45

Referee comments for my paper have just arrived: while one makes some circumstantial and useful observations, the other asks me to add a plot:

The paper still does not contain a single plot. I encourage the author to present some of their results in graphical form, which will greatly enhance the impact of this paper.

My paper regards some study of the flux of gravitational waves emitted in binary systems and is full of mathematical formulas.

While there may be a way to add a very cool and colorful plot to my paper, I don’t think it is necessary; it would be complicated to read and so I should add more text to explain how to read it.

I don’t want to add a plot. How should I respond to the referee?

  • 7
    The referee is being non-specific. You could have added a mind-map of the entire paper, a bit like a graphical abstract. – Prof. Santa Claus Mar 2 at 22:14
  • 146
    Historical anecdote: When Lagrange published his book on what is now known as Lagrangian dynamics, he was proud of the fact that it contained no diagrams at all, because the Lagrangian formulation made them unnecessary. When he came to revise it for the second edition, he discovered that he couldn't follow the details of his own work, and he had to delegate the revision to a student (who no doubt drew a lot of diagrams!) – alephzero Mar 3 at 0:51
  • 7
    Also, keep in mind something like a diagram could satisfy the referee and add clarity, it doesn't have to be a traditional graph which could be useless. – Noah Cristino Mar 3 at 1:58
  • 63
    The reviewer is saying you have presented him with a wall of text. Even people who can interpret formulae easily like a bit of leavening in the form of diagrams. A sketch for the definition of the coordinates used for the binary system and the waves might be a good start. – Neil_UK Mar 3 at 7:22
  • 28
    Your audience will be grateful if you add some diagrams instead of leaving only equations. Otherwise you risk only the most strongly motivated readers will work their way through your paper. – gerrit Mar 3 at 8:49

12 Answers 12

152

While there may be a way to add a very cool and colorful plot to my paper, I don't think it is necessary, it would be complicated to read and so I should add more text to explain how to read it.

It doesn't seem to me that you have really considered the reviewer's request seriously. You seem to think they should just be able to follow the formulas without any visual, thus putting the onus on the readers to do the work of understanding your paper. The onus should be, instead, on you to explain your ideas in the clearest way possible -- and this will greatly improve the impact of your work, as the reviewer says.

I find it hard to believe that there isn't anything visual you can add that would be helpful, no charts, figures, or pictoral examples. When I read technical papers, I have found that visuals that go through an example in detail are almost always invaluable for understanding, regardless of the type of research (very applied or very theoretical, etc.) Are you sure that there is nothing you can add?

The paper still does not contain a single plot. I encourage the author to present some of their results in graphical form, which will greatly enhance the impact of this paper.

You seem to have taken some offense that this was not a useful suggestion, but I find it very unlikely that this suggestion is not useful.

| improve this answer | |
  • 5
    Even if a figure isn't necessary to understand a paper, it can be a very useful tool to draw and focus attention. Beyond the abstract, it's not always easy to quickly get a feel for the content of a paper or its most important points - if there is only one (or very few) figures, it's a very simple way to signpost that this is the most important information, in addition to conveying the information itself. – Nuclear Wang Mar 3 at 13:55
  • 12
    Another relevant historical data point (in line with this answer): Kurt Heegner claimed to prove a long open problem formulated by Gauss, but did such a poor job communicating that his proof was never really accepted, though it was acknowledged as essentially correct ~15 years later (after his death). Wiki does a poor job summarizing, look for the bit on Heegner on this page ; the point of that section is to highlight that it is the presenter's responsibility to clearly communicate to the intended audience. – BurnsBA Mar 3 at 14:34
  • 10
    This. There are some people who can read a sheet of music and hear it playing in their head. There are some people who can read algebraic chess notation and see the entire game in their heads. There are some people who can read mathematics and immediately visualize the relationships in their heads. Most people are not like this. Simple visual, auditory, or cognitive analogies do a lot to speed digestion of complex information for most people. Adding plots is a bit like adding an executive summary - it helps people assess what they're getting into at a glance. – J... Mar 3 at 18:53
  • 3
    Succinctly, "a picture is worth a thousand words." Highly literate people often struggle to communicate well using visuals, and rationalise it by saying pictures are "not necessary." I avoid drawing diagrams because I don't enjoy it and it's time-consuming, but I force myself to because it makes a big difference. – Artelius Mar 5 at 0:35
  • 3
    Not only don't they seem to consider it a useful suggestion, their immediate thought is to create something eyecatching but pointless in order to placate this stupid reviewer.. I've always taught people that an important part of receiving feedback is not just listening to it, but actually taking it on board. – Valorum Mar 5 at 17:22
47

