NINDS's Building Up the Nerve

S5E6: Writing Impactful Publications

• NINDS • Season 5 • Episode 6

Share episode

In the fifth Season of the National Institute of Neurological Disorders and Stroke’s Building Up the Nerve podcast, we help you strengthen your science communication skills with tools and advice to use throughout your career. We know that navigating your career can be daunting, but we're here to help—it's our job!

In the sixth episode of the season, we talk about Writing Impactful Publications, focusing on how to structure and write a paper, choose a journal, and craft engaging and accessible figures.

Featuring Bang Wong, MA, MS, Senior Director at Vertex Pharmaceuticals; Marina Picciotto, PhD, Professor at Yale University; and Tanya Garcia, PhD, Associate Professor at University of North Carolina, Chapel Hill.

Resources

Transcript available at http://ninds.buzzsprout.com/.

Lauren Ullrich:

Intro music] Welcome to Season 5 of the National Institute of Neurological Disorders and Stroke's Building Up the Nerve, where we help you strengthen your science communication skills with tools and advice to use throughout your career. We know that navigating your career can be daunting, but we're here to help— it's our job![music fades]

Marguerite Matthews:

Hello, I'm Marguerite Matthews, Section Chief for Career Preparation in the Office of Programs to Enhance the Neuroscience Workforce, also known as OPEN, at NINDS.

Lauren Ullrich:

And I'm Lauren Ullrich, Section Chief for Career Advancement in OPEN, and we're your hosts today.

Marguerite Matthews:

In our last episode, we discussed securing funding for research. Today, we're going to talk about writing impactful publications, which will focus on how to structure and write a paper, choose a journal, and craft engaging and accessible figures.[ music] Joining us today are Dr. Tanya Garcia, Dr. Marina Picciotto, and Mr. Bang Wong. Let's do our introductions.

Tanya Garcia:

I'm really glad to be part of this podcast to talk about writing. It's one of my favorite topics. I'm Tanya Garcia. I'm an Associate Professor of Biostatistics at UNC Chapel Hill. And as a biostatistician, I feel that we have a lot of responsibility to carry out proper analyses, make big decisions about sample sizes, and the design of clinical trials and other important experiments. And so because we tend to play in different people's playgrounds, it's really important that we're able to articulate complex ideas in the simplest way possible. So it's been very important for me to train my trainees to communicate these really hard things in a way that people from different backgrounds and different expertise understand. And I believe I found a really effective way to do that. And it's basically based on 3 key principles. Desire, trust, and simplicity. So I believe we have to identify what the field wants to solve, creating the desire. And what evidence the field needs to be convinced of a solution. Generating the trust. And we have to communicate that desire and trust in the simplest way possible. So, I'm really excited to discuss more of those concepts that I work on with my trainees.

Marina Picciotto:

I'm Marina Picciotto. I'm a Charles B.G. Murphy Professor in Psychiatry and I'm Professor of Neuroscience and Pharmacology at Yale University. I'm also Director of our Interdepartmental Neuroscience Program. My lab studies what acetylcholine does in the brain and tries to grapple with the idea of what a neuromodulator really is. We're interested in the neurobiological basis of typical behaviors and what goes wrong in psychiatric illness. My other role in the past has been as Editor in Chief of the Journal of Neuroscience, so I'm also experienced with the other side of publishing. Most of my time there was spent handling manuscripts and then later on as Editor in Chief, watching other people handle manuscripts and dealing with problems that come up when accuracy is not followed and with other issues that come up with scientific misconduct. But in general, what I did see was how carefully people craft their science and how carefully and transparently people communicate their science and how hard people try to be of service to other scientists in the field. I would say the three words that describe my science communication style or philosophy are storytelling, transparency, and accuracy.

Bang Wong:

Hi, I'm Bang Wong. I work at Vertex Pharmaceuticals as a Senior Director in Data Technology Engineering. Over my career, I really sought ways to combine art and science. In graduate school, I discovered medical illustration and it set me on this path. At the Broad Institute, I work with researchers to apply visual approaches to explore and explain scientific data. In that process, I began to notice patterns and wrote 35 articles published by Nature Methods on data visualization. And now at Vertex, I'm applying these techniques in a therapeutic setting with the hope of speeding up research and the discovery of new medicines. I'd say the three word that describes my communication approach would be "simplify to clarify."

Lauren Ullrich:

I love it when there's a slogan.

Marguerite Matthews:

Yes![Laughter] Simplify, or clarify to simplify.

Bang Wong:

Simplify to clarify!

Marguerite Matthews:

Simplify to clarify. See, I needed you to clarify that for me![music]

Lauren Ullrich:

So we like to start out with the bigger picture to frame each of these episodes. For this one, it's why do we communicate science through publications? You know, especially now, we have a lot of different formats that are available to us, readily available to us, but publications really remain the currency of science. So what purpose do they serve that we rely on them so much as a field?

Marina Picciotto:

One of the things that I found out from our librarians, and librarians are amazing by the way, is that putting things on paper is likely going to be the best way to preserve knowledge for the longest period of time. We still have information on papyrus from many, many years BC, and it is very likely that all of the electronic methods we have of archiving and of communication will become out of date before our lifetimes are over. And can be very easily disrupted. And therefore, paper publication, believe it or not, is likely to outlast all of us. So that doesn't quite answer your question, but writing things down is really important and librarians keeping those things safe is equally important. And on the other side, why do we write papers? Why do we communicate in this way? It's because when we make discoveries, if we don't communicate them in a lasting way, somebody's going to have to go out and reproduce them and do it all again. It's as if it didn't happen. So I think that scientific communication of all kinds is important, but scientific archive is really, really important.

