PONE-D-20-04033

The role of explicit memory in syntactic persistence: effects of lexical cuing and load on sentence memory and sentence production

PLOS ONE

Dear Mr. Zhang,

I am writing with regard to your manuscript, "The role of explicit memory in syntactic persistence: effects of lexical cuing and load on sentence memory and sentence production" (PONE-D-20-04033). I have received comments from two experts in this area of research, and I have also reviewed your manuscript myself. The reviewers and I find that your experiments are tackling a critical and timely question about the mechanisms that underlie structural priming. These results will be of interest to many in this research community, and as such I would like to see them appear in the literature.

The reviewers have raised a number of issues that will need to be addressed in a revision of your manuscript. The most critical of these is raised by Reviewer 1. The reviewer notes that there seem to be inconsistencies between the way that the analyses are described in the manuscript and the analyses that were actually done. Further, the reviewer has a question about the results of Experiment 2a. To add to the reviewer's comments, you might consider generating estimated means from the model for Experiment 2a to see how well they match the observed means. As always, you should also address the reviewers' other comments in some manner -- either through a revision of the manuscript, or through your response letter. 

-------------------

We would appreciate receiving your revised manuscript by May 22 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,

Mike Kaschak

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. Please upload a copy of Supporting Information Figure S1 Fig 1., S2 Fig 2., S3 Fig 3., S4 Fig 4. which you refer to in your text on page 52.

3. Your ethics statement must appear in the Methods section of your manuscript. If your ethics statement is written in any section besides the Methods, please move it to the Methods section and delete it from any other section. Please also ensure that your ethics statement is included in your manuscript, as the ethics section of your online submission will not be published alongside your manuscript.

[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: This paper investigates the contribution of explicit memory to structural priming. The authors describe four experiments in which they investigate the relationship between structural priming, lexical boost, and the difficulty of a concurrent memory load. In two experiments (one dialogue, one monologue), participants performed a standard structural priming task, describing pictures using one of two genitive alternations. In the other two experiments (again one dialogue, one monologue), participants were explicitly told to repeat the sentence structure of the preceding sentence. In all experiments, in between the prime and target sentence, participants solved an addition problem, which was easy or hard; the difference in difficulty between the two types of math problems was increased for the third and fourth experiment.

A robust structural priming and lexical boost effect was found. The authors also found some evidence of task difficulty – i.e., degree of cognitive load – interacting with degree of syntactic priming and lexical boost, to the effect that subjects were less likely to repeat the prime structure when they cognitive load was more difficult, either because they spent longer engaged in the cognitive load task (exploratory RT analyses) or it was a more difficult math problem.

This is an interesting and well-written paper which addresses a theoretically relevant question about the mechanisms underlying structural priming and its reliance on explicit memory. The research questions are well-grounded in prior research and the experiments seem well-designed. The discussion and interpretation is well-reasoned from the results and interesting as well. However, I think there are issues with the statistics which call into question the main result – that higher cognitive load causes reduced structural priming/recall – and thus I do not think much can be inferred from the results and concomitant conclusions of this paper. I believe a complete re-analysis of the statistics is required, and that may (in my opinion, likely will) completely change the central conclusion of the paper.

Major comments:

-Table 1: This table seems to show that in the Same Head Noun condition, 36.0% of the subjects’ productions were s-genitive and 4.7% were of-genitive, and 15.7% and 5.5%, respectively in the Different Head Noun condition.

(a) Why are these values so low? That means that subjects were producing non-s/of-genitive structures on 60-80% of the trials (depending on condition), correct? What were these other structures?

(b) I can’t reconcile these data with the earlier statements that only 8.6% of responses were “other” and that 84.4% of responses were of-genitive (it looks like from Table 1 that ~5% in each condition were of-genitive).

(c) If subjects are producing a different structure on such a large proportion of the trials, please give some assurance that their behavior is indicative of a larger priming processing effect given how many trials were UNaffected by the prime manipulation.

-The authors write (pg. 20, line 473-476): “we included in the model the by-subject and by-item random intercepts as well as random slopes for all main effects and interactions in the fixed model. The full model (and all the other full models in this paper) converged.” But when I look at the code posted on OSF (thank you!), this does not seem to be the case. There are a ton of models there, but it looks like the full model that is being used in the model comparison ANOVAs is DP0122, which has all random slopes removed (i.e., everything except the intercept). It has the full set of fixed effects, but the random effects are just:

(1 | Subject) + (1 | Item)

Similarly, for all of the sub-models that were run, with the fixed effect in question removed, also have the random effects structure of just (1 | Subject) + (1 | Item).