The paper is yours and it is up to you how you revise it. You aren't required to take every suggestion, but you should, at least, consider every suggestion, including what is behind it.

Whether that makes your paper more or less useful is up to readers, of whom the reviewer is one.

But, perhaps the reviewer is more of a visual learner than you are and is making the point that a visual depiction of some data will make it more accessible to a class of users. That might be an advantage.

Respond to the referee with a new version of the paper. Follow the advice, or do not.

| improve this answer | |
  • 44
    This: "a visual depiction of some data will make it more accessible to a class of users". – Philosopher of science Mar 2 at 16:55
  • Yes, completely correct. It is good if you are able to draw the right conclusions from the mathematical expressions. However, in my field, people tend to follow a 'visualize before you analyze' approach, and this would be my general advice (which obviously might be of little help for your specific case). – Snijderfrey Mar 2 at 17:59
  • 1
    Visual effects can help break up a thick text and make it easier to consume with less effort. – Mark Rogers Mar 3 at 1:46
  • 1
    “perhaps the reviewer is more of a visual learner than you are” — They’re not: the theory of different types of learners is nowadays widely regarded as disproved. Rest of the answer is spot-on. – Konrad Rudolph Mar 3 at 16:44
  • 5
    It should be noted that there is no such thing as a "visual learner." theatlantic.com/science/archive/2018/04/… – Jonathon Mar 3 at 16:50
33

Have you consulted with others? Good data visualization is a skill in and of itself; just because you can't envision a good plot from your data doesn't mean there isn't a good plot to be made.

Remember that while you've been steeped in your paper for a while, and find all of the content intuitive at this point, you still want to make it as easy as possible for a new reader to join you in that understanding. Few things can do that like a good plot.

| improve this answer | |
13

"I do not think it is necessary" - to you or the reader?

Please note that there are many hard-to-read, sometimes obscure, papers out there. Not everyone thinks in terms of formulas, and you can increase your readership considerably by a judiciously chosen diagram.

Ultimately, you have to make the decision: if the plot is hard to create, then maybe it's not worth the extra effort compared to the increased readership. But if it is of limited complexity (e.g. because you anyway have a program computing the dynamics of your system and all you need to generate is the graphics), it could be worth adding.

| improve this answer | |
11

It's your paper and the reviewers only offer suggestions, so you are free to say "I won't do it". If you take this path you should give an explanation of why you're not doing it. Ultimately it's the editor you need to convince that a plot isn't necessary.

That said, I've heard people express sentiments such as:

You can always understand a good paper by reading the abstract and the figures.

Which is also how, I suspect, most people start: check the abstract, check the figures and the captions, and then decide if the paper is worth reading in detail. If your paper has no figures, you could be deterring readers (or giving the impression that it's an extremely technical paper filled with so much jargon that only specialists can understand it).

| improve this answer | |
  • 1
    They may express such sentiments, but the “always” is certainly false. I’ve seen abstracts that made me think the writer didn’t read the paper (or is trying to deceive us about the conclusions). – WGroleau Mar 3 at 5:18
  • 5
    @WGroleau Would you consider any of those "good papers"? – mmeent Mar 3 at 6:58
  • It's certainly a generalization, yeah - I don't expect it to always hold in practice. – Allure Mar 3 at 6:59
  • @mmeent, yes. Neither ignorance nor dishonesty on the part of the abstract writer makes the actual paper any worse. – WGroleau Mar 3 at 14:50
9

I don’t want to add a plot. How should I respond to the referee?