Tanya Garcia:

Yeah, I completely agree with Marina and I actually, I wanted to echo what Bang said in terms of simplify to clarify. I believe that writing really is, I found, one of the best ways to clarify my own thinking and my own creativity with the project. And I haven't been able to do that just in speech or other forms. Although I'd love to hear from Bang about the drawing and being able to clarify science that way. But I found that writing allows that clarity. And also, I think that with the peer review process, there is that way to really double check the science, because I do think it's important to put the right science out there, even if it isn't positive results. Like Marina was saying, we don't want to have to recreate the wheel every time. So I think writing is a great form of putting our work out there.

Bang Wong:

Yeah, I think writing, it's probably the kind of the most considered form of communications. You know, every detail has been sharpened and been honed. I think for the reader, they can really focus on that information, their own terms, without distraction. You know, there's a strong connection between seeing, thinking, and understanding. And the publication has a slow medium I think really facilitates that powerful iterative cycle of taking in information. And in contrast, I think information that are presented, say, in posters or slide presentations, are often transmitted in large public spaces. And I think it can be really hard to listen, read, and in the case of posters, have a conversation at the same time. And of course, with publication, the trade off is not having that dialogue with the author.

Marguerite Matthews:

Yeah, I think that's a good point about having time to process and process in the way you need to, whether it's rereading some things, some people like to skip around. But when you're listening to a talk, whether it's someone going through their poster or a slide, you're like having to process at whatever speed they're going. I mean, the poster gives you a little bit of static information, but it's also being really distilled down, right? You're only getting a snapshot of what's happening, whereas the publication is trying to give you a much fuller picture of what happened and what they think that the results mean.

Marina Picciotto:

And I like what that says about how everyone who reads a paper reads it in a different way and is looking for a different thing. Someone who's looking to do the same experiment is looking first at the methods. Someone who's maybe just trying to get an overview of the field is maybe just looking at the introduction first. Someone who's interested in that data is going to hone in on the figures themselves. So I think everybody can get something different out of a scientific paper.

Marguerite Matthews:

Yeah, and just as there's different forms of live communication, like talks, posters, um, we also see that publications come in a variety of forms depending on what you're trying to communicate and how. And so can you all talk a little bit about the different forms of scientific papers and their purpose? I mean, we've got reviews, brief communications, methods, papers, longer studies, meta studies, book chapters, et cetera. And do you think that those different forms of publications serve different audiences, or is it just providing the same sets of audiences with different means of communicating a particular topic?

Marina Picciotto:

As an author, I like to think about different forms of publication as different venues for getting out different aspects of a hypothesis or of a theory. So for example, if you start a whole new area of science, what you first, I think want to do is get out there some basic data, descriptive data. Things that would be in a research publication that is strongly peer reviewed, but perhaps more narrow, and then you build up a bunch of that data, and then you want to say something about how it sticks together. And so you might write a review paper, tying together all of those pieces of data into a bigger idea. And putting out there almost as a challenge to the field, here's what this could mean. Maybe we should all jump in there and try to work on this idea. And then a methods paper can be a companion to that, showing new ways to look at specific problems using different tools. So I do think of them as different tools in a toolbox to get people to understand new ideas or to pursue new concepts.

Bang Wong:

Well, the publications I've been involved in, I haven't played a large role in choosing the type of format, but I do get involved in them. But I think whatever the form, I would say that an engaging and informative overview figure, along with a descriptive title and an opening premise, shouldn't be underestimated. An illustrative schematic and an overview figure can make the material more accessible to a wider audience. It gives context and constructs a mental model for the technology or the experiments or the data to be presented.

Tanya Garcia:

Yeah, I wanted to add a different perspective to this question. So as a biostatistician, one of the key parts of my job is sometimes often developing new methods to properly analyze this really complex data that we have. And so often, for me, the methods papers is like the papers with lots of equations, lots of theory, but honestly, for the most part, mainly statisticians and data analysts can understand. But then to really get the work out there, I think it's really important to publish in a venue that allows other scientists to use these new methods and so that could be, I call them like the clinical or non-statistician papers. And so that's a way to get those methods used in the field. And then also just as a consumer of methods, honestly, although I'm trained as a biostatistician, often reading about these stat methods can be really complex and over my head. And so I often turn to one of the clinical or non-statistician papers, because often I found they write about a method that I'm trying to learn in a really beautiful and simple way. So I use that as my like "Oh, here's how I can start understanding this method and its purpose." And then if I need to dig into the details, I can go to those heavy equation, heavy theory type papers.

Lauren Ullrich:

So kind of building on that, we often talk about, like "Oh, this is a good paper, or this is a great paper." But I don't know that we necessarily talk about what made that paper great? Is there one way that a paper is great or maybe there are multiple different kinds? And then how do you decide what makes a good or a great paper?

Tanya Garcia:

I'll chime in on that just building on what I was saying with all these equations and everything. I definitely don't think that there is one definition of what makes a really good paper. But just personally, I really value simplicity and accessibility. So one of the key things that I find a good paper has is I can understand what they are saying. I can walk away from this paper and know, okay, those are the key concepts that are being explained. And also I can understand why this work matters. What's the impact of this idea. And I can then build onto it. If those two key parts are there, the why is there, and the simplicity is there, that to me is like a really nice elegant paper.

Bang Wong:

I would say a good paper has good figures.[laughter] I mean certainly data figures and graphs are essential to any scientific paper. Figure can enhance the paper. Most when I think they have a productive redundancy with the text. They don't repeat everything that the text has, but complements the written words. And if the reader were just to review the figures, and I think some people do just read the figures, they should really hold together and tell their own scientific story.

Lauren Ullrich:

Yeah, I mean, to go back to what you were saying Bang about having a nice summary figure. I just remember when I was in graduate school, there was this paper by Amy Arnsten about the PFC(Prefrontal Cortex) and just had this beautiful summary graphic at the end. And the whole paper was written very well. But like that figure was just something that I could keep coming back to to understand what was going on. And then I ended up using that paper to teach undergrads because like every figure told a story. And the methods had a figure. It was so straightforward, even though the actual experiments were relatively complicated. You know, they're talking about signal versus noise, and a bunch of different receptors, and tasks with monkeys. Yeah, trying to articulate like exactly what made that figure so easy to understand is still kind of challenging for me, 20 years later.[laughter] Maybe not quite that long, but, you know, many years later.