(a) The authors shouldn’t write that all random slopes were included, as it appears that was not the case for the actual model comparisons. It should be clear as to the model structure that was used for the statistics that are reported.

(b) If the models that included all the random slopes converged, then why is that model not being used for the statistics? Removing random slopes can make a model output anti-conservative and may skew the results.

(c) Note that this point is relevant to all of the experiments, as the code that was posted said it was used for all four experiments.

-Likely related to the previous point, I find the results for Experiment 2a to be rather surprising, given the figures and data. I’m not sure what the error bars represent, but assuming they’re standard error, I find it quite unlikely that those would produce an interaction between Prime Condition x Difficulty or a main effect of Difficulty, and yet those are robustly significant.

When I run the code using the model with no random slopes (DP0122), there is indeed a significant Prime Condition x Difficulty interaction. However, I could not get the full model to converge when I ran it myself (the maximal model with all random slopes + intercepts which the text said had converged). So to try to approximate that model, I ran an F1 ANOVA, which, given that there was a balanced design and minimal data loss (2.2%), and I wouldn’t expect big item differences in your stimuli, the ANOVA should show pretty similar results to the LMER. Some results were very similar, as expected (Prime Condition, Prime Condition x Head Noun Condition) but all of the effects involving Task Difficulty were extremely different, e.g. the main effect of Task Difficulty in the ANOVA was F < 1, p = .92; Prime Condition x Task Difficulty was F = 2.5, p = .12; Prime Condition x Head Noun Condition x Task Difficulty was F < 1, p = .75, and these results seem much more in line with the descriptive data. I am not suggesting that the authors run ANOVAs instead of LMERs; there are many good reasons to run LMERs for this type of data. However, it’s important to make sure the results make sense given what the raw data show. As is noted in the next paragraph: “Unexpectedly, we also found a significant main effect of task difficulty (χ2 = 4.966, df = 1, p = .025) and a significant two-way interaction between task difficulty and lexical overlap (χ2 = 7.213, df = 1, p = .007). There was no clear theoretical reason for these effects, and similar effects were not found in any other experiments. We decided to refrain from speculation about these unexpected findings, which might well be a result of type I error.” Given that Task Difficulty shows a tiny effect in the numerical data, but yet shows a significant (or marginal) result for all of the fixed effects that it's involved in, plus none of them were significant in the sample ANOVA that I ran, and you have the minimal random effects structure in your models, and you even note that some of the Task Difficulty effects were unexpected and might be type I error – I think there is something wrong in the LMERs as conducted. It just doesn’t pass the eyeball test. As I cannot get the maximal models to converge on my machine, I can’t see what result that returns myself, so the advice I have is two-fold:

(a) look into the analyses and really make sure they’re being done correctly and that the statistical results match common-sense intuition based on the numerical data

(b) run the model-comparison stats on the maximal model that was reported to have converged, and see if the effects still hold as currently reported.

-pg. 38, line 924-925 says: “The non-significant interaction between cognitive load and lexical overlap on structural priming was somewhat surprising…” and then discusses the implications of this non-effect. I assume that is referring to the interaction between Task Difficulty x Head Noun Condition in Experiment 2a. However, when reporting the Exp 2a results, the text says (line 740): “Unexpectedly, we also found … a significant two-way interaction between task difficulty and lexical overlap (χ2 = 7.213, df = 1, p = .007).” Please resolve this conflict or clarify which comparison you are referring to in the GD.

Minor comments:

-pg. 25, line 593: The exploratory analysis of Exp 1b investigating processing time (I assume also by using RT, as in Exp 1a) x to-be-recalled structure has df = 0.42. I do not understand how doing a model-comparison test can produce a non-integer degrees of freedom. (Unfortunately I can’t look at the model output to help me understand as I can’t find it in the submission.)

-There are a few times throughout the paper when results between experiments are compared numerically, but not statistically (e.g., pg. 26, comparing the size of the lexical boost effect in Exps 1a vs 1b, etc.). Please make those comparisons statistically explicit and report the significance/non-significance of the cross-experiment comparison.

