Saturday, September 10, 2011

How to respond to an R&R

[Note: Some of the follow suggestions I picked up from my faculty advisors and colleagues while in grad school, others I came up with on my own, and still others came from Wendy Belcher's excellent book: Writing Your Journal Article in 12 Weeks]

So, you submitted a manuscript to a journal, and the editor liked it enough to give you an R&R... now what?

First, congrats! An R&R is about as good as you can expect for a regular journal, since very few manuscripts are accepted 'as is.'

Depending on the editor's letter and reviewers' comments, it might be an easy R&R that won't take much time, or a very hard R&R that will take weeks or months to address. Also, depending on the journal, an R&R might mean the manuscript is basically accepted, pending a little polishing, or alternatively that there's a good chance you'll still get rejected later on in the process. In any case, your goal should be to address as many of the recommended changes as is reasonably possible in the revision to make the paper better, without accidentally 'messing up' your paper (or your main points) in the process. Hopefully this post helps guide you there.

So the first thing you should do is read the editor's letter in its entirely (you probably opened the letter/email and just skimmed the page(s) until you saw 'revise and resubmit'...).  The editor's job is to read the reviews, make a decision on your manuscript (in this case, R&R), and then (in my opinion) give you guidance on what are the most important things to address in the revised manuscript. If the editor didn't do that, or gave you a generic 'please address the reviewers' comments in your revised submission,' then your job will be a little harder than it probably should be.

Anyhow, read the editor's letter and see if s/he gave you specific instructions as to what to address. The editor might note the important of a missing variable in your model, as mentioned by reviewer 2, or that your treatment of the prior literature was lacking in X area, as noted by reviewers 1 and 3. Whatever the editor says in their letter regarding potential changes should be your first priority in the revision. What I like to do is open a new word document and make a bulleted list of the changes the editor specifically mentioned in the letter (these bullets go under an 'Editor' heading). This will become part of your 'potential revision' list/spreadsheet that you will continue building in the future. (More about this in a bit).

Next, read through all of your reviewers' comments, without taking any notes. Make sure you have your submitted manuscript draft in front of you, so you can reference it when a reviewer mentions something on page X or in Table Y. Keep an eye out for comments that come up from more than one reviewer, such as the suggestion to use robust standard errors in your model, to provide a correlation matrix table, or to cite the work of authors A and B.

Chances are, you won't agree with everything the reviewers recommend, and you'll form the following opinions about the reviewers [note: for ease of comparisons, all three reviewers are assumed to be male in this example]:

Reviewer 1: Even my
bowel movements are
high impact.

Reviewer 1: You'll probably think this reviewer is a genius, particularly since he gave you the most useful recommendations, and other than a couple minor suggestions, loved your entire manuscript. His suggestions are relevant, fairly easy to implement, and will really help boost the quality of your manuscript. This reviewer, while rarely seen in the wild, is the best you can hope for.

Reviewer 2: "Derp?"

Reviewer 2: You'll probably think the second reviewer is an idiot, but a fairly harmless one. He probably doesn't really know anything about either your analytical methods or the literature related to your topic, but tried to fake it so the editor didn't get angry. Most of the suggestions he offered aren't very useful and/or make no sense given your model, argument, and findings. Still, you can easily address his main concerns, and easily dismiss his unhelpful and misguided recommendations.

Reviewer 3: My comb-over
is horrible, I can't eat sweets
due to my diabeetus, and I'm
going to take my anger out
on you and your manuscript.

Reviewer 3: Upon reading reviewer 3's comments, your first thought will probably be "what a f*cking d*ckhead." You'll probably think that every suggestion by reviewer 3 is complete garbage, that he misunderstood your entire argument, knows nothing of the methods you used, and decided to use his "anonymous" review as an outlet for his deep-seated inner frustrations concerning sexual impotency and male pattern baldness. You'll likely question his academic and genetic pedigree, and wonder how the editor could have chosen such an arrogant jerk. This reaction is completely normal, and has a long historical precedent.

In fact, Einstein even developed a theory related to reviewer comments that largely supports these three general categories, as previously noted in my post on how to get over manuscript rejections:
More useful than relativity.
However, these categories are not exhaustive, nor are they exclusive. You may, for instance, get a review from a reviewer that combines traits of 2 (idiot) and 3 (jerk).

 Regardless of the types of reviewers you get, or the content of their reviews, the fact that you received an R&R meant the editor saw some promise in your manuscript. Therefore, now that you've surveyed all of the reviews, you need to get to work addressing the specific suggestions and critiques, and revising your manuscript for resubmission.

So the next step is to go back through the reviews and add to an excel spreadsheet (or equivalent) every specific suggestion or critique, from each reviewer, that requires you to change something (or might require you to say why you didn't change something.) Each suggestion/comment should be its own entry on a separate line. I recommend five columns in the spreadsheet: ID # (increments up), Reviewer (who made the comment), Comment (what was the change/critique), Importance (How important does this seem to your overall revised manuscript, or to the editor/reviewer?), and Ease (How long and/or how much work is this going to take to address). I also like to color code the importance and ease cells using the an easy to distinguish color scheme - red for important/hard, yellow for medium/moderate, and green for low/easy. If you set it up like this, your spreadsheet should look something like the following (excluding the earlier comments from the editor):

Once you've finished adding the reviewers' comments to your spreadsheet, add the editor comments you previously listed out if they're not already covered by a reviewer, and mark anything the editor said to do as high importance. If the editor says something a reviewer already said, then add 'editor' to the reviewer cell as a reminder, and mark it as high importance regardless of your original importance rank. If the editor specifically says something is not important to address, then mark it as low importance, and if they say not to do it at all, then change the font style for the associated entry to strikeout.