Bang Wong:

They do really play an essential role, and it's a shame that sometimes when you run out of space, the intro figure is the one first to be cut. But I think that they provide that scaffolding, either describing the experiment that's designed. So, then all of the data that gets presented after that, there's something to hang it on to. And so I think they play that role of making the information and the rest of the paper more accessible. I think that can be done outside of the figure as well, and how the introduction is written of course, and I think the combination of a well written introduction for setting it up and the overview figure can be a powerful intro to a good paper.

Marguerite Matthews:

I imagine also the journal you choose, if you want to really be able to describe what you're talking about with a lot of figures or make it more clear what it is that you want to communicate, something that might be really complicated that is helped by images, having a journal that allows for that versus some journals which really are maybe space saving or you're limited by the number of figures. And so I think that is something to keep in mind if you are a person who really values the ability to look at a picture, you know, and tell a story versus maybe you don't need a visual aid, it's actually not as helpful and you can perfectly communicate what you need to do in a series of paragraphs.

Bang Wong:

That's a good point. So there's a big difference between publishing in Science where every pixel counts than something like PLOS or an arXiv journal, where you have a lot more space to do that work.

Marina Picciotto:

I think for me, it's always that it made me think a new way about something that I find interesting or important. And so I think at the heart of it, it's that the scientific premise is firm enough that I believe what they wrote. But then it really intrigued me because it wasn't the way that I thought about those data in the past. And I think that also involves a degree of storytelling and storytelling can be done visually, as Bang was saying with a really good figure, the story can appear in your own mind, or if you're good with words, it can appear in the text. There's also an untapped wealth of ways to communicate visually that I don't think journals, traditional journals have taken advantage of yet. And that is, 3D visualization of data, virtual reality representations of data, I think we're increasingly going to see those kinds of modern ways of telling a story visually play out now that most journals actually do present their data online, both in text and figures classically, but also now increasingly in other visualization methods.

Marguerite Matthews:

I'd like for each of you to sort of talk about your approach to writing a paper. Maybe there's a certain methodology you use; you start from the beginning and work your way to the end, or maybe you tackle perhaps the easiest part, which a lot of people think is the method section. You know what you did.

Lauren Ullrich:

Hopefully!

Marguerite Matthews:

Hopefully you know what you did and you're able to articulate that in words. Um, And also how do you work with other people that are on the paper? Whether its co-authors, maybe a lead author, or someone who's kind of taking the lead on running the experiments or maybe you're the person who is kind of putting all of these things together. Yeah, just talk about what is your process like and what are some tips that you've learned over the years that maybe weren't taught to you, but you've found to be tried and true.

Marina Picciotto:

Well, my first ever paper, I definitely started with the methods. I needed the easiest possible way to do it. But I would say that after getting over that hurdle, the way that I start a paper and the way that I tell all of my lab members to start a paper is with the figures. And that is, if you've visualized what you found in a figure very clearly, then you can describe what you see in the figures, that's your results, and then you can introduce why you did those experiments in the first place, that's the intro. And then you can sort of sum up where you are and where we should go next and that's the discussion and then you're done except for that horrible abstract and figure legends which nobody enjoys.

Marguerite Matthews:

Yeah, what is it about figure legends that feel really hard to like encapsulate everything that's not already said in the main body of the paper and yeah, I actually found captions to be some of the hardest things to write.

Marina Picciotto:

You'd think they'd be so easy, and it's really misery!

Tanya Garcia:

Yeah, so actually the way I start with papers actually piggybacks off of Marina, what you said about believing the scientific premise. And I think if your readers don't really believe you, that can be a killer or lead to a rejection for a paper. So often I start with my own papers and with my trainees, really nailing that scientific premise. And the way that we work on it was something that I learned through this really intensive grants writing program that I did, and basically the idea is that to get your scientific premise, you want to think about creating this funnel shape of your argument where the idea is that you're narrowing in on to your specific core argument. And so the three big parts that I think about is start with the problem or the why. What is the field tackling? What's the big problem that they're trying to address? And then the next part is the background. What has been tried to overcome that problem and why hasn't it been solved yet? So those are some of the limitations. And then once you've established those limitations, that narrows in on what I call the opportunity space. This is what we're going to do to help overcome those problems that are going on in the area. And so I found with my trainees and with myself as well, having that funnel shape in mind really helps to narrow in on what we're trying to say. And in theory, it seems very easy, okay, problem, background, opportunity space, but just frankly, it takes a lot of iteration. So my writing process is very iterative. I encourage my trainees just write. It's not going to be perfect. I've written, you know, up to a hundred versions of that scientific premise, just to try to get nailed down. And so in summary, getting that scientific premise nailed and iterating over and over again to get it solid.

Bang Wong:

Well, my experience has been similar to Marina about starting with the figures, and, um, I've even witnessed people go as far as maybe drawing out the figure with fake data, what they would expect it to show, and then designing the experiments to basically fill in those figures. And for me, for the Nature Method series of articles on visualizations, it was kind of natural for me to start with the figures, because of the nature of the material. And there I really had to demonstrate the point through the visual, because it was about that topic. But on the point about writing figure legend, which I also struggle with. I was once given the advice of don't say what it is, but why people might care. So instead of trying to describe this color means this, and this thing means that, say why it matters. And I think Tanya also made that point about starting with the why.