-I wonder how difficult even the most difficult math problems were. Is carrying in addition really such a challenging task? Subjects are only spending 3 seconds doing the math problem in the hardest case. Please talk a little about why you think this might be enough of a cognitive load to disrupt processing.

-In only Exp 2b is there the expected context effect, whereby production of the less-common structure increased over the course of the experiment; in the other experiments, this effect goes in the opposite direction (the less-common structure decreases over the experiment). This is surprising, given that it’s the opposite direction as would be expected. The authors write that the increase effect (the expected direction) is “possibly indicating a practice effect.” Please discuss this further. Why would there be a practice effect in Exp 2b but not the other experiments with (approximately) the same task? Why would the other experiments show the opposite direction as previously-attested context effects?

Textual/editing comments:

-The references are not all standardized – some use the journal abbreviation (J Mem Lang) and others the full name (Journal of experimental psychology: Learning, memory, and cognition).

-pg. 14: references example stimuli sentences before the sentences are produced in the text. This makes that entire paragraph difficult to follow; I suggest moving the sample sentences before the paragraph describing the stimuli conditions (i.e., at line 330).

-pg. 17-18, lines 419-427: There’s a strange paragraph, maybe it’s supposed to be the figure caption for Figure 2? It’s presented in the middle of the text (with no accompanying figure, which is only at the end of the text body).

-pg. 21, line 501 (same for Exp 1b): The authors write that the model for the supplementary analysis investigating RT of Exp 1a is in the Supplementary Material, but it’s not in Appendix A, B, or C (which are the only attached materials), so please include an explicit pointer to where I can find these results.

-Please note in the figure captions what the error bars are: standard error, standard deviation, confidence interval, etc.

-Please make the y-axis for the two figures on the same scale. It makes it easier to compare the results across experiments.

-Grammar/wording comments:

-pg. 5, line 102-3: “It has been several decades for the debate about how such structural priming effect comes about.” -> “It has been several decades since the debate began about…” (for example)

-pg. 7, line 164: “tried to answer A similar question”

-pg. 9, line 224: “These findings are coherent with…” consistent with?

-pg. 13, line 310, 318: “figurines” -> figures

-pg. 27, line 651: “we predicted that the secondary task posited a cognitive load” -> imposed a cognitive load?

-pg. 30, line 726: “minuses” -> minus

-pg. 35, line 856: “more often correctly” -> delete “correctly” perhaps?

-pg. 37, line 881: intertwinement -> entwining

-pg. 38, line 915-956: “similar as for the recall of sentence structure” -> “similar TO” ?

-pg. 39, line 932: “whereas the memory of sentence structure is prune to the load manipulation.” -> I am not sure what this sentence intends.

Reviewer #2: • I appreciate the availability of the authors’ data.

• Overall, I agree with the authors that it is interesting to further explore cognitive load factors that affect priming. I believe the manuscript presents a nice beginning to this. I am, however, not convinced by Experiment 1a in this manuscript. The authors employ a confederate, which I appreciate. I think it is prudent to consider social factors. Perhaps some participants were high self-monitors and simply took more time answering questions to not be judged poorly by their experimental partner. If this were a factor, it may be that they were paying less attention to their partner’s descriptions in anticipation of their turns.

• Regarding the second series of experiments – I found it interesting that the authors decided to make two major changes – 1) the difficulty of the problems and 2) the dismissal of the confederate. I agree that the arithmetic problems, especially the “easy” ones, likely presented minimal, if any cognitive load on the participants. However, making two changes at once leads to questionable conclusions – how can you be sure which manipulation led to the change in significance?

• I think this is a useful addition to the literature. However, I would recommend running a third study with a confederate and the more difficult problems. It would round off these series of experiments nicely and address the concerns I (and the authors even noted) have about the manuscript.

**********

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: No

Reviewer #2: No

[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.

PONE-D-20-04033R1

The role of explicit memory in syntactic persistence: effects of lexical cueing and load on sentence memory and sentence production

PLOS ONE

Dear Dr. Zhang,

I am writing with regard to your manuscript, "The role of explicit memory in syntactic persistence: effects of lexical cueing and load on sentence memory and sentence production" (PONE-D-20-04033R1). I have received a set of comments from one of the original reviewers of your manuscript, and I have also reviewed your revision. The reviewer and I had very similar responses to your work:  you are presenting a nice set of experiments that contribute to our understanding of structural priming, but there are still issues with your statistical analyses and how they are presented. 