First of all, assume that the referee is not an idiot but has at least some very vague idea how graphics could enhance your paper. They apparently did not tell you, but I consider it acceptable to put this onus on you, so you cannot base your reply just on that.

So, here is what I suggest: Collect at least three plausible ways to add a graphic to your paper. The following may help you with this:

  • The referee said plot, which at least in my world is strictly reserved for visualising data. Do you handle any data and if yes, how could you visualise it?

  • The word plot could also have been used to refer to graphical representations in general. Is there any way to do something like this? For example, can you sketch the general structure of your model, the assumptions going into it, or the relevant forces?

  • Did you ever make any sketches, etc. to explain your work to colleagues or yourself on paper, in presentations, or on posters?

  • Do you have any colleagues who are familiar with your work and whom you could ask for suggestions?

Once you collected some ideas, try to flesh them out a bit. Spend at least half an hour for each to sketch a potential graphic. This should give you enough arguments as to why your graphic would not work or have other problems in your paper. Write them up in a response to the referee. Consider including your sketch as demonstration. Of course, if you have general arguments against some form of graphics, include those as well. For example, if your paper does not feature any data and you cannot easily produce some, that’s a good argument to exclude any plot in the narrow sense.

It should go without saying that your attempts to create a graphic should be sincere and you should be open to the possibility that your paper can be reasonably enhanced with graphics after all – in fact, I concur with the other answers that this is by far the more likely case. But, if you can document that you honestly tried and failed, you have a good chance of at least convincing the editor not to enforce the referee’s request.

| improve this answer | |
3

Other answers have suggested diagrams for communicating what you are doing, which I think is picking up on your comment that the paper is full of equations. I certainly agree with the value of such diagrams and would encourage you to include them if possible - the more people who can understand your paper the better.

However, looking at the specific comment, the reviewer is proposing figures for your "results". Your comment about equations seems focussed on the theory. It seems to me that the reviewer is focussed on whatever data you are presenting, if any. Do you have tables summarising different conditions or other data that could be presented visually? It is much easier for a reader to compare numbers visually rather numerically if there are many comparisons.

| improve this answer | |
2

You do not have to do everything a reviewer asks, but you should respond to each point they ask. If you really don't want to do a plot, don't do it and justify it to the reviewer. In these circumstances, if there is something you really don't want to do, I would make sure I address all the other points the reviewer raised, to show that you have carefully considered what they said, even if you disagree with something. (Like additional citations you don't think are relavent but the referee does).

On the other hand a good plot goes a long way to explaing what you are doing, especially in astrophysics. If you make a plot I could then put that plot in my talks (advertising your work for free) if it made sense to what I do. If all you have is equations then I'm not going to put anything about your work in my talks.

| improve this answer | |
2

If you don't want to add a plot, explain plainly why you do not want to add it. It is your duty as an author to address the reviewer's point completely and accurately. From your wording I also feel that

While there may be a way to add a very cool and colorful plot to my paper, I don’t think it is necessary; it would be complicated to read and so I should add more text to explain how to read it.

confuses the necessity of it with the (avail)ability of doing it.

Talking about necessities and abilities:

  • You will need to bring positive arguments as for why it is not necessary to have a graphical representation. You can always admit that you don't feel yourself in the position to devise a suitable graphical representation of your own work; that is, a polite acknowledgement of a genuine 'I cannot draw', which should not impair the validity of your work. Other scholars might attempt that in the future.

  • Perhaps you might also develop arguments as to why it is necessary not to have a graphical representation. This could be bold and winning, and much cooler than a plot, but here you would probably take an epistemological stand that is worth a publication in itself elsewhere, rather than a skirmish behind the curtains of the peer-reviewing process.