Lauren Ullrich:

Yeah, so moving into figures, obviously they are really important part of scientific communication. Oftentimes people maybe don't have the best tools to make really good figures, or don't know how to best display the data that they have, especially now, where a bar graph is not gonna cut it.[laughter] I need a box and whisker at the very least! But the more you try to put into that figure, like the more important the graphical elements and the simplicity, like that balance between displaying all of the data that are necessary, but also keeping it simple and to the point and understandable. So are there any guidelines that you all use when you're creating your figures or how do you make them informative, accessible, accurate, et cetera?

Bang Wong:

I think with figures, I think it's a tool that we code information first and then the reader later has to decode the information. So that the effectiveness of figure, it doesn't just depend on how clever we are in coming up with them, but it also depends on how well the reader is able to correctly interpret the information. So I think making figures informative and accessible needs to start with the audience and their needs. And for me, there's a couple of things that I asked myself as I go through the process of making figures is trying to make what needs to be seen kind of easy to see. And so in some ways, that's laying out the figure. And as you have already alluded to this, I mean, we work with lots of data, high volume, diverse, dense data. And I think the temptation is just to stuff as much as we can into a figure. And so while we can encode 10, 12 dimensions with color, with shapes, with line thickness and everything, it might not be possible to decode the information out of there. So one approach would be to break up the figure and show it in small multiples so that there are four dimensions of data that would be important to look at together. And then next figure with another two or three dimensions of data that would be valuable to look at together, but they're overlapping somehow. So like they say cell type identity or something is in common. I would say the second thing is don't reinvent the wheel. Make use of established precedents and conventions. So if certain types of trends are often displayed in a certain way in a certain field to show certain types of outcomes, build on that, don't do something completely different because I think those precedent can help communications to be highly concise and to the point. You know, they say a picture is worth a thousand words, but sometimes a word can be worth a thousand pictures. So like if you can't come up with just that perfect graphic to represent some abstract ideas, I think just use a word. Because I've seen people try to condense the whole figure into an icon and stuff it into another figure to say some analysis that they did. But it just doesn't come through. And then finally, I would say the final step is to just kind of erase back, ask the question "What can I take out of this figure? What are all the things I can take out of this figure that doesn't compromise its sophistication?" Because I think every picture, every bits of text contribute to the intricacy of the figure. And our temptation is to fill up all that white space. But I think the judicious removal of that material, it's not typically done, but I think it's going to be a very valuable part of making figures more effective. Less can be more.

Tanya Garcia:

I'm learning a lot here, and I'm taking down a bunch of notes from this. This is great! But Bang, I had a question for you to follow up on that. How do you get feedback on your figures in the sense of, you know, because sometimes we're so entrenched in our own science that it all makes sense to us when we see it. So how do you get that fresh perspective on "Oh yeah, this figure works," or it's not, you know, the decoding isn't really getting through.

Bang Wong:

It's a painful process, but you invite critique. And this is kind of user centered design at its best. Because it's really about what somebody will take from the figure. And so you put iterations out there and ask for interpretation. And it could be small things on just colors that are hard to discriminate. Or it could be that it's confusing what the trend, what the takeaway message from the trends are. And you might find different ways of plotting those data or different ways of aggregating the data. But getting feedback often, early, as painful as it is, I think it's probably the best approach for honing more effective figures.

Lauren Ullrich:

Yeah. I went to a symposium at the University of Utah this spring and they did an exercise with the fellows where they had them submit their summary figures without a caption, and a stranger went through and talked out loud through what they thought the figure meant based on just the graphical information. And then the person who actually made the figure would say "yes, that makes sense, or I had those lines because it was vibrating, not because it was moving or vice versa." And that person talking aloud through not just what they think the figure means, but why they think it means that, was just so revelatory for me. And now I just want to do it with every, [laughs] every single, like, trainee that I come into contact with. Um, And Bang, your Nature Methods papers were highly cited in the presentation to kick off the event. Um, and I think that is a theme that's been coming up again and again on this season is that you don't always know how your message will be received unless you ask someone, and multiple someone's preferably, because everyone's going to do that decoding phase slightly differently. So yeah, thank you for asking that question, Tanya, because I think it is really important that we can't just think our way into understanding how something's going to be received by somebody else. We have to actually ask them.

Marina Picciotto:

I think comprehensibility of figures and particularly putting in schematics that guide how your eyes look at the figures are absolutely essential. But I want to add one more thing, which is to include as much raw data as possible. I think that we move a lot toward schematization, which I think is important, but not sufficient. And prettying up raw data is something that we all really feel like we have to do so that everything looks perfect. But as much as possible, including raw data, if not in the figure itself, at least in the supplementary data and supplementary materials, I think is really important in showing what all of the conclusions are based on and what you can expect in terms of how reproducible or how similar data are across iterations.

Tanya Garcia:

I had another question for the group in terms of, you know, I'm thinking about folks who may not be able to see or hear. I know that, for example, with text, there are now translation services so that you can hear what's written. I'm just wondering, with figures, has anybody encountered what's the new way for people to be able to see those figures?

Marina Picciotto:

This isn't actually for visually impaired people, but because we have someone who is color blind in our lab, we've been sensitized to the fact that our traditional green-red colors for cells are absolutely miserable in terms of being able to, uh, show differences in figures. So there are websites where you can go to find colors that are friendly to people with many different forms of colorblindness. But a more profound visual impairment, I don't know what kinds of accessibility is available to, for example, describe content in figures and have that translated beyond having an interpreter who does that.

Lauren Ullrich:

Bang, I know you've written about some accessibility.

Bang Wong:

Yeah, I mean, I think thinking about colorblindness is important. As Marina mentioned, there are websites out there where we can get colorblind safe palettes. There are other websites where you can submit your figure and it'll emulate what a colorblind individual would see. And then you can see how, well someone who would be able to discriminate the different aspects of your figure. But I think, not just for accessibility purposes, but using colorblind colors, it's just good practice, because those colors just tend to be more discriminable from each other, and just for people with so called normal color vision, they can also benefit from that. We're talking about color as one aspect of accessibility, but of course it's greater than that. But there are lots of approaches to avoid color. I mean, we love color, because color tends to be our association with things that look good [laughs] and pretty. And so it tends to be the first kind of things that we, first graphical element we grab to display data. But there are lots of approaches to avoid color and to use other means.