The reviewer raises three specific points about the analyses. 

1 -- Your procedure for dropping random effects from your models is not fully specified. For example, the reviewer notes that your strategy is to drop the most complex random effects (e.g., start with 3-way interaction, then the 2-way interactions, and so on), but it is not clear how you decide which effect to drop when there are multiple effects at the same level of complexity (e.g., multiple 2-way interactions). 

2 -- There appears to be an inconsistency between your stated strategy for simplifying your models (drop random correlations, then drop random slopes) and what appears in your code and text.

3 -- There also appears to be an inconsistency between the analyses specified in your code, the analyses described in the text, and the analyses presented in your Appendix. The confusion arises because you both a) present code for an analysis that includes all of the expected fixed effects, and then present the results of the analysis in the Appendix (i.e., a "full" analysis of your design), and b) present code suggesting that you are testing for specific effects using a model-comparison procedure (compare a model that has the target effect, and a model that does not) and generally focus on these specific effects in the text (as evidenced by the presentation of the X^2 statistic in the text, rather than the Z from the model results in the Appendix). The results in the text and in the Appendix are in accord, but it is unclear why you chose this particular way to handle the effects -- it would seem more reasonable to either stick with the results in the Appendix, or do the model comparison approach for all of your predictors (and present the X^2 values in your tables). I also think it would be helpful to have the full models reported in the text (i.e., put the tables in the text, rather than in the Appendix). 

Beyond these main points, the reviewer also provides a set of minor comments for you to consider and address.  

Please submit your revised manuscript by Aug 22 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're 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.

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). You should upload this letter as a separate file labeled 'Response to Reviewers'.
• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

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

We look forward to receiving your revised manuscript.

Kind regards,

Mike Kaschak

Academic Editor

PLOS ONE

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: (No Response)

**********

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: Partly

**********

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

Reviewer #1: No

**********

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 paper is still an interesting topic and contribution to the literature. However, there remain several inconsistencies or errors with the analyses which have the potential to affect the results. I feel the authors need to fix the analyses before I can have confidence that the results are in fact as they have been presented (and thus of course whether the discussion/interpretation follows).

MAJOR COMMENTS:

-There still is confusion about the running and reporting of the lmer models. Due to the discrepancies between what is reported in the paper for the overall analysis strategy, what is reported for each individual experiment, and the R code, in addition to the fact that the reported results often seem at odds with the means and sds shown in the tables/figures (as even the authors themselves note at times), I believe the authors need to very carefully check their analyses, and this must be resolved before I can have confidence in them. I detail several (potentially related) concerns below.

The authors write (pg. 22, line 112) "If the maximal model could not converge or showed singularity, we first dropped the random correlation terms, and then dropped one random factor at a time, starting from the interaction terms, until the model converged and no warning of singular fit was reported." They write something similar in their Response letter.

a) How did you decide which order to drop terms if a model did not converge? A common strategy is to iteratively drop the random effect which accounts for the least amount of variance, but it does not sound like this was what was done. The authors say their strategy was to start by dropping the most complex interaction term, but how did you decide the order of dropping terms at the same complexity level (i.e., the order of dropping among the 3 2-way interactions, or among the 3 main effects)? It looks like terms were dropped based on the order that they were written in the model starting from the bottom-up; in order words, arbitrarily within a complexity level. That is not a good strategy, as dropping random terms can affect significance and thus should not be done arbitrarily. Dropping a random effect which accounts for a substantial amount of variance can have a big effect on the observed significance of fixed effects, and while this is necessary sometimes for convergence/singularity reasons, it needs to be done systematically.

b) The authors write that they first dropped random correlations, and then started dropping random slopes terms. However, most of the models (the main model for Exp 1b, 2a, and 2b) - both in the R code and as written in the paper text - contradict this stated strategy: Those models RETAIN (e.g.) the by-subjects random correlations, but DROP almost all of the by-subjects random effects. (But Exp 1a follows this stated strategy, so they differ between each other.) That is, this model did NOT first drop the by-subject random correlations and then start dropping random effects. It’s really important to both (1) have the description in the paper body match your actual model, and (2) to have a systematic and consistent strategy for pruning a model’s terms, as this can have a major effect on the model’s output.

c) In both the R code and the text body, only some of the factors appear to be tested. In both Exp 1a and 1b, there are models testing only these fixed effects:

-Prime Structure

-Trial Number

-Prime Structure x Head Noun

-Prime Structure x Task Difficulty

-Prime Structure x Head Noun x Task Difficulty

Where are the models testing the other fixed effects? Did you test, for example, for the main effect of Head Noun or Task Difficulty? If you didn’t test all factors, you should say that explicitly in the paper and justify why not. Even if some factors are not theoretically interesting, it is very non-standard to not even test for/report those factors if they were entered into your analysis.

