code on the OSF repository, and provide additional details in the Supplementary Methods as well, where we write: To determine sample size we simulated data. We targeted an effect size of 0.04 (in √ µS) for the target interaction between a CRQA metric and CS status, with a standard deviation of 0.015 for the per participant varying coefficients. Model priors were same as for our subsequent analyses. We analyzed 400 simulated datasets and assessed if BF 10 > 3. Our simulations indicated that 65 dyads would provide 90% power to assess an effect.