Lauren Ullrich:

Yeah, and in terms of visually impaired or blind readers, I think that is a huge challenge. I know of one researcher who has used sound to sort of represent her data in different ways. But many times, images in particular are just very inaccessible. Even though you have 508 compliance and different things like that, the descriptions are not really sufficient. So I think that is a good thing to keep in mind when you're writing your paper, like that should be able to stand on its own. Like everyone should be able to get the information just from the text. We have someone in our office also who is visually impaired and he's like "I don't understand figures. Give me a table! That's how I understand data and process it." So having access to the raw data or summary tables, I think the bottom line is probably presenting the information in as many different ways as you can within a given paper or a series of papers is probably the way to make things the most accessible for everyone.

Marguerite Matthews:

Yeah, and I think your point, Lauren, also ties in sort of a lot of things we've talked about presenting raw data, be able to accurately describe what it is that you are trying to communicate in the figure. Whether that's a legend, a caption, just describing the results so that the reader can understand it. Because even some folks who may not be visually impaired still may not be absorbing what it is that you're trying to communicate. Their decoding has, you know, there's something fell apart there. And it may not even be the fact that it wasn't a well designed figure, right, or a well designed image. It just there's some pieces missing that the reader may not be able to just immediately decode or decipher upon first glance. And so I think it can be helpful to make sure that all of these things are complementing each other and there's consistency throughout, right? Like, there's clarity everywhere. And for those who may be missing a piece of the information for one reason or another, can still figure out... haha that's kind of funny, figure out![laughter] You know, how all of these pieces are coming together and what is being communicated to you?[laughter] I don't know why that is cracking me up, but I thought that was unintentionally witty. So when you're thinking about a publication, regardless of what the publication type is, how do you approach discussing the significance of the work, without overselling it, without oversimplifying it, without overcomplicating it? And how do you think it fits into sort of like a broader scientific landscape? I mean, I love, Tanya, what you said about an opportunity space. So it may not necessarily be a gap you're filling, it's you are approaching a problem in a different way or with a different set of tools, just kind of approaching it maybe in a way that has not been done before. And how do you think that fits in sort of with the larger literature? Like how do you make sure that what you're producing or what you're communicating to your audience is appropriate, but also like you've found the right rhythm of what you want to communicate at any given time?

Tanya Garcia:

Yeah, that's why I like that term opportunity space because it does open this opportunity to move science forward and whatever that means. There's a couple of things that I think about when talking about how does my science move the field forward? I read a really great book, "Writing Science" by Joshua Schimel, and he had this idea of at the end of the paper, instead of thinking that you're closing the loop in your paper, this idea of coming back to the beginning, but having an opening at the end. So you're opening up to what is that next step. And that's what other authors can continue to build on the science that you've created. And I think it's important to know like the impact of your work. I do think it goes back to what I mentioned earlier about understanding what are the limitations that exist in the field right now. Because then you can create contrast of, okay, this is what's been done. Here's my work. How does it compare to what's been done? How did what we do change what's been done or did it build on something that's been done or did it reaffirm something that's been done? So all of those are ways to move the science forward. I love your question about not overselling your work, because I mean, I still have this challenge of "oh, I want my paper to be in the best journal, and so I'm going to say that I did something magnificent!" So honestly, I do what Bang suggests in terms of iterating on the figure, like I iterate on my message. I do different versions in terms of you know, I cured this disease [laughs] versus, you know, other variations of that. And I open myself up to different critiques to really hone in on what's the key message. And ultimately what I find is I end up saying, " what is the next logical step that my work led to?" That ends up being the one that's most honest and exciting for the work.

Marina Picciotto:

I think that any experiment that is done well and well controlled is worth publishing and is important to put out there because either it's new, or it's a confirmation, or it's a refutation, or it's a great idea. And I think that what's important when you're writing a paper is to know who your audience is going to be. When I was a postdoc and I was entering a completely new field, the papers that were most useful to me were not the big stories, the great new advances. They were the really compact, well done, sometimes very descriptive papers that told me something about for example, an animal might typically behave, or how a molecule might work in a cell, or how a drug might affect behavior. And being able to look at that one result, look carefully at the methods, allowed me to have a base to stand on when I looked at what might a molecular genetic manipulation do to this well-characterized thing. And those papers are, for me, the bedrock of what science is about. They're not Nature papers. They're not Science papers. They're not the ones that attract "oohs and aahs." But they're the ones that allow me to do my work and take the next step. And maybe that step's a little one. Maybe it's a big one. But I think that when you think about your own work, you should say, "who's going to be interested in this?" Is it the people who are doing pretty much the same experiment and want to have the next step in this? Okay, that's one kind of way of writing a paper. Is this the people who are in a broader field and need an idea that's new? And I've really supported that, and here's the idea. Or is this so big that it's going to change the lives of everyone out there, like DNA or Theory of Relativity, in which case you put it into broad, broad interest journal or even a newspaper. But each of those has a really important place in science.

Lauren Ullrich:

And kind of following up on that, how do you decide you have enough for a paper? Because I think that kind of ties into this, like, why are you writing the paper? What are you hoping to accomplish? Or who are you talking to with the paper. Because papers can be this one little chunk or they can be seven figures and they have, you know, panels A through ZZ. How do you all think about what makes a given paper? Where do you end?