The cross-experiment analysis of 1a vs. 1b similarly omits testing of a number of factors.

Strangely, in Exp 2a, the authors suddenly do test for the main effect of Problem Difficulty (but still not Head Noun), and Problem Difficulty x Lexical Overlap [sic].

But then testing for those two fixed effects disappears again in Exp 2b, which only tests for the 5 fixed effects I noted above.

So again we have a consistency problem.

MINOR COMMENTS:

-Table 1: I understand, I had misunderstood what the numbers in Table 1 represented. The authors respond that they changed the title of the table to clarify this; I admit the title looks identical to the previous one except that one sentence was moved from the top to the bottom of the table, but I leave this to the authors to figure out.

-There are some drastic changes between different versions of lme4, so I suggest you report the version of the package that you used.

-pg 29, line 292-293: The authors write "structural priming effect may persist over at least one intervening task, irrespective of what type of interference is employed." I do not think this is a fair statement; the authors have tested one type of secondary task (addition math problems); they don’t have evidence how any other type of task will or will not interfere.

-The authors reference a condition called "Lexical Overlap" and "Overlap Condition" (pg. 36, line 439 and continuing on that page). Is this supposed to be “Head Noun Condition” or is this a new contrast they have introduced?

-The authors write (pg. 29, lines 281-283) that the "Inverse of Bayes factor" is 0.023 or 0.020. Do they really mean “inverse”? If that’s accurate, that means the Bayes factor is around 50 (1/0.020), which is very strong evidence in favor of H1 (cf Jeffreys 1961).

-The authors provide a short response to my previous question about whether the math problems provide enough of a cognitive load to disrupt processing in their Response letter. However, I didn’t see this in the manuscript itself. While I appreciate the personalized response, it would be useful if other readers could see this discussion as well.

MINOR WORDING EDITS:

-pg. 29, line 291-292: “suggest that THE structural priming effect may persist”

-pg. 109, line 326-327: “One unexpected result in the priming experiments was a negative correlation between the critical trial number and the likelihood of s-genitive production.” --> “priming experiment” [singular], as you only found this negative correlation in Exp 1a, not 1b.

**********

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

[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.]

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 PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PONE-D-20-04033R2

The role of explicit memory in syntactic persistence: effects of lexical cueing and load on sentence memory and sentence production

PLOS ONE

Dear Dr. Zhang,

I am writing with regard to your manuscript, "The role of explicit memory in syntactic persistence: effects of lexical cueing and load on sentence memory and sentence production" (PONE-D-20-04033R2). I have received comments the reviewer who handled your last submission of the manuscript, and I have reviewed your manuscript myself. The reviewer and I find that you have done a good job of handling the majority of the concerns raised in the last round of reviews, leaving only a few relatively straightforward concerns to address. 

The main point to address comes from the reviewer, who notes that there appears to be an inconsistency between the sign of the effects listed in your tables and your stated procedure for defining a reference level for the comparisons. Please double check this to make sure that the results are presented correctly. Beyond this, there are a few other smaller concerns to address. 

Please submit your revised manuscript by Nov 06 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're 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.

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). You should upload this letter as a separate file labeled 'Response to Reviewers'.
• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

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

We look forward to receiving your revised manuscript.

Kind regards,

Mike Kaschak

Academic Editor

PLOS ONE

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

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: (No Response)

**********

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: I Don't Know

**********

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: The authors have largely addressed my statistical comments on the previous version of the manuscript. There remain a few concerns but most of them are minor.