Talking about opportunities:

  • Making a plot is sometimes a way to discover errors, or confirming that there are none, or finding special circumstances. I would suggest to give it a try regardless of if you want to publish in this paper. Graphical representations, to be sure, can also hide exceptions. It is a complementary tool. My two cents.
| improve this answer | |
2

Ultimately is your decision, but maybe an additional perspective I have not seem addressed by others answers may help understanding why the reviewer wants you to add a figure.

First I'd like to point out that in no way I'm judging you nor your research, but i'm only taking an utilitarian approach. Thus I like to remind you that reading a paper is a very time consuming task and people's time is limited. No one can keep up with the speed at which papers are published so potential readers will try to be as efficient as possible, while spending as little time as possible figuring out if the paper will be relevant to them or not. In a less elegant way to put it "No one wants to read your paper, You want people to read it"

You have to put the utmost effort that any potential reader that finds your paper, will read it. You want to capture your reader's as soon as possible. If a reader did not bookmark or download your paper when first found, will probably never read it. Anything you do to (honestly) build up the interest is valid, relevant titles, clear abstract good figures.

Maybe there is no place in your paper for a figure, but if there is, consider it

| improve this answer | |
0

Remember the four basic styles of learning.... "reading, watching, listening and doing"

  • Words are for Readers. The vast bulk of academics would be strong in reading.
  • Pictures/graphs/plots are for those who learn well by Watching. These people like to study maps and identify points of interest.
  • Listeners respond well to discussion and lectures
  • Doers tend to be hands-on.

Some things can't be effectively taught with words, like welding for example, so a Doing or hands-on learner has an edge here.

WRT the advisor, they're suggesting that your work has a lot of words with less to keep the interest of other styles of learners. Adding something visual will help to break up a potential wall of text and increase the reach and approachability of the work.

| improve this answer | |
-1

The Referee suggests adding a plot simply because the paper does not contain a single plot. The Referee did not suggest that some specific information in the paper could be better presented using a plot. This is not useful feedback for a technical paper. You can respond by saying that you did consider adding a few plots, detailing exactly where in the paper, but that you ended up deciding against doing that, and then explaining why on balance the plot would make the paper worse.

What matters is that you can explain that the content of the paper is what it is for good reasons. So, plots are not included, not because you are too lazy to bother making plots that would likely make the paper a lot better, but rather because it would actually make the paper worse.

Technical papers don't need to look attractive, you don't need to catch the attention of readers using nice-looking figures and plots. It's unfortunate that there is now a trend where people do include figures for the wrong reasons in technical papers. Editors of journals don't act against this because figures add to the publication charges, particularly color figures.

Your audience are researchers who want to read your paper for the technical information contained in it. It's then best that the information is presented in a form that's most appropriate for them. If a plot were useful, then you would likely already have included a plot in your paper. If a potential reader would not have read your paper just because it doesn't contain a plot then obviously that reader is not doing serious research for which he/she needs to read your paper.

The logic of including plots to catch the attention of readers is therefore only appropriate when you're writing for a wider audience of people who are not directly involved in your research topic, for example in a review paper.

| improve this answer | |
  • 2
    Negative for "technical papers don't need to look attractive". Selling the technical paper in the best way is one of the most crucial factors that influence the impact of the paper. – KratosMath Mar 4 at 9:31
  • @KratosMath The impact of the paper as measured by meaningless measures like how many times the paper is cited, which does not take into account the real impact of the paper in the field. It is true that scientific careers are based on such measures but this means that you should also write a lot of low quality papers with lots of stupid figures in there that then get cited a lot by your peers who don't actually read your paper. – Count Iblis Mar 4 at 22:13
  • @CountIblis yep, exactly. Which is how we got where we are today. – barbecue Mar 5 at 16:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.