Marina Picciotto:

The overall goal of each paper is different. And if you are trying to influence a large number of people across fields, you're going to need a lot of evidence, probably going to need those 12 figures with 13 panels each, and supplementary, to convince people that this new thing is true, important, worth following up on. And those papers are both skill and luck because you need to have discovered something new, that's luck often, but you also need to have the resources and the rigor to support a completely new idea. And if you're aiming for that, my advice is that you have many, many backups so that you don't depend on luck and you don't depend on unlimited resources in order to publish that one home run. Make sure you also have solid singles in your arsenal so that you can publish something that is bounded. Maybe it is descriptive. Maybe you have one idea that you can support with a smaller number of experiments. Or not even an idea, but an improvement on a method or something that allows a new vision of things that we already know. And there is no answer to when you're done because it depends a little bit on how big the idea is, depends a little bit on who else gets excited about what you've done. So I think it's important to be flexible.

Tanya Garcia:

Yeah, I'll add to what Marina said. It is very difficult to know when you're done -done. But one thing that does help me is I try to think about what is the core message I'm trying to say in this paper? And then also think about, okay, what could be the big questions that my reviewers or my readers will have about my work? And often that's where I find, you know, working with co-authors, trainees, thinking about okay, what are the biggest critiques I think we're going to get. And what are the ones that we really need to be solid on, so that we can convince our readers? We're not going to be able to answer every single detail, but these are the big three or whatever that we can really answer. And then of course we'll get critiques back and then we'll know more from there. So that gives me some way to know of when I'm done. Like I've answered these three big critiques I'm going to get, and let's try this out.

Marguerite Matthews:

So when it comes to citing other papers in your work, are there specific things you definitely want to say? Like perhaps maybe there's some peers that you feel like maybe your work will not be taken as seriously if you omit certain papers or you have collaborators that you definitely want to acknowledge the work that they've done or even thinking about diverse perspectives and knowledge bases. I mean, I know you talk Tanya about being a biostatistician. So I imagine maybe you're pulling on specific folks who've utilized the work that you're doing in a very specific way and maybe how it's similar or different. So yeah, can you all talk about what is your approach to citing maybe specific papers or specific people that are in the work that is relevant to the work you're doing.

Marina Picciotto:

I'd like to talk a little bit about equity in citation because I think it's something that we've neglected until quite recently, and it's super important. As you probably all know, the majority of individuals who are cited are of historically overrepresented backgrounds. And people may have published around the same time and those who are from minoritized backgrounds, or who are women, are cited less often than their counterparts. And I think that there are a lot of reasons for this. One is that we tend to cite the people we've heard speak, for example. And there's a bias in who gets invited to present their work. We tend to cite people who we know and who perhaps we've collaborated with and those who are of majority backgrounds often collaborate more widely or in more prestigious spaces. So, I'm only at the beginning of this, but I'm trying to take a look at how historically I've chosen which laboratories or which studies to cite. I know that the core papers that I cite are often those that I heard at a meeting that, again, changed my whole perspective. And I'm so blown away by how it changed how I think that that becomes unconsciously one of the bases for how I design the next experiment. So I always cite those because they're so salient to me. But then what I'm trying to do now is to go back and see if I can be more broad in how I search for citations that support those ideas. No idea comes out of a vacuum. Ideas come from percolation between brains and from grassroots studies that are out there before that one big starburst of a paper gets published. And I'm right at the beginning of trying to figure out for myself how to do this, but I'd like to urge everybody out there to start thinking about how to be as equitable as possible in their citations.

Marguerite Matthews:

All great points. Thank you, Marina, for bringing light to that.

Lauren Ullrich:

So I think maybe one of the most important parts of the paper is the abstract because sometimes that's the only thing that people read. And it oftentimes is the first thing that people read, so you're really setting the tone. And you have to summarize your, how many thousands of word paper into a couple hundred words. Or maybe one picture if you're doing a graphical abstract. How do you approach the abstract and what do you think makes a good or effective abstract?

Bang Wong:

Well, I can speak to the graphical aspect of the question, but of course, I welcome your perspective on it too. I mean, this is kind of the ultimate abstraction [laughter] uh, literally! And maybe I'll just start over.

Lauren Ullrich:

No, we have to keep that part in!

Bang Wong:

Okay![laughter]

Lauren Ullrich:

This is the pun episode.

Bang Wong:

Yeah, the typical graphical abstracts that I've seen have usually about procedures or comparison of conditions, normal or experimental, or this is the normal pathway, and this is the pathway that we tweak. And these figures tend to portray a continuous process in discrete steps, and I think that's the challenge there. So it's really imperative that we create continuity through the imagery and the written description that goes along with those imagery, so that each step in the procession is really understood by relating it to the previous step and the subsequent step. So in the design of these kind of procedural schematics, often I like to adopt kind of A to B structure in which A and B are states that are connected by an action. You know, the states are often depicted graphically, and the action is a piece of text describing that transformation from A to B, say, " cut with restriction enzyme." And to create good visual linkage between the steps, I tend to redraw all the elements from the previous steps, just highlighting only the things that's changed. Because when a reader needs to follow a series of events, it's helpful to account for all the graphical elements introduced and removed from figures. So when these number of elements do not match up from one step to the next, it's really confusing and really can compromise the utility of these graphical abstracts. And here I think there's lots to learn actually from comics, they do this really well, taking a kind of a continuous story and breaking into discrete frames.

Lauren Ullrich:

I love that. I never thought about comics before, but that's such a good tip. And what about the written abstract?

Marina Picciotto:

Well, I was going to say that some journals still have their abstracts broken up into pieces like their introduction, methods, results, and conclusions. And in some ways, although that can be kind of clunky, it does provide a route to starting to write your abstract. Can you say why you did it, what exactly you did, what you found, and why that's important in 200 words? If you can, then you're done. Usually what happens is you've got then about 850 words and you have to start chopping out every other word. But if you can keep in mind that what you're trying to do is to distill why, how, what, and why is it important, then I think you can at least get the elements of what you need to get into the abstract. Sometimes methods are so complex that trying to get them all into the abstract is simply impossible. And so you have to figure out what to leave out.

Tanya Garcia:

Yeah, so actually in the past year, a colleague and I, Nicole Dalzell from Wake Forest, she and I developed this tutorial of how to teach mentors how to train their trainees on writing. And the abstract is one big thing that came up. And so I just wanted to share a couple of things that I believe piggyback off of what Marina and Bang said. One is what Marina said- the why, the what, the how, and the impact that's equivalent to this funnel shape I was talking about. The why, the background, opportunity space. And then at the end of it, what are your results? So a little taste of what's to come. And so thinking of that funnel shape to construct the abstract, I think also is a nice way to sell the story. What Nicole and I found was that, especially with trainees, it's really helpful, although it sounds a bit constrictive, when they're writing the abstract, to make sure they stick to the 250 and 300 word limit, because it forces them to focus on what's the core message. And then with trainees and anybody else, we have like, we call it the wishlist. You know, what are the other things that you wish you could have added? You could put that in that list. And so then when you're iterating on the abstract, it can help to write that compelling abstract. I do want to share as well, some of my colleagues here are very much into AI and ChatGPT and all of that. I learned that there are apps that you can plug in your whole paper and it will spit out a 250 or 300 word abstract. So, you know, that's a wonderful tool, but I very much believe it's still important to know how to write a solid abstract because it's going to spit something out and the quality of it is up to you to really understand how good it is. So yeah, whatever your thoughts on AI tools, I still think it's really important to know how to do that. And then just another caveat, that app does require that you have a whole paper ready to go. And often for conferences or sometimes grants, we don't have that. So still learning how to write a solid abstract is key.

Marguerite Matthews:

Yeah. Although I will say abstract is where I shine. I can take a lot of text and whittle it down. Cutting out fluff is one of my favorite pastimes. I love looking at personal statements for this very reason. Like, oh, you don't need all this. This is just nonsense![laughter] Take that out. Or clarify this, but give me more information, more details here. So, I found that the abstract was the least stressful part to write because, I just went for writing all of my main ideas down. Like what is this paper saying? And even if it was a really long summary, it was easier to start cutting things out. Does this still make sense if you're missing all of these sentences or you're missing a lot of extra detail? Because it will be somewhere else, you know. The abstract is not the end all be all, but it will be the difference between whether someone wants to pick it up, or someone wants to say "hey, I'm really curious what that means." Same thing with the poster, if someone is just reading in a booklet, if they're going to stop by your poster, how can I make this, like, you want to know, you want to know how I got to this place. Audience, please do not ask me to help you whittle down your abstract, those days are over![laughter] But if you were in a lab with me, you probably have had my help in getting that down to the word count.

Lauren Ullrich:

Having a good editor is key. Like having that outside perspective. Sometimes it is so nice to just be like "I'm too close to this. Can you tell me what you think are the main take home points?" And then they might tell you something and you're like "oh, that's not what I wanted at all [laughter], I need to do some editing." Or you might be like "yes, that's right, and you articulated it in a way that I'm just gonna steal and put right into my abstract."

Marguerite Matthews:

So, once the paper is written, all the co-authors, everyone has put their input in, you submit it to the journal, and then you get back your comments. Maybe you have to revise, maybe it gets rejected, and you need to submit it somewhere else, or you need to resubmit it to the same place. How do you approach that? Like, where do you start after you've cussed out all the reviewers cause they did not get your brilliance! Uh, [laughs] what happens then? And how do you communicate to say, if you're resubmitting, or you have revision, and the same people may be reviewing it, like, how do you even approach the comments and how to orient people to what changes you may have made. I think there may be someone on this podcast who could really speak to what that looks like![laughter]

Marina Picciotto:

Well, I've been on many sides of this, obviously. My papers and grants have been thoroughly critiqued. I've critiqued other people's and I've seen behind the scenes as an editor what critiques look like and what responses look like. And I think the one really important message I want to put out there is that if the reviewers didn't understand what you wrote, it's because you didn't write it so they could understand. And so rather than cursing them out for not understanding your brilliance, go back a step, and do what we've been talking about, which is to get someone who's not doing the work, who's not intimately related to it to read what you've changed, read what you wrote, and make sure that what you're trying to convey actually comes across to your reader. And the simpler, the clearer, the better. I also think that it's important to try to take-- and it's so hard-- as much emotion as possible out of the reply process. I almost never read through the whole critique the first day. I dip in, and then I go away and I sleep, and then I dip in a little more, until my heart rate goes down and I'm not gasping for breath. And then by the third or fourth time, I can read all the way through as if they were not ripping my heart out. And at that point, I know that what I read may actually be more accurate because often on that first day, the things that I'm trying to reply to are inside my own head. They didn't even ask it![mmhmm] and on day four, I'll look back at what I wrote and I'm like "wow, that is not at all what they asked." And so what I would like to sort of recommend is to take time to process what you've gotten back until you can read for accuracy rather than for pain.

Marguerite Matthews:

We've had a lot of that advice for our grantees or applicants to let them know maybe it is personal, but you can't decide that and having a snarky response to the reviewers is not going to help your grant get reviewed any better. So you can take it personal, but at some point you're going to have to like get down to what is being communicated to you about what you tried to communicate and maybe failed, and then really get down to the business of the words on the page. And I love your idea too, Marina, is having someone else read it that is very divested from the work that can say like, "yeah, I'm not really sure what you were trying to communicate here, or this doesn't make sense with what you said over here, the two aren't connecting, so maybe you need to add more, or it's not that you're scrapping the science, but you're saying okay, well, I'm not going to touch that particular area, because, like, I'm not yet able to communicate what's already out there that people really make that connection for me. And I think also being able to cut out some of your own interpretations of your work can be very hard to do because you feel like if it's not in the paper, it doesn't exist. And sometimes the scientific community isn't ready yet for your brilliance. Like you said at the very beginning where you gotta walk them to it. We're going to get there eventually, but maybe this is some data or some ideas, theories, we hold onto a little bit until we've really established ourself in this space to be like, Gotcha! Now I'm going to reel you in because I've set the stage. It's taken me maybe a couple of papers, a couple of presentations, but now here it is.

Tanya Garcia:

I totally agree with what both of you said and I also brood and get very angry at reviewers. So I'm right there with you. One thing that I found quite helpful as well is taking this as a chance to learn. And I think you said this Marguerite as well, but just looking at the reviewer comments is okay, this is my chance to get to learn more about my science. So that mindset allows me to not rush through the process and really just take a step back and think about what else can I add that I get to learn more about this and respond to these comments. That really helps me to slow down, calm down, and approach this more carefully and thoughtfully.

Lauren Ullrich:

Thank you all for sharing your wisdom today. Can I ask each of you for one last piece of parting advice for our listeners?

Tanya Garcia:

I'm a perfectionist, first off. And so, I want my first draft to be perfect. So it's taken me a long time to realize that one of the best ways for me to write is just to iterate. Just keep iterating, building on the work and not be afraid to get those critiques so that my work can be better. For me now, it's not about perfection, it's progress and just iteration to build up to that nice work that I submit. Reviewers will still not like it, but it'll get better after the revisions!

Marina Picciotto:

That's such good advice and I think to piggyback on that, I like to remind my co-workers that it's easier to edit than it is to address a blank page. Get something down on there, make it sloppy, messy, spill everything in your head and then start to work on it, edit, hone it, which is a very hard thing to do if you're a perfectionist.

Bang Wong:

I would add that we've talked about some different roles of technology in this podcast. But I would say that we should continue to embrace pencil to paper. Maybe not literally just pencil to paper, but I think that process of drawing out a schematic, drawing out a figure, is really a thinking process. And I think it's no different than writing itself has a way to expose all the things that are yet to be clear. And I think this kind of human aspect of it we should not forget, and rely on ChatGPT, AI, LLMs to write our, titles for our figures or abstract. I mean, I think that is something we should continue to embrace as long as possible.

Lauren Ullrich:

And Marguerite, what's your advice?

Marguerite Matthews:

Well, Tanya and Marina have taken on the role of Lauren in past seasons to say exactly what was in my brain. So, I was going to say "just write." And that is not advice that I have taken ever in my life, to be quite honest. So I'm going to say do as I say, not as I do! But I think my other piece of advice bringing in together all these great topics that we've touched on is to trust yourself as the scientist. Whether you are a co-author, whether you just kind of helped and you're really more just as this extra pair of eyes, or if you're taking the lead as the lead author, trust what you know. And you'll get help to whittle down, to edit, to clarify. Sometimes you have to learn that maybe how you're communicating isn't best received. And sometimes you don't know that until you get outsider's perspectives, whether that's the reviewers or that someone maybe in your department that's able to take a look. So I think trusting yourself, having a healthy dose of confidence, like "I can do this!" It may take me a little bit while to become an expert at it or to become really comfortable at it, but I am able to do this. I am a scientist. I can think critically. I've done really great work. Um, I've contributed in a really meaningful way, no matter what the size of your contribution is. And I think that's what helps push the science forward. And that pushes you also into the opportunity space, to steal a few things from Tanya. You're able to trust yourself and say "I want people to enjoy or to be as excited as I am about this work." And the only way that happens is if you go to it like "I have something good to say. I have something good to contribute. There's good work happening in the space that I'm in." What about you, Lauren? What's your advice?

Lauren Ullrich:

I think the former writing associate in me will provide some very concrete, very specific advice. Which is one, judicious use of the parking lot, so that if you have a piece that you love and you just don't want to delete it, but it doesn't fit, you just cut and paste it to a different part of the document or a different document so that you've saved your little baby and you can maybe reuse it in a different place or a different spot. But sometimes like you're just like hitting your head against this phrase that you fall in love with and it's not actually serving you well. So that's my one piece of advice. Second piece is the post outline, which I am like an evangelist for the post outline, which is after you've written your paper, you want to make sure that it actually says what you want it to say, so you make an outline from your original draft. And you just take the high level summary of every paragraph and then you have, you know, your 10 page, 20 page draft, you've turned it into a one page outline after you've already written it. And that can help you get some distance and also make sure that you're actually saying what you want to say. And that can also help you if you need to reorganize because you're working with the skeleton rather than the full sentences.

Marguerite Matthews:

I feel like you have not evangelized that to me, Lauren![Laughs] I would've liked that advice a long time ago. That's brilliant, though. Like, I love that idea. Because you're right, it gives you an opportunity to say "Am I communicating what I think I'm communicating?" Like, this outline doesn't even make sense. Two, three, and four, like, are three different papers.[mmhmm] Um, so yeah, so thank you for Season 95 of the podcast that I'm finally getting a piece of your writing genius![laughter]

Lauren Ullrich:

Alright I won't gatekeep post outlines anymore! Sorry, Marguerite.[laughter] So that's all we have time for today on Building Up the Nerve. So thank you so much to our guests this week for sharing their expertise. Thank you to Ana Ebrahimi, Mariah Hoye, Jimmy Liu, Joe Sanchez, and Tam Vo for production help. And thank you to Bob Riddle for our theme song and music. We'll see you next time when we tackle engaging with non-scientists. You can find past episodes of this podcast and many more grant application resources on the web at ninds.nih.gov.

Marguerite Matthews:

And be sure to follow us on X @NINDSDiversity. You can email us with questions at NINDSNervepod@nih.gov. And make sure you've subscribed to the podcast on Apple podcasts or your favorite podcast app. So you won't miss an episode. We'll see you next time.