After everything is down on your spreadsheet, you can get a sense of the overall work that's ahead of you by looking at the importance/ease categories for each entry. A roadmap to these combinations, and what they mean for your revision, is below:

High/Hard: These are the changes that will give you the most headaches. You need to do them (high importance), but they're going to take a long time and generally be a P.I.T.A. (hard to do).  Doing these first will make your life much easier later on, and if they're major changes, some of the less important changes may be obsolete once you do these. If you decide not to do one of these, you need a very good reason why not.

High/Moderate: These are changes that will be a little easier than the High/Hard changes. You still need to do them, and they're going to take you some time, but they won't be absolutely horrible to address.

High/Easy: These are the changes that will make you feel most productive, since you can address these important 'need to do' changes very quickly and without much work.

Medium/Moderate: You should probably do these changes, especially if you have the time.  If you don't do one of these change requests, then you should definitely address why you didn't in your cover letter (more about the cover letter in a minute).

Medium/Easy or Low/Easy: Since these are easy to do, you might as well do them, especially for the ones of medium importance. If you choose not to do these, you should at least acknowledge why you're not doing the medium importance ones, if not the low importance ones as well (since not doing the latter will make you look lazy).

Low/Moderate or Low/High: These might be changes you decide not to do, given their relatively low importance compared to the moderate-to-high amount of work they'd require of you. Again, you may want to address why you didn't do these in your cover letter, but given their low importance, if you miss discussing some of these it probably won't be the end of the world.

Ok, so you have your potential changes all in order, you've ranked their importance and ease, you know which ones you need to address, which you should address, which you can easily address, and which you're not wasting your time with. Now comes the hard part: Make these changes in the manuscript. That's about as specific as I can get for this step, but this is what will take the longest.

Once you're done making (or not making) the various changes, you need to reread your manuscript from start to finish to make sure things still make sense. This is especially true if you had to do some serious changing of text. Edit as necessary.

"Dear Editor: Thank you for
the opportunity to revise
my manuscript. Unfortunately,
as I am not fluent in dumbass, I
have been unable to interpret
Reviewer 3's critiques."
After your revised manuscript is re-polished into a final version, you need to draft your cover letter to the editor. I recommend starting the letter with a brief paragraph thanking the editor for the opportunity to revise and resubmit your manuscript for their journal. You should also thank the editor and reviewers for their helpful suggestions and critiques (if even you think most of the suggestions were garbage). Then I recommend stating something along the lines of "I've provided specific responses to each of the suggestions and critiques below. I look forward to hearing your final decision." This will provide a transition between your formal paragraph and the subsequent list of changes that's about to follow, as well as obviate the need to write a closing paragraph after your list.

Skip a line or two, and then list out each change you want to address (all of the high importance ones, most/all of the ones of medium importance, some/most of the low importance ones). Briefly summarize the suggestion or critique, identify who said it, its location in the manuscript, and what you did to address it (or, alternatively, why you did not address it). Examples of wording for this section are provided below:

Reviewers 1 and 2 recommended adding r^2 values to Table 3 (p. 21). This has been done for all three models.
 Reviewer 2 recommended revising the paper to focus on racial threat theory. While this is a valid alternative theoretical focus, I feel that racial threat is not appropriate for the current study given the current focus on non-racial factors, as well as the comments of Reviewer 1 regarding the appropriateness of the current theoretical model. Further, using racial threat would generate the problem of..."
 Reviewer 2 questioned the need for spatial lag measurements in the final model. I have expanded the justification for this measurement on page 4, but believe that the inclusion of this measurement is essential in order to control for the spatial autocorrelation of the observations that was indicated by the Moran's I tests discussed on page 12, footnote 3. As such, this measurement has been left in the final model.
Reviewer 3 recommended deleting the third paragraph on page 2, as it was considered unnecessary. On the contrary, I believe the key points of that paragraph are essential to understanding why the study includes measurements for different types of vehicles as well as the differences in operator characteristics. Without it, the subsequent three paragraphs appear disjointed from the rest of the literature review. As such, this paragraph was left in the article as was originally written.
 Make sure you come across as respectful in your comments, especially when discussing suggestions that you didn't implement, and certainly don't word your responses as attacks on the reviewers' or editor's suggestions. If it's clear that one of the reviewers is clearly confused on a certain point, model, or statistic, spin the wording of your response in such a way so that it seems like it was your fault for not being clear, and then justify your original point/method:

Reviewer 2 seemed confused by my use of the term "spatial proximity" on pages 3 and 6. In order to address this potential point of confusion for readers, I have added footnote 2 on page 3 so as to clarify what this term precisely means for this study. 

When you're done, reread your cover letter to make sure, one more time, that you come across as respectful, attentive, and appreciative of all of the suggestions. When you're satisfied that everything sounds good, send the revised manuscript and cover letter to the editor (or upload them to the journal website, as appropriate), and cross your fingers. With any luck, you'll hear back in a few weeks (or a few months) with good news! If not, then see if you can revise the manuscript any further, and then submit it to a different journal to start the process over again.

Image Credits: The Most Interesting Man in the WorldMr. Bean, Reviewer 3Einstein, Letter Writer

No comments:

Post a Comment