STATISTICS/ANALYSIS COMMENTS:

1. Table 2 lists the fixed effects estimates for Exp 1a with the caption, “Prime condition (S-genitive as the baseline level), head noun condition (Same Head Noun as the baseline level), problem difficulty (Easy Problem as the baseline level) were in mean-centered form.” But the estimates and z-scores listed in the table all seem to have the wrong sign. If S-genitive is coded as the baseline (reference) level, then the estimate shows the change from S-genitive to of-genitive prime. Subjects produced more S-genitives following S-genitive primes than of-genitive primes (see Table 1); that means the change from S-gen prime to of-gen prime is negative. But the estimate listed in Table 2 for Prime Condition is positive. There is a similar problem for Head Noun Condition – more S-genitives were produced in the Same than the Different Head Noun Condition, and Same was coded as the baseline level. That means the estimate for Head Noun Condition should be negative, but it is positive. This backwards-sign problem appears to be the case for all of the models reported in the paper (the four experiments plus the cross-experiment analyses). (Though perhaps not for the effect of critical trial number, which is continuous rather than categorical.)

This speaks to some underlying error that I’m having trouble diagnosing. I don’t know if the error comes about because of (a) type-os in the text of inverting all of the signs, (b) miscoding your factors in the code and analysis, or (c) me misunderstanding what you did (in which case it should be clearer because I am trying really hard to figure it out). But whatever is causing this problem, the output numbers don’t make sense.

2. The authors write, “Based on the standard interpretation of Bayes factors as evidence for null hypotheses [65], BF01 that ranges from 1 to 3 can be taken as weak evidence for the null hypothesis. The higher a BF01, the more evidence in support of the null hypothesis (3-20: positive evidence; 20-150: strong evidence; > 150: very strong evidence).”

I’ve always seen Bayes Factor for experimental hypothesis testing written as strength of evidence in favor of the Alternative (H1), rather than evidence in favor of the Null (H0). For example, as noted in [Kass, R.E. and Raftery, A. (1995) Bayes Factors, Journal of the American Statistical Association, 90: 773-795.], “When comparing results with standard likelihood ratio tests, it is convenient to instead put the null hypothesis in the denominator and thus use B10 as the Bayes factor.” Additionally, Jeffreys (1961), where Bayes Factor was originally proposed, lists BF values as strength of evidence for H1, so numbers >1 show evidence in favor of H1 and numbers <1 show evidence in favor of H0. Anyway, you do write explicitly that you’re using BF01 so I suppose that’s alright, but I suspect it will be confusing to readers who see a decimal number and yet the interpretation is that the evidence favors H1 (as I was on the previous version of the manuscript).

3. There is no report of all of the raw condition means in the paper, which makes interpreting some of your results difficult or impossible. For example, in Exp 2a, there is a main effect of Problem Difficulty, but I don’t know how that is actually manifest in the data. The text lists the p-value but no condition means. Table 5 lists the estimate which ostensibly tells me the direction and magnitude of the effect in standardized terms, (though see earlier point regarding sign) but no raw values. Table 1 lists raw values but only for Prime Condition x Head Noun Condition, collapsing across Problem Difficulty. The figures only show Head Noun Condition x Problem Difficulty because the DV is the Prime Condition difference score. So there’s a main effect of Problem Difficulty and I have no idea what that actually looks like in terms of raw condition means.

MINOR COMMENTS:

4. The code to normalize the key press RT and trial number variables is missing from the model selection code so it doesn’t run as-is. (You can leave it if you want but just so you know that I had to modify your code to run it.)

5. There are quite a number of type-os, extra or missing words, and grammatical errors throughout the text. I list a few here but this is non-exhaustive (just the ones I happened to notice) and I recommend careful proof-reading.

-pg. 22: “This is because that by using contrastive coding, the fixed effects of the model are informative about the main effects and the interactions” --> remove “that”

-pg. 29: “Similar to Experiment 1a, there was also a main effect of the head noun condition (pz

630 < .001), which might also BE due to that the head noun overlap effect…” --> add “be”, remove “that”

-pg. 32: “In addition, the theoretically interestING interactions that involved the contrast between the two experiments were examined by estimating Bayes factors (BF01) using Bayesian Information Criteria.” --> add “ing”

**********

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

[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.]

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 PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

The role of explicit memory in syntactic persistence: effects of lexical cueing and load on sentence memory and sentence production

PONE-D-20-04033R3

Dear Dr. Zhang,

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

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. 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 help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- 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.

Kind regards,

Michael Kaschak

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments: