Create Visuals That Have Impact
Transcript
[Michael’s presentation: Quantum Mechanic comic]
Michael Mrak:
…quantum mechanics for a quantum mechanic.
[Second slide: Black Circle]
But I am also in charge of visualizing Scientific American [coughs] and the country's oldest continuously published magazine. Some of you may wonder what this is.
[Third slide: Black Circle Labeled “Black Hole”]
And of course most of you knew that this was a black hole. [coughs] This one, of course, is not rooted in any data, but this is a good idea of how we start to think about what we depict. This dumb idea can usually lead to any other type of visualization, and [coughs] for those we do use data.
[Fourth slide: Correlations of Dynamically Measured Black-Hole Masses graphic]
This is, of course, a presentation that you might be familiar with. We do this sometimes at Scientific American, though this one appeared in Nature. But for a generalist science magazine, this isn't always the most instructive way for our readers to see the data and that it represents. And we look to find different creative ways to show the same thing.
[Fifth slide: Entangled Black Holes]
For a generalist audience, [coughs] you might want to think of data more like this, which is an interpretation of quantum entangled black holes. This is a lead image for the story, but it is rooted in data. [coughs] And later in the article we go more into depth about what is scientifically happening with this type of phenomenon.
[Sixth slide: Gargantua from Interstellar]
Or you might want to really visualize what a black hole might look like. This one was constructed for the movie Interstellar by astrophysicist Kip Thorne and a team of other physicists for director Christopher Nolan. He's a real stickler [coughs] for the science in his film, and he really likes it to be mostly accurate. But this moving image is completely rooted in data. It was modeled and crunched through a computer to make it really look like this, and it's pretty awesome too.
[Seventh slide: Black Hole at the Center of Galaxy M87]
I think the ultimate construction using data is the recent photograph of a black hole at the center of M87. Using the Event Horizon Telescope, this was constructed by combining array of smaller telescopes that were synchronized to focus on the same object at the same time around the world and act as a giant virtual telescope. Then the data was combined to form this remarkable image. I know I'm simplifying that here, and you all probably know more about this than I do, but you get the idea.
[Eighth slide: Panelists’ Names and Titles]
Today, this panel will walk you through some of the best practices and reasons [coughs] for how your data might be best visualized and displayed in the real world. First, we'll have Creative Director for Nature and Nature journal’s Kelly Krause speaking on the making of a Nature cover.
Next, Columbia's Nicoletta Bartoli, Baraloni, Barolini, excuse me, Art Director for the Office of Communication and Public Affairs, and Alissa Park, Director of the Lenfest Center for Sustainable Energy, will speak about effective data visualization for the dissemination of technology and science research.
And finally, Jennifer Christiansen, the Graphics Editor at Scientific American, will speak on developing scientific data visualizations for a non-specialist audience.
And then we'll have time for questions and answers at the end. Again, please [coughs] remember to put them into the chat window. I'm going to be reviewing them as we go and try to bring them up to the panelists as we go about the presentation.
[Ninth slide: Thank You]
Thank you, and enjoy the presentations.
[Screen share stops]
And now we're going to have Kelly Krause join us.
Kelly Krause:
Hello. Thank you so much, Michael. And thank you to Columbia for having me and giving me the opportunity to speak to you all today. I'm just going to share my screen here. There we go.
[Screen share begins]
[Kelly’s presentation: Title slide]
Okay. So today, we're going to be talking about something that's always popular, which is looking at Nature covers and what makes a good cover. There probably will be questions around how do we select covers and that kind of thing, and I'm happy to answer those, but this is going to be looking at what we look for in covers and the sort of things we're thinking about when we are creating or commissioning our covers.
[Second slide: Part 1—Introduction]
There we go.
[Third slide: The best journal covers]
So a cover will have three things for a journal cover. One of the things it must do is it must tell a story that represents the research accurately. So it must be built around data and sort of accurate science. At the same time, though, we want to layer in a bit of creativity. It's not a figure in your paper. It is something where we can be a bit more abstract and creative, but we need to make sure that there are new misleading elements. And then thirdly, I think—and perhaps obviously—it should be aesthetically pleasing, shall we say, or attractive.
[Fourth slide: Covers are a perfect blend of science and art]
It's safe to say that I think all covers should be a perfect blend of science and art. So you know, it's not just the science; it's not like a research figure. And it's not just a piece of art. It needs to communicate something, you know, in a scientific context. It needs to communicate the science, and it needs to be very high quality as well.
[Fifth slide: Covers are a bit like movie posters—Star Wars poster]
And you know I always tell authors and and people where we're trying to visualize the research into one image that covers can be thought about a bit like movie posters. And so if you think about a movie poster, you know, something that you know if you're seeing it for the first time and you haven't seen the film or you haven't seen any trailers, you know, what it is really is it's a teaser.
And it's sort of this artistic interpretation of the film in an image that gives, you know, some key elements, but it's not telling the whole story. It's not burdened with telling the whole story. You can kind of pick and choose the elements that you want to feature. And so I have a few classic movie posters with a few of our covers here.
[Sixth slide: Covers are a bit like movie posters—Indiana Jones]
Sorry, my Irish setter is getting a little aggressive. [laughs] And so, you know, we liken our covers to have a bit of drama and a bit of action if relevant.
[Seventh slide: Covers are a bit like movie posters—Batman]
Or our covers can sometimes be iconic, you know. So if we have sort of big projects or things that we want to sort of build a more iconic approach around.
[Eighth slide: Covers are a bit like movie posters—West Side Story]
Also, you know, we could use the sort of graphic devices such as typography rather than a sort of strict object-based approach to create a beautiful cover.
[Ninth slide: Part 2—Concepts: Objects and Metaphors]
And in this part I'm going to go into sort of the things that Michael was just talking about in terms of black holes. It's sort of, how do you go from sort of raw data into visualizing a concept that is aesthetically pleasing? This sort of goes into the concept of so you have objects, and then a lot of times you have sort of metaphors, so I'm going to kind of walk through that.
[Tenth slide: Some things are easier to visualize than others]
It is true that some things are much easier to visualize than others. And how, generally speaking, it goes is if your research is sort of based around an object—like in this example here we have the HIV virus, things like, you know, cells, planets, things—those things tend to be easier to draw and easier to conceptualize regardless of the medium. When you're dealing with things that are more sort of phenomena or process, such as quantumness for example, which I'm going to show here.
[Eleventh slide: However, not all research can be visually summed up by showing relevant objects]
You know, like that. That actually requires a bit more creativity. And what we tend to do is we tend to rely on metaphors. And in order for a metaphor to work, it has to be shared widely enough for people to understand what you're doing.
This paper here for this cover was about quantum cryptography. Well, you know, what does quantumness look like? Does it look like a bunch of sort of waves or colors? [laughs] Bursts of light? And what does cryptography look like? And so obviously, you know, we sort of went with the lock metaphor here but in a kind of abstract way with the faces are supposed to be Alice and Bob. So building in things that are sort of metaphorical to tell the story. This kind of thing is not as easily done, but it can be.
[Twelfth slide: We use symbols and metaphors]
This is another example of the use of metaphors. So this was a special issue on microfluidics and specifically looking at sort of– They’re very sort of practical. And so we were, you know, sort of kicking around what are some metaphors for “practical,” and the Swiss Army Knife was sort of maybe an obvious concept. But one of our in-house illustrators—you know, the red bit there is a microfluidic device—and then sort of tacked on the classic Swiss Army accoutrement, whatever. It made a nice cover. So that's another example of a metaphor.
[Thirteenth slide: We use symbols and metaphors—GMOs]
Also, you know, the drama of GMO and sort of taking some drama symbols—the Janus masks, which are often used as a metaphor for drama—making it as some nice bread. But then again you get the idea here of metaphors.
[Fourteenth slide: Some visual metaphors have become clichéd]
So this was a cover submission; this is not a Nature cover; this is a submission. So oftentimes people, when they submit Nature covers, they'll put the Nature stuff around it. [laughs] But this was a submission, and it's a question mark. And I think I just wanted to highlight this because these kinds of things come in often, that there are some metaphors that are overused. And one of them is a question mark because I think all science can probably be reduced to a question mark in some form. [laughs] All research, I should say.
So, you know, to just think about things that we tend to avoid that we get a lot are question marks, puzzle pieces, breaking something into a puzzle, jigsaw puzzle, chess. Or like we get scissors a lot in biology, you know, for sort of editing like genome editing, that kind of thing. So we try very hard to come up with more interesting ways of showing these things.
[Fifteenth slide: Part 3—From Concept to Image: Executing the Art]
Once you come up with a good concept and a good metaphor, you know, the other piece is to execute it beautifully. So these are tips here for that.
[Sixteenth slide: Covers should have]
So in order for something to be on a cover—covers are digital things, but they are still at the moment printed—and so in order for a cover to print beautifully, it has to have certain kind of technical quality. You know, you could have the most beautiful, interesting image in the world, but if it doesn't have good technical quality, we more than likely would not be able to use it.
And then the second thing is a sort of more vague [laughs] and much discussed idea of aesthetics—aesthetic value or beauty. And I think in this context, “aesthetic,” what I mean by that is just a sort of attractiveness, which is subjective but I think can more or less be shared.
[Seventeenth slide: Breaking that down into basics]
And so we're kind of further breaking that down. What do we mean by attractiveness or quality? So this brings in, like you know, classic principles of art and design like composition. So composition, meaning the sort of arrangements of elements and space within an image. Then there is also– (Sorry, let me just dismiss this). There is also things about, like, just the basic quality such as clarity, resolution, and then there's a bit about color, which I'll talk about.
[Eighteenth slide: Composition: The Rule of Thirds—Deer]
There we go. So this is something in terms of composition that some people probably have heard of called “the rule of thirds.” When you're out, let's say, photographing something, it is a very typical, almost instinctual thing to put the subject in the center of your frame. But actually, images tend to be a bit more dynamic and interesting to people, to observers, when things are not centered. So there is this sort of handy rule of thirds thing where it's a very simple grid that you can draw over a composition and try to get away from centering things.
[Nineteenth slide: Composition: The Rule of Thirds—Nature covers]
And this is just a quick example of how most of our covers probably use this. We will sometimes center things, but not very often.
[Twentieth slide: Composition: The Golden Ratio]
There's also, which you probably have heard about, the golden ratio and the idea of the golden ratio. I'm not going to go into that because it's a bit controversial.
[Twenty-first slide: Composition: The Golden Ratio—Person’s head]
Actually, some people say that the golden ratio can be applied to anything. [laughs] So anyway, just haha. But I'm not going to go into that too much.
[Twenty-second slide: Composition: Simplicity]
Something important is the idea of simplicity, particularly when we're talking about an image that's going to be used for a cover. So this was a cover submission. And on its own, it's kind of interesting. But if you think about a cover, a lot of stuff goes on a cover. You know, you have to have a sort of cover line, and the logo goes up top, and all kinds of other things that we call “page furniture” in the business.
I always recommend—because you know, speaking only on behalf of Nature and the Nature journals—but you can submit as many cover submissions as you want. And I always say strip something back as far as you can, even till it almost kind of seems bare to you, and that actually probably is the thing that will make the best cover image.
So what we did was we worked with the research group on this one, and we asked them to kind of zoom in on a space. And then we dropped part of it back and sort of made it grayscale and sort of added a bit more detail to other bits and to make it a bit more suitable for a cover, a little less complicated and a little more simple.
[Twenty-third slide: Quality: Clarity and Resolution]
Another point about clarity. Much of this actually starts in the lab or in your experimental setup or in the field. You will need some high resolution, high quality raw materials to begin with. So when you use photography or video or microscopy or however you're capturing your images, give a lot of thought to it in the beginning because then you will have some high quality material to make your covers or websites or press releases or that sort of thing.
And so in this example on the left, this is kind of a photo illustration actually, but the bit with the graphene—that block with the graphene on it—is a photo from the lab. They actually did photograph the sort of blue bits—those are microfluidic layers—but then we ended up computer rendering those. But we had like an actual thing to work from, and so that was, you know, really useful and made a nice cover. And then on the right, it's just sort of classic microscopy, but it's very high resolution, which gave us a lot to work with and we were able to put that in print.
[Twenty-fourth slide: Quality: Colour Blindness]
A bit about color and color blindness. So, you know red and green tends to be default for some imagery in some fields, but because of red green color blindness, we do not feature red and green combinations on the cover. So we just did a simple hue shift in photoshop and a little tilting of it to give it a more interesting composition in the final, which is the submission on the left, and the final version’s on the right. But that's just something to think about—color blindness.
Photoshop, if you use it, has some built-in tools that can help you determine whether or not your images are colorblind friendly or not. But also a very old kind of tried and tested thing is to just make something grayscale, and if you can't see the differences, then often than that that type. If you don't have the sort of technology, if you just have, like, a black and white printer or something, that actually gives you a lot of information about how a colorblind person might see something.
[Twenty-fifth slide: The is no “best” artistic media—Drawing]
And then I get a lot of questions about, you know, what is the best kind of artistic media. You know, what do you like? They notice, and this is true, that on Nature a lot of our covers are computer art, basically. Computer-aided arts would be from a 3D render or, you know, some other kind of computer-assisted art. And I think because Nature itself is a weekly, that sort of art is really easy to kind of modify quickly, and so that tends to be sort of for practical reasons.
But there is no best artistic media, and we accept all kinds from classic painting and drawing to the more sort of, I won't say sophisticated, but different kinds of 3D rendering and then there's also as I'll go through–
[Twenty-sixth slide: The is no “best” artistic media—Watercolour]
Yeah, so here was a recent cover that was based on a cryo-EM—so a sophisticated microscopy—but it's actually a watercolor work, which is really lovely.
[Twenty-seventh slide: The is no “best” artistic media—Digital illustration and photography]
This is a photo illustration done in photoshop, so no modeling really needed. It was just a beautiful illustration/photo mashup.
[Twenty-eighth slide: The is no “best” artistic media—3D rendering]
This sort of 3D rendering, again, is something– We use it quite a lot because it does tend to suit the ability to– You know, much of science is sort of investigating things that you can't see with a naked eye. Right? So computer 3D rendering—be it something in space or something, you know, at the level of microscopy—this sort of 3D realistic rendering gives us the feeling of going in places that we can't ourselves actually go.
[Twenty-ninth slide: The is no “best” artistic media—Bespoke modelling software]
And then sometimes when we're really lucky, we get groups that have built sort of bespoke modeling software, and they're showing interesting things in really original ways for the first time, and that is a real wonderful thing about working at Nature. It's a real privilege to be working with some of the world's best scientists on their amazing visuals.
[Thirtieth slide: Moving Image]
I'm going to end this bit by just saying that we still print it, and it's still a printed thing, but a lot of it is also consumed digitally. And we think about that when sort of commissioning our covers and working with artists. This is an example of– This is actually a still from a high-resolution animation that we commissioned as part of a package that was the cover and an animation.
[Thirty-first slide: Nature cover—150 Years of Nature]
We did the same thing for our recent 150 anniversary cover. This is actually a still from an animation rather than the other way around—building a print cover and then thinking, “How do we animate it?” So that's much more of our thinking these days.
[Thirty-second slide: Dynamical constraints on the mass…]
And it was also interactive.
[Thirty-third slide: Bonus Round—Rejections]
And then very quickly, my last few slides are just giving you a little bonus, a little look into some of the common reasons for rejections.
[Thirty-fourth slide: Erotica]
We try to keep it clean at Nature, so please send us [laughs] images without any kind of, you know, romance please.
[Thirty-fifth slide: Copyright infringement]
This happens a lot. We get a lot of kind of innocent, but it is still copyright infringement, so people will reference, like, Star Wars or Ghostbusters or this or that. But we can't use those things.
[Thirty-sixth slide: Massage scenario using mouse model]
And last slide. This was just kind of random, but really one of my favorite rejections just because it's sort of a mouse on a [laughs] massage table with all kinds of detail and stuff. It was like, you know, obviously not great for the cover but fun and did get our attention. And we did actually feature their paper on the cover, but we used a nice photograph instead.
[Thirty-seventh slide: Thank you]
So, that's all. Thank you very much.
Michael:
Thank you, Kelly. That was really fascinating.
[Kelly laughs]
[Screen share stops]
I particularly liked the last image myself, so.
Kelly:
Thank you.
Michael:
Next up, we have Alissa Park and Nicoletta Barolini. They're speaking about effective data visualization for the dissemination of technology and science research.
Nicoletta, you're muted at the moment.
Nicoletta Barolini:
Thank you. Sorry. Just one one sec here while I–
Good morning, everybody. Okay. Welcome to this talk on effective data visualization for the dissemination of technology and science research.
Visualization of data has been used throughout human history to streamline and communicate complex information as you can see from these antique illustrations. In the 21st century, thanks to so many digital tools and platforms of communication, visuals that transmit data and scientific discovery to the public are not prolific.
Alissa Park:
Niki, are you trying to move the slides?
Niki:
Am I what?
Alissa:
Where are the slides?
Niki:
Are you not seeing– Oh, oh I'm sorry. Is it not showing up?
All right, just a second. Sorry for the technical mishap. All right. There we go.
[Screen share begins]
[Niki’s presentation: Title slide]
Okay, we're in business now.
[Second slide: Six images of historical data visuals]
Okay. Visualization of data has been used throughout human history to streamline and communicate complex information as you can see from these antique illustrations. In the 21st century, thanks to so many new digital tools and platforms and communications, visuals that transmit data and scientific discovery to the public are now prolific.
[Third slide: What’s the story?]
So which graphic tells the story better? The narrative of the image on the right is easier to interpret as opposed to the one on the left. So what makes it better? Here we will address these questions and more.
[Fourth slide: Hello, I’m Alissa Park]
So I'm here with Alissa park. Alissa, would you like to introduce yourself and tell us about yourself?
Alissa:
Hi, my name is Alissa Park. I'm a faculty member in Earth and Environmental Engineering and Chemical Engineering at Columbia, and I'm also the Director of the Lenfest Center for Sustainable Energy.
And as you can imagine, we have developed engineering solutions for a number of sustainable energy and environmental problems. Particularly, my group is very interested in addressing climate change by developing technologies for carbon capture utilization and storage. Throughout this technology development, we had to talk to a wide range of audience, including the students to other researchers, which I'm more used to. But at the same time, we also had to talk to industry partners as well as the potential investors to our technology we are trying to commercialize.
So we started talking to Niki about how to effectively communicate what we are developing in our lab to show that what could be done not only impacts but what we are visualizing to have what kind of system we want to have. And that has been really for a full collaboration, and this presentation is to show that how we evolved over the time kind of trying to understand how to talk to each other to come up with the visualization. And this is not as straightforward as you’d think.
So Niki, here you are.
Niki:
Yeah, exactly.
[Fifth slide: Hello, I’m Nicoletta Barolini]
So I'm Niki. I'm the Art Director in the Office of Communications and Public Affairs. One of my tasks at Columbia involves working with researchers like Alissa who have a limited time and means to dedicate to creating art. Yet their research requires a certain amount of visual dissemination that is accessible to more than the scientific community.
[Sixth slide: Over the years, we have successfully worked together]
So over the years, Alissa and I have successfully worked together and created clear visuals of some of the innovative research in sustainability that's done at the Park Lab. And Alissa, can you tell us about the Park Lab and some of the work that you do there?
Alissa:
So this particular work shown on this slide is how to extract the metals—like, for example, copper and gold—from electronic waste. So when you think about electronic waste, I can always show the picture of my iPhone. [laughs]
That's one way to do it, but it doesn't really illustrate scientifically how our reactions and approaches alter the physical and chemical properties of these materials, which is highly heterogeneous, and how we are enhancing the extraction of these materials, which we call it “urban mining.” So concept is interesting, but we wanted to somehow show that in progress how our supercritical CO2 base, the green solvent, can be used to extract the copper and gold from these systems.
So we went through a number of iterations, and as you can see on the left, even illustrating what—for example, PC board, printed circuit board—it wasn't as easy. Because the color means something as well as the content: how much thickness of each layer, and how quantity of each material exists in a PC board. And from there, we had to decide how we're going to illustrate first and second stage reactions.
And at the end, you see that there's different colors as well as a different form factor of the products coming out. So some are solids, some are liquid. I think a lot of these visualization really help us when there's a really complex reactive system and there's multiple impacts of both physical and chemical changes, and that illustration really helps more than just words or graphs. We can definitely have graphs of what was the extent of extraction or extraction yield as a function of different solvents or time and so on, but this was one of the way for us to show that: What are the challenges? And what are the opportunities? And how we are addressing those changes.
[Seventh slide: Pictures can be more efficient than words]
Niki:
As Alissa was saying, pictures can be more efficient than words, especially when communicating complex ideas. Images are universal and can transcend barriers. They allow us to zoom in and out and to see the whole story and the details all at once. Moreover, detecting patterns is an important part of how humans learn to make decisions. Patterns within images allow our brains to know what is important and what is not.
In this example, the combined carbon capture and conversion is the focal point. And of course, it's central.
Do you want to say something about this image, Alissa? We worked on it recently.
Alissa:
I know. So I think this is the 13th version [laughs] of the graph–
Niki:
Yeah.
Alissa:
–we came up with. So if you look at here, all the colors and thickness of the arrows and everything matters. So basically, this particular technology is coming up with the capturing CO2 directly and convert it to chemicals and fuels, which is shown bottom right. So we wanted to show molecules to show that we are going from one carbon with only oxygen attached to we are adding hydrogen to it. So now you have materials with carbon and hydrogen. So there's less red dots on the bottom right. Or there is red dots, but still you see white dots as well. So we show that there's a chemical change.
But a lot of these systems—as Kelly mentioned—you sometimes have to use the graphics because you cannot see with your naked eye. So for example, that inset you see right there in the middle is our nanoscale materials—how it's facilitating capture and conversion of CO2.
And the arrows going from left to right, which is kind of red to blue, we are trying to illustrate that there is a high concentration CO2 coming in, but then you will have a CO2 lean air leaving the system. But as you can see, we also wanted to illustrate that this approach might be able to show the circular economy of carbon. That's why there's overall circle. But at the same time, we wanted to show that top red arrows are much lighter color because the concentration in that stream is a lot lower in CO2 concentration than what you're starting with at the bottom.
And we also put the renewable energy at the top, but it's not within that green box because that's not the system boundary we are looking at.
So each of the box, each of the color, each intensity of the colors means something when you try to illustrate your technology or your concept.
[Eighth slide: Current Technology vs. Columbia Engineering Technology]
Niki:
So I wanted to go over the first assignment I ever got from Alissa, which was quite a long time ago now. It was to create a graphic that depicted the process of recycling slag, which is the byproduct of steel. This is the final image. Alissa, can you talk a little bit about this graph?
Alissa:
Yeah. So this technology we developed was– If you look at the left and right, this is how we decide so that it's like a current versus future technology. So on the left hand side, or right in the middle, is how iron and steel plant are making three major products or examples of products. That doesn't change. So we still have the steel manufacturing technology, but the question is how to make them more sustainable, and that's one of the technology we develop in our lab.
So if you look at the left, you want to illustrate what is the current practice of making steel. So, what we wanted to highlight was not how steel is being made, but what are the environmental impact of those steel making. So you see that we picked the three major waste streams, in a sense.
So first waste stream is flue gas—contains high concentration of CO2, which is greenhouse gasses. Second is the waste heat, and third one is solid waste. And what we realized in our technology development was we can combine flue gas and solid waste, and both together, and then maybe we can make value-added materials while managing carbon emission.
So you can see that on the left hand side and right hand side they're mirrored in these three arrows. And there's a reason waste heat is smaller arrow because it's less of a quantity or the impact in a sense. But then the color on the left hand side and right hand side is different because left hand side, these three major streams from steel plants are considered to be waste stream. They either landfill, or they emit into the environment. So that's why we put into gray on the left hand side. That's our current practice.
But now, what our visualization or vision is that instead of looking at them as a waste stream, can we look at them as almost unconventional resources? So on the right hand side, now as you can see, we changed the color. So now CO2 is not a gray color; it's blue. And the slag is green color instead of gray again. So then by starting from same stream but then now we are viewing differently by putting the colors to them, it means something.
And then from there, we had three sets of the reactors. The colors inside the reactor is also very carefully selected. And we try to illustrate how we do the reactions. And one of the important thing is on the most top right or most right-hand side of the reactor: there are three major reaction products. The size of the arrows are designed so they illustrate what is our major product. And then color for minor element is select to be really vibrant because we really wanted to show that that's one of the area we are investigating to improve the overall sustainability of the process.
Niki:
Alissa, how did you use this graph? How did you end up using it?
Alissa:
So generally when I give academic talk or seminars at the conferences and invited lectures at the different universities, I don't usually use the graphic like this. Or only until the end, maybe, I will show as this is how we are commercializing or this is our vision. So this is more vision statement, not the data figure, so I don't usually go into this.
But this was the really nice way of showing industrial partners or somebody who really want to just get the concept very quickly. This is how I showed because with this figure itself, I can show that how reaction can be processed and how many reactors are involved and what kind of products you can make using what feedstock. And what quantity of materials may go through in these of the entire processes.
So usually when I work with somebody who's not really in the field, I usually start with this picture.
Niki:
Yeah, great.
[Ninth slide: Sketch of slag chart]
So making that slag chart went through many iterations before it was perfected. This was a new technology that I wasn't familiar with. So in order to explain the process to me, Alissa drew this sketch. So this is the first sketch that you gave me when I met you.
[Ninth slide: Zoomed-in sketch—top of sketch]
I've kept the sketch for many years cause I've always loved it. I just thought, “Wow, this is a beautiful sketch!” So in the most literal sense, a sketch is simply a set of marks laid out in a meaningful pattern like a map. So let's dissect what Alissa's sketch communicated to me.
The entry point of my gaze was at the very top since that object was centrally located and apart. Then it moved to the dotted lines around the three rectangular shapes on the right, and then to the bracket on the left and the objects within that bracket.
[Tenth slide: Zoomed-in sketch—bottom of sketch]
Finally, my gaze went down to two distinct sections at the bottom. Basically, I chunked the image into bite size sections in order to make the whole sketch easier for me to digest.
[Eleventh slide: Zoomed-in sketch—Green One Process]
Overall, the three rectangular shapes within the dashed lines are the largest and most detailed part of the whole sketch. Clearly, they were the focal point. Two rectangles contained textures while the third was blank. Textures indicated content within a liquid, depicted by the wavy line at the top. Right? That always depicts liquid. The textured contents of the first rectangle had a darker value, and they were bolder. This indicated weight or abundance and became lighter in the second rectangle.
The arrows moving from the left to the right conveyed direction, path, and process. This mechanism was going from dark to light or from more solid to less. Finally, there were arrows on top and on the bottom that indicated input or additions to the liquid and content.
[Twelfth slide: Audience, Context, Path, Actors, Clarity]
In order to clean up and refine Alissa’s sketch, I needed to consider these things: audience, context, path, actors, and clarity.
[Thirteenth slide: Audience]
So when creating visuals, it's important to know who the target audience is. Here are some questions to ask yourself. How familiar is your audience with the subject matter? Who is your audience? What is your desired outcome?
[Fourteenth slide: Sweet Spot between Simple and Complex Visuals]
So it's not an easy task for scientists to simplify or reduce complex research. Finding the sweet spot can be difficult because there is a possibility of either being too simplistic for experts or too complex for the general public.
Alissa, would you like to talk a little bit about– Is it simple for you to condense your research into something very simplistic that someone like I could understand?
Alissa:
I'm not sure. I mean, this is always a learning process, I think. What I realized, especially at Columbia, I end up interfacing or communicating with a wider range of audience, in my opinion, because even some of the scientific talks I give—this one is unique [laughs] because I don't usually do visualization talks—as well as a lot of our panel discussions and presentations on campus are open to public. And I think that's a good learning curve for us or learning experiences for us because at the end of day, science should be communicated among experts.
But in order to really have a high impact, we need to learn to communicate our science with not only public, but also policy makers. For example, a lot of environmental technology development—for example, climate change related carbon capture and storage in my group—we do want to talk to policy makers eventually or at least communicate with them so that they understand where the technology stands so that whether they can make a policy so that we can move the technology forward.
A lot of technology development is kind of pulled by three dimensions: Technology should be there advancing science and engineering, but there gotta be a market as well as policy. So I think Columbia is very unique in that even engineers have in mind or do we do really fundamental engineering and science research, we have a good communication with the other two pillars.
So for example, in this reactor system you're showing left as just a big black box versus right hand side with a lot of details. Definitely we teach our students and do research on every little bits and pieces of the right hand side. There's how to design the impeller, what dimension of this reactor has to be, how to introduce different reactors, and remove different products. There's a lot of complexity going into, and in fact even a little piece of that can be entire PhD thesis.
But today when you try to address the question what they really want to know, what are the grand challenges and how we are approaching the problem and what are the critical piece we really need to focus on. And I think that without losing the details of how we advance science, communicating that is very important and not straightforward.
Niki:
Yes.
[Fifteenth slide: Context]
So context is how content is packaged. For example, think of a book. It has a front cover and a back cover, and in between is the subject matter. Well, the same is true for picture. It has an entry point and an exit point, and in between is the important content.
[Sixteenth slide: Path]
Within the content is a narrative or a distinct path from entry to exit. Clear paths are crucial for directing the flow of information. So here are a few examples of common paths. There's linear, serpentine, convergent, circular, and radiating.
[Seventeenth slide: Example of a linear path]
Here is an example of a linear path. So as you can see, it is one straight line.
[Eighteenth slide: Example of a serpentine path]
And here is an example of a serpentine path. And it zigzags.
[Nineteenth slide: Example of a convergent path]
Here is another example from Alissa's lab—a convergent path. The arrow is converging there.
[Twentieth slide: Example of a circular path]
A circular path.
[Twenty-first slide: Example of a radiating path]
There's a radiating path. Of course you can have two paths within one, so this is a radiating and a circular path.
[Twenty-second slide: Actors]
Next is establishing the actors. Who was the story about? Who are the players that will travel along the path we have chosen? Just like real movie actors, who's the star, the co-star, the extras? Who is the villain? And what is the relationship between these actors? So the batch reactor is the star in this case, and the slag stockpile is the villain.
Is there anything you wanted to say about that? No.
[Twenty-third slide: Clarity]
Alissa:
[laughs] So should we call them villain? I'm not sure. Slags? [laughs]
Niki:
Yeah.
Alissa:
I guess in a sense the way we view things is that I was even telling my students in the class the other day that [a computer alert sounds], “Is coal a bad guy?” Well, depending on how you view it. Or do we put the gray color because I want to say that's the old story, a lot of times.
Coal made us to—or even petroleum—made us to build our modern society.
Niki:
Right.
Alissa:
So we should be thankful that we were able to find these high energy density materials. The problem was not about this material: These are just materials.
Niki:
Right.
Alissa:
The problem was how we use those materials. And I think that's why arrows and other boxes are very important. And putting them as a villain is fine, but in a sense that I would say the way we were using those materials were villains [laughs] more than just the material itself. So I think having those comments or those kind of ideas expressed well within the picture is very important.
Niki:
Yeah. No, but that's a very good point. It was the way we are using them.
Finally, clarity is achieved by not adding extra clutter to the diagram. Strong graphics are clean. This doesn't mean that the image should be stark, but rather that each element should have a distinct purpose and not be superfluous. Shapes, colors, and patterns and fonts should be used to drive the message along the path and not just to be pleasing to the eye.
[Twenty-fourth slide: Pitfalls that limit effective collaboration]
So collaboration between scientists and artists is tricky and requires clear communication.
[Twenty-fifth slide: Pitfalls]
Scientists work under tight deadlines and their budgets are capped. They often don't have the means, opportunity, or time to engage with an artist. The main thing that drives scientists is the love of discovery and sharing those discoveries with the people who can understand the intricacies.
Do you want to talk about some of the pitfalls of working to have graphics made with artists and non-scientists?
Alissa:
As I mentioned, we don't speak the same language. [laughs]
Niki:
Yeah.
Alissa:
So I think that language barrier is hard. But it's not only between scientists and visual artists. I think even when you try to do interdisciplinary research between scientists and policy makers or business school researchers or even Teachers College in education, there's a lot of difference in we don't speak the same language.
Niki:
Yeah.
Alissa:
So you need to really create something together creatively. You need to talk to each other enough length to understand each other's language—what we are emphasizing, how we are emphasizing, and how also that the visualization Niki prepared will be read by mostly engineers, for example. Then Niki needs to understand how engineers would understand the visual. So I think depending on that kind of audience is very important so she can learn from us how I would react to some of her visuals. And I think more learning we do, the process gets easier, I'm sure.
At Columbia, it was really nice because we had Niki's group to help us. I think it's getting more expensive, [laughs] so that's a challenge for us. Time, definitely, it does take more than you think because we have to go through a number of iterations. As I mentioned, one of the latest figures we use for paper just submitted took us 13 iterations. So I think start early.
And I think in my group, we also use this visualization exercise to kind of help us to solidify our vision—long-term vision or big vision—of our research itself and what is the potential impact. I think as we go through iterations, we even changed our mind on what component we want to add or what component we want to remove. Because as Niki mentioned, what's in the picture versus what is omitted—both of them—are very important, in a sense. So that decision was made kind of working with Niki, and it's a good learning experience for scientific researchers as well as a visual artist, in my opinion.
[Twenty-sixth slide: The main goal of the scientist is to discover the knowledge]
Niki:
For most scientists, art is an afterthought. The main goal of the scientists is to discover the knowledge. The dissemination is the second priority. Whereas the main goal of the artist is that of disseminating concepts, ideas, and narratives visually.
[Twenty-seventh slide: Many of the exciting discoveries are smaller steps along the way]
It's important to bring art into the process early on, as Alissa just said. Many of the exciting discoveries are smaller steps along the way that are hard to explain. Visualization of data mapping out research can lead to new hypothesis and insights for scientists themselves.
[Twenty-eighth slide: Benefits]
Not only that, but artwork can bring in additional funding. In other words, if the artwork enabled success at getting funds that wouldn't be obtained otherwise, the artwork would become the driver. Artwork needs to move from simply disseminating knowledge that is already codified to becoming part of the scientific process of model building.
[Twenty-ninth slide: Include Art at the Start!]
Finally, as with most things, the more attention and time given to its development, the better the results. So include art at the start.
[Thirtieth slide: Columbia University resources for graphics]
Here are our in-house Columbia resources. That's my email, and everybody's welcome to contact me. And there's also our team at Columbia Creative who can also lend a hand. As an external resource–
[Thirty-first slide: An external resource for creating science visuals]
There's an external resource for creating science visuals. There's an online course that's endorsed by the Medical Illustrator Association, and they are offering a special discount rate to Columbia community. So they're offering this course at $79. It was $199. And you can just log into [a computer alert sounds] their website there.
Okay. Well, thank you, Alissa. Thank you very much. And I have to say that of all the professors I've worked with, you are really one who's very keen on art and graphics, and you really promote that. And I really appreciate that about you.
Alissa:
Thank you. It's always a pleasure to work with you.
Niki:
Thank you, Alissa.
Michael:
Thank you both. That was fascinating. I mean, we work through that kind of thing every day, of course, at our publication, but it's really interesting to see how the process is there. So thank you again.
[Screen share stops]
So lastly, we have—or not lastly because she's awesome—we have Jennifer Christiansen from Scientific American. She's the graphics editor there. And she's going to be discussing how to develop scientific data visualizations for the non-specialist audience. This sort of dovetails well into Nicoletta's and Alissa's discussion.
So Jennifer.
Jennifer Christiansen:
Great, thanks so much. Let's see. I'm gonna try sharing my screen here.
[Screen share begins]
[Jen’s presentation: title slide]
And there we go. Can I have audio confirmation that you can hear me and see the slide?
Michael:
Yes.
Jen:
Excellent. Thanks so much.
Yeah, so thanks for including me this morning, and thanks to the folks listening in and watching for carving out some time to look at your screen some more today. It's really interesting to hear what other people have to say that are coming at things from slightly different angles and with slightly different purposes. But everything I've been hearing so far resonates a lot with the kinds of things that I think about and work with and do on a daily basis as well.
So we started out this morning talking a little bit more about conceptual art in the service of science with Michael's talk about kind of concept pieces and Kelly's discussion about developing covers. And then we just heard some more about creating schematics and what you think of when you think of schematic diagrams and graphics. I'm going to be speaking a little bit more solidly to data visualizations today. So we're going to take things into very kind of quantitative realm now.
[Second slide: Title slide with “data visualizations” highlighted]
So let's start here. What are data visualizations?
[Third slide: Data Visualization]
Let’s see. There we go. I'm gonna modify my screen a little bit here. There we go. So data visualization is essentially information that's chiefly numerical made visible. But that's a pretty broad definition.
[Fourth slide: Data Visualization with bar chart and dinosaur visuals]
Encompassing everything from bar charts to even dinosaur reconstructions, which are illustrations rooted in scientific measurements of things like bone length.
[Fifth slide: Data Visualization with square around abstract representations]
In general, however, folks usually think of data visualization as abstract representations of data rather than figurative illustrations, and that's what I'll be speaking to this morning.
[Sixth slide: Content]
Now for sure, data visualization is strongly rooted in content. That's the information that you're visualizing.
[Seventh slide: Content and Context]
But this morning, I'd like to encourage you to take a step back and also think about the ultimate context. For whom, and why?
[Eighth slide: Content]
The content is often fairly straightforward, especially if you're a scientist working with the data that you've collected.
[Ninth slide: Content—Data/Information]
It's the information that you're setting out to visualize.
[Tenth slide: Content with line graph]
From time series–
[Eleventh slide: Content with connected circles]
to interconnections–
[Twelfth slide: Content with bar graph]
–category comparisons, and beyond. The data set itself influences the form of your visualization.
[Thirteenth slide: Content with visuals and Context]
But I don't think that people think about context enough.
[Fourteenth slide: Content and Context definitions]
The intended audience and purpose should also be considered. Form follows function. I think it's more intuitive when people think of text. I think many people understand that there are different writing styles that are used for advertising, narrative non-fiction books, PR, scientific journals, and newspapers. You'll see different styles of writing in each of those different categories.
[Fifteenth slide: Developing scientific data visualizations for non-specialist audiences]
So this talk is divided into two sections.
[Sixteenth slide: Content and Context]
I'll provide some best practice tips and resources that relate to content. And then I'll move on to discussing the idea of context with the goal of getting you to think critically about how you can present your data to the public and engage audiences beyond your peer group. I should note that I'll be focusing on data visualizations that are developed to communicate results to an audience, not visualizations that are used as tools for analysis.
[Seventeenth slide: Organizing your content]
And rather than focus on the tools and the technical details related to building those data visualizations from scratch, I'll present strategies that should be helpful for a wide range of researchers regardless of your existing skill set. I'm operating under the assumption that some of you might rely on Excel for chart output while others might develop custom code for novel visualization forms. So for that reason, I'll focus on higher level strategies that should apply to folks across the board regardless of the tools that you use.
[Eighteenth slide: Organizing your content and customizing for a different context]
Then, I'll use a few before and after examples from Scientific American to provide you with concrete strategies for presenting your data to lay audiences, ranging from just kind of simple edits to moderate reworks to really kind of more complex and bespoke solutions.
[Nineteenth slide: Design tips]
Shortly, I'll share some references like books and blog posts that include some design best practices and tips. But first, I'd like to present one case study. So this example is courtesy of Robert Simmon. He's a data visualization engineer with Planet Labs. He wrote a blog post for the Scientific American website using an example rooted in sea surface temperature. You can read the full post at the URL shown here. I'll also have all of these URLs on my last slide so you can have a moment to catch them all if they buzz by too quickly here.
[Twentieth slide: NOAA Sea Surface Temperature Anomaly Chart]
So despite being relatively simple—this is just one variable changing over time—the data is a little bit difficult to read in this original chart by NASA’s Scientific Visualization Studio. Now, part of that might have to do with kind of their built-in defaults, so this isn't necessarily a critique of a particular studio; it's just using it as an example to show you how things can be fine-tuned and pushed further.
So Robert improved on it by creating visual hierarchy with subtle changes and typography, color, and arrangement.
[Twenty-first slide: NOAA Sea Surface Temperature Anomaly Redesigned Chart]
Here's the redesigned chart. As he wrote in that blog post, quote “A visual hierarchy brings the most important elements of a graphic into the foreground and pushes less important elements into the background.” End quote.
[Twenty-second slide: Original vs Redesign]
So, what did Robert change in order to shift attention from chart architecture to the data?
[Twenty-third slide: Original vs Redesign with removed frame]
Well, he removed the frame around the graph entirely. it flattens the graphic and kind of makes the page at the same level of importance, kind of flattens things out as I mentioned.
[Twenty-fourth slide: Original vs Redesign with adjusted color]
So he also adjusted the red, neutral, and blue colors so that they'd have equal visual weight. In the original, that white central zone is dominant and kind of glows and it really pops out. But in the revised version, the red and the blue tones don't fade into the background, and it's a little easier to read this as a continuous line as opposed to having your focus drawn to that central white band.
[Twenty-fifth slide: Original vs Redesign with removed many tick marks]
He removed intermediate tick marks and some of the labels, which made the remaining labels easier to read and reduced kind of the visual noise. Visual noise—you can think of it as clutter that doesn't really help you see the signal but just might distract from that signal.
[Twenty-sixth slide: Original vs Redesign with moved labels]
He shifted some of the labels off to the side of the chart entirely.
[Twenty-seventh slide: Original vs Redesign with URL]
Now you'll notice that these were all pretty subtle design tweaks. The chart form remains intact because a line chart is a great way to show change over time. His adjustments are all related to clarity and improved legibility.
[Twenty-eighth slide: Chart Choosers]
There are a lot of great guides to help you make chart type choices. Like I indicated in that last one, a line chart is a great way to show change over time but it's not necessarily the best way to show change differences in buckets from, like discrete categories. A bar chart might be your best choice. Again, there's lots of choosers out there that can help you kind of make these decisions.
Here's one from the journalists at the Financial Times. At the top of each column here, there's a different data relationship or goal that's described. Like the one here that I highlight speaks of the idea of flow. Below that description there are a series of recommended chart types that they suggest that you use. And this is kind of just to help with the initial brainstorming stage. This isn't necessarily hard and fast rules.
[Twenty-ninth slide: Chart Choosers—Effective Data Visualization book]
If you tend to stick to very kind of classic chart forms like line charts, bar charts, and the like, and if Excel is your primary charting tool, this book by Stephanie Evergreen is a great guide for getting you to think clearly about the point that you hope to make with your chart.
[Thirtieth slide: More Resources]
And these two titles from Alberto Cairo are particularly useful for journalists, but chapters on topics like perception science makes them useful to a much wider audience. The Functional Art also gets into kind of schematic graphics as well, but The Truthful Art and a lot of The Functional Art are focused on data visualization. But I kind of often recommend The Functional Art as your general, kind of all purpose, you want to get going, to understand some fundamentals across all types of graphics.
[Thirty-first slide: More Resources—two other books]
Now these two books go further into the types of design decisions that Robert Simmon exhibited with the sea surface chart makeover that I shared earlier—that temperature chart, excuse me. Both of these start with the basics. The Fundamentals of Data Visualization, I think, is particularly relevant for scientists.
Earlier, Alissa mentioned that sometimes the language of artists and scientists kind of can provide a block, but Claus—Claus who wrote the Fundamentals of Data Visualization—is a research scientist first, and so the language he uses and the examples he presents might feel more familiar to another research scientist.
[Thirty-second slide: More Resources—Andy Kirk’s book and website]
Andy Kirk's book and website are incredibly deep and wide resources for all things dataviz. If you don't know where to start, you can go to his site and kind of figure out what you need from there.
[Thirty-third slide: More resources: How to pick more beautiful colors]
And here's one of my favorite articles on color. It was just published a week or two ago. It's by Lisa Charlotte Rost. It walks you through common color mistakes and how to avoid them. It's a really kind of conversational and easy to understand and very kind of clear post that I think has a really wide relevance.
[Thirty-fourth slide: Customizing for a different context]
Okay. Let's move on to specific strategies for sharing your data with non-specialist audiences.
[Thirty-fifth slide: Customizing for a different context—simple edits/reframing]
I'll start with an example that simply reframes an existing chart. So perhaps you have a data visualization that was developed for a peer-reviewed article, and you're hoping to make it either more accessible to a broader audience or a quicker read for your peers in something like a poster presentation.
[Thirty-sixth slide: Reframing]
So this chart was provided as the reference material for an article on the rings of Saturn. The top chart shows how a star dimmed in a spotty manner over time. It's one of the early lines of evidence that suggested something irregular might be passing quickly in front of a distant star, dimming its apparent brightness as viewed from Earth. Other graphics in the article would get into the research that followed, but first we wanted to present the weird data that captured the scientist's attention.
[Thirty-seventh slide: Reframing—Perplexing Pattern]
So we presented the central part of the chart pretty much as is. But as a welcoming gesture to the non-specialist reader, we flagged the take home message clearly with a title to signal that this data presented scientists with an odd pattern, kind of a riddle to solve. Then we shook the jargon out of the labels and added two annotations or notes tied directly to the chart patterns as seen in the tan circles here. That would help explain clearly what people are looking at without requiring them to kind of bounce back and forth between, like, a really long caption and the image.
[Thirty-eighth slide: Reframing—Original chart and Perplexing Pattern visual]
Here's a side-by-side look. Now, I'm not trying to suggest that the bottom version should replace the top version in all cases. You'll note that we removed some pretty critical information–
[Thirty-ninth slide: Reframing—Original chart and revised one with squares]
–such as actual flux values, which are boxed here in red, and specificity with regards to time, which is boxed in orange. But because that top version exists in the primary literature, I have the freedom of stripping out some of that content here knowing that a source citation will lead a reader that needs to know more to the more complete story.
[Fortieth slide: Reframing—Monthly track directions of larger insects]
Now, many scientists use some really sophisticated and complex visualization strategies when analyzing their data. Kelly noted a few of those that actually then end up turning into covers with really kind of rich and engaging forms. And often those visualizations can cross over into being really engaging and effective communication tools as well.
Here's one example. It's the sort of visualization form that takes a lot of effort to figure out how to read, quite frankly. But the payoff is worth it because once you learn how to read it, you're kind of rewarded with the ability to discover all sorts of interesting patterns within the image.
[Forty-first slide: Reframing—Trillions of Insects Migrate]
So here's the version as it ran in our magazine. So this Scientific American rework clearly includes an aesthetic shift. There's been a change in color palettes and typography, for example. Data designer Jan Willem Tulp made a lot of subtle design changes, but the core of it is pretty darn faithful to the original. The critical change here lies in the addition of some playful insect illustrations that provide a bit more context for an otherwise really abstract set of charts. And then along with the title, those illustrations sort of help signal that these charts represent data that is related to migrating insects.
There's a fine line sometimes between adding kind of illustrative details like this and kind of tipping over into—what Edward Tufte has kind of coined the term as being—“chartjunk.” But I think there's a there's a really kind of a lovely use for them and kind of helping engage a reader who might otherwise be put off by the abstract nature of things and to help people kind of remember, you know, this in their mind's eye when they're thinking back on what they have read in the past.
[Forty-second slide: Reframing—Trillions of Insects Migrate with red square]
We also added a clear “How to Read This Graphic” panel. Kind of a welcoming gesture. It's a clear starting point for the reader. If you turn to the page and you don't know what you're looking at and you're trying to figure out where to start, we kind of clearly flag that as a “Start here. We're going to walk you through how to read this graphic in a very kind of a casual and kind of narrative way, much in the way that you might talk a friend through reading it.” And that kind of sets up the reader with the tools to then dive into the actual graphic itself.
[Forty-third slide: Reframing—Original graphic and revised one]
So here are those two versions side by side.
[Forty-fourth slide: Customizing for a different context—Moderate reworks]
Now sometimes, a bit more work is involved in customizing an existing data visualization for a broader audience. Often, information needs to be added to help the reader more fully understand what they're looking at, or the information needs to be kind of reorganized for a clearer look at the broader story being told.
[Forty-fifth slide: Rework—star brightness chart]
Here's another chart that shows an apparent change in brightness of a star as viewed from Earth. The source of this particular light pattern is Boyajian's Star, named in honor of astrophysicist Tabetha Boyajian.
[Forty-sixth slide: Rework—Enigmatic Light Patterns]
Here's the final magazine graphics box that was rooted in that chart.
[Forty-seventh slide: Rework—Zoomed-in right side of Enigmatic Light Patterns]
The right hand side is similar to some of the examples I showed earlier. Jargon has been replaced with plain descriptive language, and a few on-art labels point out the critical details.
[Forty-eighth slide: Rework—Zoomed-in left side of Enigmatic Light Patterns]
The key addition here is the diagram on the left. With this schematic, we've added some background information for the non-specialist reader, specifically an indication of why the apparent brightness of stars often changes and what that sort of pattern usually looks like.
[Forty-ninth slide: Rework—star brightness chart and Enigmatic Light Patterns]
Armed with that information, the reader can then see why the pattern on the right is unusual. Again, here's the original at the top and the magazine version on the bottom.
[Fiftieth slide: Customizing for a different context—custom solutions]
Finally, custom solutions are an option. We heard about some of those from Niki and Alissa earlier. So sometimes a visual isn't necessary for analytics purposes, but it's a handy way to communicate findings to broader audiences. Often, that can cost time and money, but not always a lot. [laughs] There's kind of our scale there.
[Fifty-first slide: Custom—Chart on author position on rosiglitazone safety]
For example, here's a table from a paper on conflicts of interest in publishing on cardiovascular risk and specific medications. So this table is perfectly suited for its purpose—a research paper in the primary literature. However, a non-specialist reader might get lost, kind of unsure of the take-home message, which numbers should they really be focusing on here, and compared to what.
[Fifty-second slide: Custom—Chart on author position on rosiglitazone safety with red box]
So in this case, a simple visualization rooted in a small portion of that table can help make the primary point clear.
[Fifty-third slide: Custom—Case Study: Conflicting Interests]
Here's the visual we ran in the magazine. Now here we show authors of scientific papers that declared a conflict of interest with a particular class of drugs in red. Authors with no financial conflict of interest are shown in blue. As it turns out, those same authors—the ones in red here with the conflict of interest—were more likely to publish favorable conclusions related to that class of drugs. You can see that in the top row here where most of those published favorable accounts or have a conflict of interest and are shown in red.
[Fifty-fourth slide: Custom—Chart and Case Study]
Here are the two items side by side. Now this chart wasn't complicated to build. Nor, frankly, is it as rich as the original table, which contains more information. But it does provide an engaging entry point for an audience that might not be willing to wade through a table, or really have the background to understand what they're looking at in that table.
[Fifty-fifth slide: Custom—Cumulative contact networks or individuals]
Not all custom projects are so straightforward. In some cases like this one, you may have a clearly identified and discrete data set, and you may have developed some serviceable visualizations for publication. But it may be worthwhile to hire a data visualizer to take things to the next level with a different audience in mind.
This study found that nurses interact with the widest variety and largest number of individuals across the hospital. Now, that's information that could provide useful in developing strategies to contain disease outbreaks within hospitals. The research scientists used visualization as a tool for analysis and published the results in this form.
[Fifty-sixth slide: Custom—Tag—You’re Sick]
If the goal is to then communicate those results to a general audience in an aesthetically pleasing and engaging manner, you can consider hiring a professional. In this case, I obtained the data from scientists and asked data designer Jan Willem Tulp to visualize that information for a lay audience. Here, each line represents a face-to-face interaction among individuals in a hospital as detected by RFID tags. Each gray node shows an individual's total number of interactions. The colorized nodes show the number of interactions with a specific group.
[Fifty-seventh slide: Custom—both graphics]
So as an art director, I hire other artists to develop solutions like this all at the time. And that option is also open to research scientists, assuming you build in the budget to do so.
[Fifty-eighth slide: How to proceed?]
So how do you proceed?
[Fifty-ninth slide: How to proceed? Build in time for many iterations]
Well first of all, build in time for many iterations, as Alissa noted. Even straightforward text-based modifications can take some time, and many of the examples I've shown were not the first attempt at that solution.
[Sixtieth slide: Rework—Profile of a paleo-orogen]
For example, for a piece on reconstructing the Sierra Nevada mountain range based on ancient rainfall data, we started with the data from the chart on the left. Data designer Tiffany Farrant-Gonzalez broke things down into separate panels so that we could walk people through the different time periods one at a time. But this was not the first attempt at a solution.
[Sixty-first slide: Variations of charts]
Here are some of the steps along the way in which Tiffany explored different ways of displaying the same data. The decision to show each time period as its own chart was made pretty early on, so we broke it down into three pretty quickly. But lots of subtle variations were explored before we honed in on the final version.
[Sixty-second slide: How to proceed? Look at your content with fresh eyes]
Perhaps the best piece of advice I can give you is to simply look at your material with fresh eyes. Let go of the jargon, and think about how you can use plain, descriptive language to approximate the message for an audience that doesn't speak your jargonese.
This goes for visual jargon as well. Does your field rely on a certain chart form because they're simply a part of the lexicon? So rather than just kind of flock to that default, think about the story that you have to tell with your data and how you can best feature that story. Look at your material through the eyes of your intended audience.
[Sixty-third slide: UNCG Research pages and chart]
For example, here's a chart that appeared in a university research magazine. It's the same version of the chart that appeared in the peer-reviewed journal, but the audiences are not the same. The university went to the trouble of using a professional photographer to feature the scientists in action, but they didn't give the graphic that same level of consideration. And that does a disservice to the scientists, their work, and your audience.
[Sixty-fourth slide: Zoom in on previous slide]
Here's a closer look. If you hope to engage the public, it's important to look at the figure with fresh eyes. Why are you including it in your article, and how can you help your readers interpret it? At its simplest, this can simply mean translating the caption at the bottom and creating smaller notes tied directly into the graphic with leader lines.
[Sixty-fifth slide: Distinct Urinary Metabolic Profile chart]
And shaking the jargon out of the labels. You simply can't present the public with access labels, such as tPS[1].
[Sixty-sixth slide: Detecting cancer through urine tests]
But I'd advocate making the time to take things to the next level. Set up a title and an introductory text block. Then deconstruct the graphic and walk people through the results. Here in this mock-up, I simply separated out the samples used to construct the model and the samples used to test the model, which was one of the points they were making with the original. The end goal is to provide an engaging, clear, and informative graphic.
And you don't need to, like, water anything down or simplify in your efforts to clarify. Just approach the source material with a fresh eye and think about how you can best guide a non-specialist reader through the content.
[Sixty-seventh slide: How to proceed? Remember to ask]
So always remember to ask, “For whom and why?” and allow those answers to inform the shape and details of your final visualizations.
[Sixty-eighth slide: URLs]
So here in the blue bar is a link to a Google sheet that I maintain that includes more resources, including the names and information about the books and posts that I had earlier. But those book and article links are also in teal below on this deck.
All right, let's see if I can stop sharing my screen.
[Screen share stops]
And hand things back over.
Michael:
Thank you, Jen. That was awesome. I have to say, I'm lucky enough to work with Jen on a daily basis. And this particular presentation is always interesting to me. You know, I do a little bit of this, but I rely on her much smarter look at these kind of things.
So we've arrived at the Q&A period of this. And if you have questions or comments, please put them into the chat box. We've gotten a few questions and a few clarifications. So I'm gonna open this up. It'd be great if Kelly, Niki, and Alissa might join us back so you can see them.
So one question is [coughs]—it’s really for everyone—is, “Can the panel talk about the notion of dumbing down a graphic?” A frequent complaint and source of problems from researchers—and we have the same problem—they don't want to dumb their ideas down. So readers can't all be expected to be specialists, of course, in any particular field, but that doesn't mean that they're dumber than you.
You folk, just think about the question for a minute. And then you might want to just jump in and see and tell us how you might address that. I think the best–
Alissa is raising your hand. Please go ahead and talk.
Alissa:
I would not call it dumbing it down because in my opinion, making your science communicated very concise manner—either verbally or written form or graphic—is harder than explaining in pages. It's always harder to do a concise manner. So I think that makes you think harder about what are your key messages. A lot of times when you're doing research— especially graduate student—you're so deep in the details, which is required, that you lose track of what is the big message you're trying to have.
I think graphics cannot pack all the information, and that's not the purpose of visualization. [A phone chimes] It's more about showing the vision or direction you're going and then start the conversation. So I think as long as you understand the purpose of the particular graphic you're trying to show, and I'm sure that there are some graphics you're trying to really illustrate molecular level interaction of something, and that may be something you can't see by photograph.
So there are different purposes, but a lot of graphics we show or we are trying to show here is that just because you have more information, it does not mean that's better communication. So it's not dumbing it down. It's a harder decision making as a communicator of science and engineering, “What is the most important thing we need to communicate?” in my opinion.
Michael:
Jen, go ahead, and you speak to that as well.
Jen:
Yeah, no, I'd love to build on that. Oops, am I mute? No I'm not, excellent. Nigel Holmes, is a designer for magazines for decades, and Alberto Cairo speak to this idea in terms of, you don't want to simplify; you should focus on clarifying. So I often, like, try to keep that mentality. It's like, what are you clearly trying to state? And as Alissa was alluding to, you kind of have to edit some things out of that, but the goal isn't to, like, simplify the process; it's to clarify and kind of understand what it is you're trying to show and show that clearly.
I like to think, also, as graphics editors as being translators. So, you know, you're going from jargon to plain language, and that happens with words and with imagery. And jargon is really useful because it sums up a lot of really specific things in one word. So sometimes using plain language, you might lose a little bit of that nuance, but probably only for other specialists. That level of nuance might not be critical to that kind of central point you're trying to make.
Niki:
I'd like to add to that that I like to chunk things up into bit pieces because I find that when I get these enormous concepts—scientific concepts—that are really hard to digest, I start to work on them in smaller ways to organize them into something that makes sense to me in a cleaner, more streamlined fashion. And I found the way to do that is chunking them into little parts and, you know, sewing them together in something bigger that's clean.
Michael:
That point—there's a comment in the box—because people worry about including enough information. Would you recommend having two figures as opposed to one figure if it helps to communicate it better rather than condensing it into one?
Jen:
I think that can be a good strategy, yeah. It's kind of, “Here's some background information, and now here's the nugget of what's new.” I do that a fair amount for articles in the magazine.
Was Kelly getting in there? No, sorry.
Kelly:
I was gonna say–
[Jen laughs]
It is a thing that you do, like, to say, Jen, it is all about context, and sort of what we're talking about. Right? So I think it's not necessarily always about the level of comprehension where someone is reading. I know that we're talking about dumbing down, which I assume that we're talking about language and education level, but also the idea of simplifying and clarifying, and I think you have to really think about your audience.
So is your audience– Is it policy makers? Because they're actually very time poor. They may be very intelligent, but they don't have a lot of time. So, you know, you may have to think about simplifying just to the very basic message that you want to get across. Or thinking about the ecosystem in which your graphic is going to live. So will it be in a printed report where someone can kind of spend time with? Will it be in a Twitter feed where it's just going to kind of roll around, where you'll need to have some different visual techniques? So it's just sort of mapping out all of these things.
I hear that a lot as well from scientists, you know. When we talk about Nature methods, as a journal, has a column called “Points of View,” which speaks specifically to scientists communicating to other scientists, so not necessarily the sort of context of a wider audience. But even in that context, I think that you get a lot of, you know, “We've spent years collecting this data. We want to use it all.” [laughs]
So I think, you know, that in a scholarly context, there's always extended data. So you know, you can put your main points in the main paper, and use extended data for that kind of bit.
Michael:
Yeah, thank you. That's great. [coughs] Kelly, while we have you on the screen, a number of questions have come in directly related to what you do.
I think that this is a particularly good question because it affects how we communicate in general. The question is, “How do you balance the connotations or multiple meanings of metaphor for readers in different geographical locations and cultures given that Nature is read by people around the world?”
Kelly:
Yeah, that's a great question, and it's something that we think a lot about and sometimes not always enough about. You know, so things like very simple metaphors. Like I showed in the presentation, there was a lock. And sort of like, well, you know you make these assumptions that are sort of “lock and key are universal.” But actually when you really dig deep into it, they're not necessarily the way they look is universal.
When we use sort of metaphors, very often we will sort of shop them around to editors around different parts. So we have editorial offices around the world, and so we will actually kind of try to get feedback. It's hard to know what you don't know, but it takes a lot of work and a lot of effort to make sure that we're actually reaching everyone and that it's not a kind of western-centric, like U.S. or, you know, Anglophile-type thing.
I think also, you know, some of our cover lines can be very sort of culture centric, sort of Anglo-centric, American culture type things, and I think we work really hard to flag those up. But you know you're always working within your own context, and so it does take extra effort to reach out and make sure that the things that we're using– It's not just kind of symbols and shapes; it's also colors. Different colors have different connotations in different cultures, and so, yeah, we try to be aware of that as we can.
Michael:
And I'd like to add on to that, because we deal with that same problem at Scientific American very often, and I think that all the communicators here have that issue. I think the biggest thing that you have to think about is to just ask the question. Right? Again, it's about thinking about the audience rather than just thinking about what it is you're producing, and to who that audience is going to be. It maybe seems simplistic, but, you know, just by stepping back and saying, “What does this mean?” is a good place to start.
So another question for you, Kelly, is– [coughs] The person asks, “I'm curious how the Nature team decides when to assign an in-house illustrator for cover design versus using submissions, and how many submissions do you usually get for any given issue?”
Kelly:
Okay, I can sort of answer [laughs] the question. So we get a lot of submissions. I wouldn't know how many we get. And this is a question that someone's probably going to ask anyway, so I'm just going to sort of digress slightly. You know, people ask how do we select what goes on the cover. For Nature, it's complicated because we're a multidisciplinary journal. We want to make sure that we are, you know, very balanced in the fields that we're showing from week to week.
Also, a lot of it comes down to just kind of scheduling. So this has particularly been made more complex over the years as some papers go online way ahead of when they go into print. And so if they go, you know, let's say a month or so before they're scheduled into a print issue, then they won't likely have the cover. So to get back to that question, I don't actually see all of the cover submissions because some of the papers are not an option for me because of scheduling. So I only look through a sort of subset.
So I would say that if—as a piece of advice, and I think we could probably at Nature do a better job of publicizing this—but if you as a researcher value very highly having your paper on the cover more so than, say, speed of publication, because different fields are have different levels of competitiveness, then flag that to your editor early in the process. Because some people are not aware of the fact that if they go on a super advanced publication schedule where they go, you know, online very quickly that they have near zero chance of a cover. So that's something that has not been asked here, but is asked of me a lot.
In terms of how we decide about commissioning illustration versus using submissions as is, it is really a mixed bag. I think there are different kinds of covers. Some of the covers go with special issues that are planned well in advance, and for those we have the luxury of time. And also we are covering a lot of different pieces and maybe a lot of different research groups that are maybe competing and are in the same issue, or maybe it's a combination of journalism and peer-reviewed research. And so we would need some other kind of third-party visual for that, so that's usually when we commission.
Whereas if we're showing a particular bit of research on the cover and the group has sent in a really lovely submission, and either it's just beautiful as is, which sometimes happens, or, you know, we can see potential in it, and we can go back to their artists and say “Can you move this here?” or “What do you think about showing that?” Or sometimes I'll look in the paper at some of the figures and I'll say, “What do you think about maybe building visualization around this?” It tends to be a bit obvious by the submission when the group is working with a really good artist, and we're able to work with them.
So it really is kind of all over the place, and also we do sometimes do our own cover. So our artists and designers will illustrate our own covers. Thank you.
Michael:
Yeah, thank you, Kelly. I appreciate that. A number of questions have come in about tools. So I know that people who are [coughs] doing that, and I will respond to another question in the chat box momentarily, but the number of people are asking about “How do you create the visuals?” So notably, do you use PowerPoint or Excel? Are there free or low cost options for people? Do people use storyboards to think these things through?
I think these are all pertinent questions. And maybe Nicoletta, you can start by speaking to that and then Jen perhaps, please.
Niki:
I'm sorry, I was answering a question online [laughs] and I apologize. Do you mind just–?
Michael:
The question is what tools you might suggest. A number of people–
Niki:
Asking for tools, yes.
Michael:
Do you suggest using PowerPoint or Excel?
Niki:
Yes, I saw those questions in the chat, and that's exactly what I was looking at. Using PowerPoint is not ideal. It really isn't ideal. And I would suggest that there's a number of free, very good courses on Illustrator that you can get online. I think even LinkedIn does a whole tutorial on it. But Illustrator would really be the best way to go, and then obviously having some Photoshop skills.
But making graphs in PowerPoint is really difficult. It's like drawing with a brick, in my opinion, because it's a very templated software. If you want to have nuances and be able to do more detailed things, Illustrator is really the easier choice. And you have layers in Illustrator, and you have ways of making masks and cutting and stuff that you don't have in PowerPoint. PowerPoint is really a place where you want to assemble all of those things for your presentation. It's not the frying pan that you would cook things; it's more the table that you would put the food on.
So I would say Illustrator and Photoshop, for sure. And again, LinkedIn and lynda.com are great resources.
Michael:
What do you think, Jen? Because you deal with this on a daily basis.
Jen:
Yeah, I mean the tools of the publishing industry tend to be the Adobe Creative Cloud tools, so I use Photoshop, Illustrator. But those are expensive packages. I don't know how they work with university licenses. So it could be worth–
Niki:
Well–
Jen:
–just looking to see. I don't know.
Niki:
You can get them monthly now. So if you have a graph that you need to do, you can register with Adobe Creative Cloud for just simply a month to use their package, but it's true they are expensive. I'm sure that there's educational discounts, probably, through Columbia. We could probably look into that and find out. But there are ways to sign on for a month if you just need to get certain graphics out and then come back to it, and I think it's something like 30 bucks.
Jen:
So for other options, I put in the chat box RAWgraphs.io is a great alternative to– Well, it's a kind of a halfway point between something like an Excel and coding things, so if you're comfortable with spreadsheets but not comfortable with coding visualizations, you can output things for download as SVGs and PNGs. It's a way to get to very sophisticated data visualization output for somebody who might not consider themselves a data visualizer or coder for that. So I recommend checking that out. It's free.
Niki:
There’s–
Jen:
If you stop using the Excel defaults—Excel defaults kind of make everybody's look the same and aren't always the best defaults—one of the books I suggested by, I think it was Story with Data, she goes through how to break away from those defaults. So I'm sorry, Niki. You were–?
Niki:
Oh, I was just going to mention that there is a new web resource. It's relatively new. It's called BioRender. I believe it's by render.com. It's a bit cookie cutter, but if you're, you know, in dire straits. It's a bit of a drag and drop. They have all sorts of icons that are already prefabricated that you can drag and drop and make, you know, a graph out of. But again, this is if you're really in dire straits and you need to get something out pretty quickly, but it's a good resource.
Jen:
So if I could just– I'm sorry
Niki:
Yes, go ahead. [laughs]
Jen:
I'm just going to add one more thing.
Niki:
Yeah.
Jen:
I almost always start with pencil and paper quite often.
Niki:
Yes, absolutely. Yes.
Jen:
And your phone has a camera in it, and so I'm often taking, you know, taking photos of sketches. I have tracing paper that I overlay–
Niki:
Yes.
Jen:
–because even though I use digital tools all of the time, it's still faster and more intuitive for me to– And I feel like I'll get less caught up in all the details. I think ideas flow a little bit more seamlessly.
Niki:
Yeah, nothing like pen and pencil. That is definitely the number one go-to. And all your ideas, you should—no matter how badly you think they look—you should just scratch them out with pen and paper. They always come out better if you start out that way.
Michael:
I would echo that too. I think that it's really important to– We're visual communicators. Right? But it's the easiest way to sort of disseminate a message. Sometimes, you know, you can even think through keywords and then start sketching around that. It's a very easy way to solve a problem.
Niki:
Sketching gets the juices going in your visualizing, and it really helps to map out and clarify in your mind what you're trying to say. And obviously when we make these sketches, they're not perfect; they're not great. They're just sketches, and that's the whole point. They're exercises to really try and get you to understand in a non-written way what you want to say, and they're very important.
Michael:
So another couple of questions [coughs] are how to do things. One question was if “I'm making a flowchart, how many words or descriptions should I include for an explanative purposes?” And somewhat related, I think, “If someone's to split one big figure into several smaller figures, in what situation would it be efficient to do that? What's the reasoning behind doing that?”
Niki:
Sorry, who–?
Michael:
[indiscernible]
Niki:
Okay. So the first question was how much text, basically? Is that–?
Michael:
Yeah. In a flowchart, about how many words might you use?
Niki:
Yeah. I would try and keep the language as concise as possible and to the point. And as I think Jen has pointed out in her presentation, the graphic should lead you around, and so the text is almost just a label or a title. So you don't want to overwhelm the image with text. Everything needs to be as clean as it possibly can. And I'm sorry, what was the other question?
Michael:
The other question is, “In what situation is it efficient to split the one big figure into several smaller figures?”
Niki:
See, that's a tricky question because I need to look at the figure. Each figure really has to be addressed individually. It's kind of hard to say that without seeing a figure.
Michael:
Yeah, I'd agree with Niki on all of these too. I think it's important that to understand, seeing whether or not it makes sense to do that. Right? Because sometimes it doesn't, and you have to see what the message is. I think Jen, you might be able to speak to that too.
Jen:
Yeah, I think I'd echo probably what you all are saying, but I also see Kelly leaning to– [Kelly laughs] Kelly, did you want to take over on that one?
Kelly:
I was just going to say a very simple rule is when you start to have overlapping data sets, when you start to, you know, on one framework have a bunch of different data sets—that's more of a dataviz thing than a sort of figurative thing—but that's where it becomes difficult for someone unacquainted with your research to sort of– It takes a really long time to figure out [laughs] what you're doing. So I think that's what I would say.
Niki:
There's a question, “Is there a distinction to be made between labels and narrative?” And going back to that first question, I would say that there is. I would say that the narrative should be within the image itself. The narrative—you know, we're talking about visual language now, so the narrative is not in written words.
The words that you have should be assisting the visual, the graphic. And the graphics should be telling the majority of the story. And again, we come back to path because the path is what the narrative is. It’s very important to when you're doing those pencil sketches in the beginning to come up with what you want your path to be, and that's where you clarify your thinking in a visual way.
Jen:
Yeah, I'd add to that by saying that yeah, I definitely agree that the form of your graphic should be echoing and reinforcing that narrative. But sometimes I find that, like, “step one,” “step two,” “step three” kinds of labels that might be more like annotations or notes are useful to help walk somebody through that narrative.
Niki:
Absolutely, yeah.
Jen:
But in that case, creating a hierarchy so that your labels that are very descriptive and just kind of like name tags for different objects are kind of at one level, whether they're like a slightly lighter font or a little bit smaller, and that your narrative-based notes kind of exist on a different plane of some sort, whether they're bold or in a circle, so that way somebody knows that they're reading like signposts versus, like, little–
Niki:
Right. The chunked narrative notes are usually, you know, off in the side or in a corner. I mean, they can be within the body of it too, but there is a distinction between the labels and a smaller explanatory function.
Michael:
So there's been a lot of discussion here about the language between scientists and engineers and art folk. You know, I'm gonna just speak to one quick thing. You know, I have fond memories about putting scientists and data visualizers together and having them talk because I find them not being very dissimilar in a lot of ways that they think.
Jen mentioned a data visualizer named Nigel Holmes earlier, and earlier in my career I got Nigel Holmes and Michio Kaku in the same room. Michio was doing a very simple– He was trying to explain relativity to a very lay audience, and we hired Nigel to do the work.
How do you guys feel about that kind of interaction? You know, the direct interaction between a data visualizer or artist and scientist?
Kelly, what do you think about that?
Kelly:
So that's pretty much what we do all the time. [laughs] Nature does obviously have some journalism where it's more akin to the work that Scientific American does where, you know, you have a sort of professional science journalist, but most of our content is either primary research or expert authored. While we may not be dealing with the author of the paper, we'll be dealing with someone who may have peer-reviewed it or someone in the circles or whatever.
So I think to me, that is sort of where things become the most fruitful because– I mean, I know we have said over the course of this event that there are sort of different languages, but I find that people treat scientists as if they're not creative or they don't, you know, get art or design or whatever, and I have found that to be not true at all.
Niki:
[laughs] I agree with that. I completely agree with that, Kelly.
Kelly:
And sometimes they have artistic references that actually are new to me or whatever. And I think also we have these sort of professional editors as translators as well, which is sort of a great thing about working for our publication is that, you know, you have more people to bounce ideas around.
And I would say that generally, no matter what you're making and whoever the audience is and whatever you're doing, bring someone in that's fresh that doesn't know what you've been working on or– Always have a second pair of eyes or a third pair of eyes and that will improve everything you're doing, whether it's a poster for a session or whether it's something for a press release or whether it's something that you'd like to have published. This is my biggest bit of advice, usually, to people.
Michael:
Jen, what do you think? I mean, you tend to do this too, I know..
Jen:
Yeah, I mean I think I don't know if I can beat Kelly on that one. [Kelly laughs] That was a good way of articulating it all. But yeah, like it's fun and surprising to see the kinds of results you can get when you have collaborations with almost anybody. But, you know, when you have a collaboration with a scientist who's very excited and passionate about the content, it's fun to see conversations and results that kind of come out of that.
Michael:
Yes. And as Catherine Mc—pardon me if I pronounce your name incorrectly—McRobie points out in the chat box that, you know, freelance scientific illustrators exist and are usually delighted to work with scientists in any field. I will echo that that is 100% true. And in fact, most artists are very scientific in a lot of ways, and they like to take those kind of conceptual ideas and help scientists and engineers to show things like this.
So we're coming down to the last seven or so minutes of our time here.
Another question that came in and I think is important to the folks who do graphics is, “Can you talk about the role of visual analogies between science concepts and more general real-life images, i.e. a motherboard equaling a subway map or stuff along those lines?”
Niki:
I'm not clear: motherboard equaling subway map?
Michael:
I think–
Niki:
Oh, I see.
Michael:
You know, the idea that, like, how do you see–
Niki:
Right. Got it. Yeah. Actually, I would say that that's something that Kelly touches on a lot with the covers—these analogies and these conceptual ideas, you know.
Kelly:
Yes.
Niki:
Yeah.
Kelly:
I'm sorry to interrupt, but it is an interesting point that like when we do things that—like let's say the microfluidic device that was a Swiss Army Knife—sometimes when we do things and it's clearly an object-based metaphor, we have to be careful that we don't make them into, like, a photorealistic kind of style that people think they exist.
Niki:
Absolutely.
Kelly:
So then we'll just choose a sort of illustrative style rather than a photorealistic kind of 3D-rendered style for certain things. Sometimes it's fine but, you know, I remember we had a paper on time crystals, and we sort of made the cover– It looked like these pieces of glass and the fog and things and with a few in-jokes with the numbers.
But, you know, we talked about, “Oh, well maybe we will animate this for, you know, social media or whatever.” But then we were like, “If we do that, it will look real and we don't want people to think that this is what they look like.” I mean, they are real, but it's like they don't actually look like this. You know? So that's a good point.
Niki:
Yeah, that's a good point.
Michael:
That point too, you know, we also did a cover about time crystals. The difference between, I think, Kelly’s publication and my publication is more nuanced because we're talking to a very educated lay audience. Part of the reasoning behind the way that we do things as opposed to Kelly's team, I think in some cases, is because we are trying to grab people's attention. When there were newsstands in pre-COVID times, I think that [Niki laughs] we are trying to grab the attention of people as they pass them and getting them to pick those things up, which is not a problem that everybody on this call will really need all the time. Again, it comes down to who your audience tends to be.
And then another question came in related to that, so I'm gonna just finish up on this point, but it’s “How do we at Scientific American choose covers and choose, like, what stories end up in things?”
Similar to Kelly's point, I think it's a little all over the place. We have an editorial board, and we schedule stories far out in time. Right? So between two and six months out. We do a lot of online stories too, but the cover stories tend to be, you know, things that we think are gonna pull in more people, are gonna relate to the subscribers of the magazine better.
And those stories are sometimes written by journalists, but about 50% of those things are written by scientists and other science-related people. If you submit something to Scientific American or if somebody comes to you from Scientific American, that may be on the cover if we find it seems like it's interesting and it'll reach a large audience for us.
Kelly, I don't know. You don't have to worry that much about that, I don’t think. Right? [coughs] But your magazine ends up in everybody's, like, scientists’ inbox all over the place.
Kelly:
[laughs]
It's true we don't have to worry about the newsstand. [laughs] So changes the process quite a bit.
Niki:
I can imagine, yeah. Well–
Michael:
[Coughs]
So I think coming close to the end of this, and I want to just say thank you to everybody who attended and certainly to the speakers. I really appreciate everybody who's been here, and to Robert Hornsby who's been behind the scenes sort of helping this whole time. I want to just say thank you.
I want everybody to know that we've been recording the full session and a video is going to be available afterwards. I think that we're going to email that to all the people who have registered to this. And we're also going to save the chat with all the links so you can see– Like there are a number of helpful links. I noticed the Adobe Suite things popped up in a couple of places and some other helpful tools.
I'd suggest if you have the time to go back and, you know, go and find the person that you think is going to really be helpful. I know Jen's presentation had a number of helpful links in it, and I think that I've used about half of them. So, you know, even if we work on a daily basis together, I usually forget these things. So these are all useful tools.
But yes, again, thank you to Jen, Kelly, Nicoletta, and Alissa for being on this panel and really being very helpful. And I hope that this has been a good help to everybody else who's been on this call.
Niki:
I'd like to say thank you to you, Michael, and I appreciate having you and Jen and Kelly. So, awesome. Thank you for coming to Columbia and speaking.
Michael:
Our pleasure.
Kelly:
Thanks for having us.
Niki:
Okay, well, goodbye. [laughs]
Unidentified attendee #1:
Thank you all.
Unidentified attendee #2:
Thank you.
Unidentified attendee #3:
Thank you.
Unidentified attendee #4:
Bye.
Michael Mrak:
…quantum mechanics for a quantum mechanic.
[Second slide: Black Circle]
But I am also in charge of visualizing Scientific American [coughs] and the country's oldest continuously published magazine. Some of you may wonder what this is.
[Third slide: Black Circle Labeled “Black Hole”]
And of course most of you knew that this was a black hole. [coughs] This one, of course, is not rooted in any data, but this is a good idea of how we start to think about what we depict. This dumb idea can usually lead to any other type of visualization, and [coughs] for those we do use data.
[Fourth slide: Correlations of Dynamically Measured Black-Hole Masses graphic]
This is, of course, a presentation that you might be familiar with. We do this sometimes at Scientific American, though this one appeared in Nature. But for a generalist science magazine, this isn't always the most instructive way for our readers to see the data and that it represents. And we look to find different creative ways to show the same thing.
[Fifth slide: Entangled Black Holes]
For a generalist audience, [coughs] you might want to think of data more like this, which is an interpretation of quantum entangled black holes. This is a lead image for the story, but it is rooted in data. [coughs] And later in the article we go more into depth about what is scientifically happening with this type of phenomenon.
[Sixth slide: Gargantua from Interstellar]
Or you might want to really visualize what a black hole might look like. This one was constructed for the movie Interstellar by astrophysicist Kip Thorne and a team of other physicists for director Christopher Nolan. He's a real stickler [coughs] for the science in his film, and he really likes it to be mostly accurate. But this moving image is completely rooted in data. It was modeled and crunched through a computer to make it really look like this, and it's pretty awesome too.
[Seventh slide: Black Hole at the Center of Galaxy M87]
I think the ultimate construction using data is the recent photograph of a black hole at the center of M87. Using the Event Horizon Telescope, this was constructed by combining array of smaller telescopes that were synchronized to focus on the same object at the same time around the world and act as a giant virtual telescope. Then the data was combined to form this remarkable image. I know I'm simplifying that here, and you all probably know more about this than I do, but you get the idea.
[Eighth slide: Panelists’ Names and Titles]
Today, this panel will walk you through some of the best practices and reasons [coughs] for how your data might be best visualized and displayed in the real world. First, we'll have Creative Director for Nature and Nature journal’s Kelly Krause speaking on the making of a Nature cover.
Next, Columbia's Nicoletta Bartoli, Baraloni, Barolini, excuse me, Art Director for the Office of Communication and Public Affairs, and Alissa Park, Director of the Lenfest Center for Sustainable Energy, will speak about effective data visualization for the dissemination of technology and science research.
And finally, Jennifer Christiansen, the Graphics Editor at Scientific American, will speak on developing scientific data visualizations for a non-specialist audience.
And then we'll have time for questions and answers at the end. Again, please [coughs] remember to put them into the chat window. I'm going to be reviewing them as we go and try to bring them up to the panelists as we go about the presentation.
[Ninth slide: Thank You]
Thank you, and enjoy the presentations.
[Screen share stops]
And now we're going to have Kelly Krause join us.
Kelly Krause:
Hello. Thank you so much, Michael. And thank you to Columbia for having me and giving me the opportunity to speak to you all today. I'm just going to share my screen here. There we go.
[Screen share begins]
[Kelly’s presentation: Title slide]
Okay. So today, we're going to be talking about something that's always popular, which is looking at Nature covers and what makes a good cover. There probably will be questions around how do we select covers and that kind of thing, and I'm happy to answer those, but this is going to be looking at what we look for in covers and the sort of things we're thinking about when we are creating or commissioning our covers.
[Second slide: Part 1—Introduction]
There we go.
[Third slide: The best journal covers]
So a cover will have three things for a journal cover. One of the things it must do is it must tell a story that represents the research accurately. So it must be built around data and sort of accurate science. At the same time, though, we want to layer in a bit of creativity. It's not a figure in your paper. It is something where we can be a bit more abstract and creative, but we need to make sure that there are new misleading elements. And then thirdly, I think—and perhaps obviously—it should be aesthetically pleasing, shall we say, or attractive.
[Fourth slide: Covers are a perfect blend of science and art]
It's safe to say that I think all covers should be a perfect blend of science and art. So you know, it's not just the science; it's not like a research figure. And it's not just a piece of art. It needs to communicate something, you know, in a scientific context. It needs to communicate the science, and it needs to be very high quality as well.
[Fifth slide: Covers are a bit like movie posters—Star Wars poster]
And you know I always tell authors and and people where we're trying to visualize the research into one image that covers can be thought about a bit like movie posters. And so if you think about a movie poster, you know, something that you know if you're seeing it for the first time and you haven't seen the film or you haven't seen any trailers, you know, what it is really is it's a teaser.
And it's sort of this artistic interpretation of the film in an image that gives, you know, some key elements, but it's not telling the whole story. It's not burdened with telling the whole story. You can kind of pick and choose the elements that you want to feature. And so I have a few classic movie posters with a few of our covers here.
[Sixth slide: Covers are a bit like movie posters—Indiana Jones]
Sorry, my Irish setter is getting a little aggressive. [laughs] And so, you know, we liken our covers to have a bit of drama and a bit of action if relevant.
[Seventh slide: Covers are a bit like movie posters—Batman]
Or our covers can sometimes be iconic, you know. So if we have sort of big projects or things that we want to sort of build a more iconic approach around.
[Eighth slide: Covers are a bit like movie posters—West Side Story]
Also, you know, we could use the sort of graphic devices such as typography rather than a sort of strict object-based approach to create a beautiful cover.
[Ninth slide: Part 2—Concepts: Objects and Metaphors]
And in this part I'm going to go into sort of the things that Michael was just talking about in terms of black holes. It's sort of, how do you go from sort of raw data into visualizing a concept that is aesthetically pleasing? This sort of goes into the concept of so you have objects, and then a lot of times you have sort of metaphors, so I'm going to kind of walk through that.
[Tenth slide: Some things are easier to visualize than others]
It is true that some things are much easier to visualize than others. And how, generally speaking, it goes is if your research is sort of based around an object—like in this example here we have the HIV virus, things like, you know, cells, planets, things—those things tend to be easier to draw and easier to conceptualize regardless of the medium. When you're dealing with things that are more sort of phenomena or process, such as quantumness for example, which I'm going to show here.
[Eleventh slide: However, not all research can be visually summed up by showing relevant objects]
You know, like that. That actually requires a bit more creativity. And what we tend to do is we tend to rely on metaphors. And in order for a metaphor to work, it has to be shared widely enough for people to understand what you're doing.
This paper here for this cover was about quantum cryptography. Well, you know, what does quantumness look like? Does it look like a bunch of sort of waves or colors? [laughs] Bursts of light? And what does cryptography look like? And so obviously, you know, we sort of went with the lock metaphor here but in a kind of abstract way with the faces are supposed to be Alice and Bob. So building in things that are sort of metaphorical to tell the story. This kind of thing is not as easily done, but it can be.
[Twelfth slide: We use symbols and metaphors]
This is another example of the use of metaphors. So this was a special issue on microfluidics and specifically looking at sort of– They’re very sort of practical. And so we were, you know, sort of kicking around what are some metaphors for “practical,” and the Swiss Army Knife was sort of maybe an obvious concept. But one of our in-house illustrators—you know, the red bit there is a microfluidic device—and then sort of tacked on the classic Swiss Army accoutrement, whatever. It made a nice cover. So that's another example of a metaphor.
[Thirteenth slide: We use symbols and metaphors—GMOs]
Also, you know, the drama of GMO and sort of taking some drama symbols—the Janus masks, which are often used as a metaphor for drama—making it as some nice bread. But then again you get the idea here of metaphors.
[Fourteenth slide: Some visual metaphors have become clichéd]
So this was a cover submission; this is not a Nature cover; this is a submission. So oftentimes people, when they submit Nature covers, they'll put the Nature stuff around it. [laughs] But this was a submission, and it's a question mark. And I think I just wanted to highlight this because these kinds of things come in often, that there are some metaphors that are overused. And one of them is a question mark because I think all science can probably be reduced to a question mark in some form. [laughs] All research, I should say.
So, you know, to just think about things that we tend to avoid that we get a lot are question marks, puzzle pieces, breaking something into a puzzle, jigsaw puzzle, chess. Or like we get scissors a lot in biology, you know, for sort of editing like genome editing, that kind of thing. So we try very hard to come up with more interesting ways of showing these things.
[Fifteenth slide: Part 3—From Concept to Image: Executing the Art]
Once you come up with a good concept and a good metaphor, you know, the other piece is to execute it beautifully. So these are tips here for that.
[Sixteenth slide: Covers should have]
So in order for something to be on a cover—covers are digital things, but they are still at the moment printed—and so in order for a cover to print beautifully, it has to have certain kind of technical quality. You know, you could have the most beautiful, interesting image in the world, but if it doesn't have good technical quality, we more than likely would not be able to use it.
And then the second thing is a sort of more vague [laughs] and much discussed idea of aesthetics—aesthetic value or beauty. And I think in this context, “aesthetic,” what I mean by that is just a sort of attractiveness, which is subjective but I think can more or less be shared.
[Seventeenth slide: Breaking that down into basics]
And so we're kind of further breaking that down. What do we mean by attractiveness or quality? So this brings in, like you know, classic principles of art and design like composition. So composition, meaning the sort of arrangements of elements and space within an image. Then there is also– (Sorry, let me just dismiss this). There is also things about, like, just the basic quality such as clarity, resolution, and then there's a bit about color, which I'll talk about.
[Eighteenth slide: Composition: The Rule of Thirds—Deer]
There we go. So this is something in terms of composition that some people probably have heard of called “the rule of thirds.” When you're out, let's say, photographing something, it is a very typical, almost instinctual thing to put the subject in the center of your frame. But actually, images tend to be a bit more dynamic and interesting to people, to observers, when things are not centered. So there is this sort of handy rule of thirds thing where it's a very simple grid that you can draw over a composition and try to get away from centering things.
[Nineteenth slide: Composition: The Rule of Thirds—Nature covers]
And this is just a quick example of how most of our covers probably use this. We will sometimes center things, but not very often.
[Twentieth slide: Composition: The Golden Ratio]
There's also, which you probably have heard about, the golden ratio and the idea of the golden ratio. I'm not going to go into that because it's a bit controversial.
[Twenty-first slide: Composition: The Golden Ratio—Person’s head]
Actually, some people say that the golden ratio can be applied to anything. [laughs] So anyway, just haha. But I'm not going to go into that too much.
[Twenty-second slide: Composition: Simplicity]
Something important is the idea of simplicity, particularly when we're talking about an image that's going to be used for a cover. So this was a cover submission. And on its own, it's kind of interesting. But if you think about a cover, a lot of stuff goes on a cover. You know, you have to have a sort of cover line, and the logo goes up top, and all kinds of other things that we call “page furniture” in the business.
I always recommend—because you know, speaking only on behalf of Nature and the Nature journals—but you can submit as many cover submissions as you want. And I always say strip something back as far as you can, even till it almost kind of seems bare to you, and that actually probably is the thing that will make the best cover image.
So what we did was we worked with the research group on this one, and we asked them to kind of zoom in on a space. And then we dropped part of it back and sort of made it grayscale and sort of added a bit more detail to other bits and to make it a bit more suitable for a cover, a little less complicated and a little more simple.
[Twenty-third slide: Quality: Clarity and Resolution]
Another point about clarity. Much of this actually starts in the lab or in your experimental setup or in the field. You will need some high resolution, high quality raw materials to begin with. So when you use photography or video or microscopy or however you're capturing your images, give a lot of thought to it in the beginning because then you will have some high quality material to make your covers or websites or press releases or that sort of thing.
And so in this example on the left, this is kind of a photo illustration actually, but the bit with the graphene—that block with the graphene on it—is a photo from the lab. They actually did photograph the sort of blue bits—those are microfluidic layers—but then we ended up computer rendering those. But we had like an actual thing to work from, and so that was, you know, really useful and made a nice cover. And then on the right, it's just sort of classic microscopy, but it's very high resolution, which gave us a lot to work with and we were able to put that in print.
[Twenty-fourth slide: Quality: Colour Blindness]
A bit about color and color blindness. So, you know red and green tends to be default for some imagery in some fields, but because of red green color blindness, we do not feature red and green combinations on the cover. So we just did a simple hue shift in photoshop and a little tilting of it to give it a more interesting composition in the final, which is the submission on the left, and the final version’s on the right. But that's just something to think about—color blindness.
Photoshop, if you use it, has some built-in tools that can help you determine whether or not your images are colorblind friendly or not. But also a very old kind of tried and tested thing is to just make something grayscale, and if you can't see the differences, then often than that that type. If you don't have the sort of technology, if you just have, like, a black and white printer or something, that actually gives you a lot of information about how a colorblind person might see something.
[Twenty-fifth slide: The is no “best” artistic media—Drawing]
And then I get a lot of questions about, you know, what is the best kind of artistic media. You know, what do you like? They notice, and this is true, that on Nature a lot of our covers are computer art, basically. Computer-aided arts would be from a 3D render or, you know, some other kind of computer-assisted art. And I think because Nature itself is a weekly, that sort of art is really easy to kind of modify quickly, and so that tends to be sort of for practical reasons.
But there is no best artistic media, and we accept all kinds from classic painting and drawing to the more sort of, I won't say sophisticated, but different kinds of 3D rendering and then there's also as I'll go through–
[Twenty-sixth slide: The is no “best” artistic media—Watercolour]
Yeah, so here was a recent cover that was based on a cryo-EM—so a sophisticated microscopy—but it's actually a watercolor work, which is really lovely.
[Twenty-seventh slide: The is no “best” artistic media—Digital illustration and photography]
This is a photo illustration done in photoshop, so no modeling really needed. It was just a beautiful illustration/photo mashup.
[Twenty-eighth slide: The is no “best” artistic media—3D rendering]
This sort of 3D rendering, again, is something– We use it quite a lot because it does tend to suit the ability to– You know, much of science is sort of investigating things that you can't see with a naked eye. Right? So computer 3D rendering—be it something in space or something, you know, at the level of microscopy—this sort of 3D realistic rendering gives us the feeling of going in places that we can't ourselves actually go.
[Twenty-ninth slide: The is no “best” artistic media—Bespoke modelling software]
And then sometimes when we're really lucky, we get groups that have built sort of bespoke modeling software, and they're showing interesting things in really original ways for the first time, and that is a real wonderful thing about working at Nature. It's a real privilege to be working with some of the world's best scientists on their amazing visuals.
[Thirtieth slide: Moving Image]
I'm going to end this bit by just saying that we still print it, and it's still a printed thing, but a lot of it is also consumed digitally. And we think about that when sort of commissioning our covers and working with artists. This is an example of– This is actually a still from a high-resolution animation that we commissioned as part of a package that was the cover and an animation.
[Thirty-first slide: Nature cover—150 Years of Nature]
We did the same thing for our recent 150 anniversary cover. This is actually a still from an animation rather than the other way around—building a print cover and then thinking, “How do we animate it?” So that's much more of our thinking these days.
[Thirty-second slide: Dynamical constraints on the mass…]
And it was also interactive.
[Thirty-third slide: Bonus Round—Rejections]
And then very quickly, my last few slides are just giving you a little bonus, a little look into some of the common reasons for rejections.
[Thirty-fourth slide: Erotica]
We try to keep it clean at Nature, so please send us [laughs] images without any kind of, you know, romance please.
[Thirty-fifth slide: Copyright infringement]
This happens a lot. We get a lot of kind of innocent, but it is still copyright infringement, so people will reference, like, Star Wars or Ghostbusters or this or that. But we can't use those things.
[Thirty-sixth slide: Massage scenario using mouse model]
And last slide. This was just kind of random, but really one of my favorite rejections just because it's sort of a mouse on a [laughs] massage table with all kinds of detail and stuff. It was like, you know, obviously not great for the cover but fun and did get our attention. And we did actually feature their paper on the cover, but we used a nice photograph instead.
[Thirty-seventh slide: Thank you]
So, that's all. Thank you very much.
Michael:
Thank you, Kelly. That was really fascinating.
[Kelly laughs]
[Screen share stops]
I particularly liked the last image myself, so.
Kelly:
Thank you.
Michael:
Next up, we have Alissa Park and Nicoletta Barolini. They're speaking about effective data visualization for the dissemination of technology and science research.
Nicoletta, you're muted at the moment.
Nicoletta Barolini:
Thank you. Sorry. Just one one sec here while I–
Good morning, everybody. Okay. Welcome to this talk on effective data visualization for the dissemination of technology and science research.
Visualization of data has been used throughout human history to streamline and communicate complex information as you can see from these antique illustrations. In the 21st century, thanks to so many digital tools and platforms of communication, visuals that transmit data and scientific discovery to the public are not prolific.
Alissa Park:
Niki, are you trying to move the slides?
Niki:
Am I what?
Alissa:
Where are the slides?
Niki:
Are you not seeing– Oh, oh I'm sorry. Is it not showing up?
All right, just a second. Sorry for the technical mishap. All right. There we go.
[Screen share begins]
[Niki’s presentation: Title slide]
Okay, we're in business now.
[Second slide: Six images of historical data visuals]
Okay. Visualization of data has been used throughout human history to streamline and communicate complex information as you can see from these antique illustrations. In the 21st century, thanks to so many new digital tools and platforms and communications, visuals that transmit data and scientific discovery to the public are now prolific.
[Third slide: What’s the story?]
So which graphic tells the story better? The narrative of the image on the right is easier to interpret as opposed to the one on the left. So what makes it better? Here we will address these questions and more.
[Fourth slide: Hello, I’m Alissa Park]
So I'm here with Alissa park. Alissa, would you like to introduce yourself and tell us about yourself?
Alissa:
Hi, my name is Alissa Park. I'm a faculty member in Earth and Environmental Engineering and Chemical Engineering at Columbia, and I'm also the Director of the Lenfest Center for Sustainable Energy.
And as you can imagine, we have developed engineering solutions for a number of sustainable energy and environmental problems. Particularly, my group is very interested in addressing climate change by developing technologies for carbon capture utilization and storage. Throughout this technology development, we had to talk to a wide range of audience, including the students to other researchers, which I'm more used to. But at the same time, we also had to talk to industry partners as well as the potential investors to our technology we are trying to commercialize.
So we started talking to Niki about how to effectively communicate what we are developing in our lab to show that what could be done not only impacts but what we are visualizing to have what kind of system we want to have. And that has been really for a full collaboration, and this presentation is to show that how we evolved over the time kind of trying to understand how to talk to each other to come up with the visualization. And this is not as straightforward as you’d think.
So Niki, here you are.
Niki:
Yeah, exactly.
[Fifth slide: Hello, I’m Nicoletta Barolini]
So I'm Niki. I'm the Art Director in the Office of Communications and Public Affairs. One of my tasks at Columbia involves working with researchers like Alissa who have a limited time and means to dedicate to creating art. Yet their research requires a certain amount of visual dissemination that is accessible to more than the scientific community.
[Sixth slide: Over the years, we have successfully worked together]
So over the years, Alissa and I have successfully worked together and created clear visuals of some of the innovative research in sustainability that's done at the Park Lab. And Alissa, can you tell us about the Park Lab and some of the work that you do there?
Alissa:
So this particular work shown on this slide is how to extract the metals—like, for example, copper and gold—from electronic waste. So when you think about electronic waste, I can always show the picture of my iPhone. [laughs]
That's one way to do it, but it doesn't really illustrate scientifically how our reactions and approaches alter the physical and chemical properties of these materials, which is highly heterogeneous, and how we are enhancing the extraction of these materials, which we call it “urban mining.” So concept is interesting, but we wanted to somehow show that in progress how our supercritical CO2 base, the green solvent, can be used to extract the copper and gold from these systems.
So we went through a number of iterations, and as you can see on the left, even illustrating what—for example, PC board, printed circuit board—it wasn't as easy. Because the color means something as well as the content: how much thickness of each layer, and how quantity of each material exists in a PC board. And from there, we had to decide how we're going to illustrate first and second stage reactions.
And at the end, you see that there's different colors as well as a different form factor of the products coming out. So some are solids, some are liquid. I think a lot of these visualization really help us when there's a really complex reactive system and there's multiple impacts of both physical and chemical changes, and that illustration really helps more than just words or graphs. We can definitely have graphs of what was the extent of extraction or extraction yield as a function of different solvents or time and so on, but this was one of the way for us to show that: What are the challenges? And what are the opportunities? And how we are addressing those changes.
[Seventh slide: Pictures can be more efficient than words]
Niki:
As Alissa was saying, pictures can be more efficient than words, especially when communicating complex ideas. Images are universal and can transcend barriers. They allow us to zoom in and out and to see the whole story and the details all at once. Moreover, detecting patterns is an important part of how humans learn to make decisions. Patterns within images allow our brains to know what is important and what is not.
In this example, the combined carbon capture and conversion is the focal point. And of course, it's central.
Do you want to say something about this image, Alissa? We worked on it recently.
Alissa:
I know. So I think this is the 13th version [laughs] of the graph–
Niki:
Yeah.
Alissa:
–we came up with. So if you look at here, all the colors and thickness of the arrows and everything matters. So basically, this particular technology is coming up with the capturing CO2 directly and convert it to chemicals and fuels, which is shown bottom right. So we wanted to show molecules to show that we are going from one carbon with only oxygen attached to we are adding hydrogen to it. So now you have materials with carbon and hydrogen. So there's less red dots on the bottom right. Or there is red dots, but still you see white dots as well. So we show that there's a chemical change.
But a lot of these systems—as Kelly mentioned—you sometimes have to use the graphics because you cannot see with your naked eye. So for example, that inset you see right there in the middle is our nanoscale materials—how it's facilitating capture and conversion of CO2.
And the arrows going from left to right, which is kind of red to blue, we are trying to illustrate that there is a high concentration CO2 coming in, but then you will have a CO2 lean air leaving the system. But as you can see, we also wanted to illustrate that this approach might be able to show the circular economy of carbon. That's why there's overall circle. But at the same time, we wanted to show that top red arrows are much lighter color because the concentration in that stream is a lot lower in CO2 concentration than what you're starting with at the bottom.
And we also put the renewable energy at the top, but it's not within that green box because that's not the system boundary we are looking at.
So each of the box, each of the color, each intensity of the colors means something when you try to illustrate your technology or your concept.
[Eighth slide: Current Technology vs. Columbia Engineering Technology]
Niki:
So I wanted to go over the first assignment I ever got from Alissa, which was quite a long time ago now. It was to create a graphic that depicted the process of recycling slag, which is the byproduct of steel. This is the final image. Alissa, can you talk a little bit about this graph?
Alissa:
Yeah. So this technology we developed was– If you look at the left and right, this is how we decide so that it's like a current versus future technology. So on the left hand side, or right in the middle, is how iron and steel plant are making three major products or examples of products. That doesn't change. So we still have the steel manufacturing technology, but the question is how to make them more sustainable, and that's one of the technology we develop in our lab.
So if you look at the left, you want to illustrate what is the current practice of making steel. So, what we wanted to highlight was not how steel is being made, but what are the environmental impact of those steel making. So you see that we picked the three major waste streams, in a sense.
So first waste stream is flue gas—contains high concentration of CO2, which is greenhouse gasses. Second is the waste heat, and third one is solid waste. And what we realized in our technology development was we can combine flue gas and solid waste, and both together, and then maybe we can make value-added materials while managing carbon emission.
So you can see that on the left hand side and right hand side they're mirrored in these three arrows. And there's a reason waste heat is smaller arrow because it's less of a quantity or the impact in a sense. But then the color on the left hand side and right hand side is different because left hand side, these three major streams from steel plants are considered to be waste stream. They either landfill, or they emit into the environment. So that's why we put into gray on the left hand side. That's our current practice.
But now, what our visualization or vision is that instead of looking at them as a waste stream, can we look at them as almost unconventional resources? So on the right hand side, now as you can see, we changed the color. So now CO2 is not a gray color; it's blue. And the slag is green color instead of gray again. So then by starting from same stream but then now we are viewing differently by putting the colors to them, it means something.
And then from there, we had three sets of the reactors. The colors inside the reactor is also very carefully selected. And we try to illustrate how we do the reactions. And one of the important thing is on the most top right or most right-hand side of the reactor: there are three major reaction products. The size of the arrows are designed so they illustrate what is our major product. And then color for minor element is select to be really vibrant because we really wanted to show that that's one of the area we are investigating to improve the overall sustainability of the process.
Niki:
Alissa, how did you use this graph? How did you end up using it?
Alissa:
So generally when I give academic talk or seminars at the conferences and invited lectures at the different universities, I don't usually use the graphic like this. Or only until the end, maybe, I will show as this is how we are commercializing or this is our vision. So this is more vision statement, not the data figure, so I don't usually go into this.
But this was the really nice way of showing industrial partners or somebody who really want to just get the concept very quickly. This is how I showed because with this figure itself, I can show that how reaction can be processed and how many reactors are involved and what kind of products you can make using what feedstock. And what quantity of materials may go through in these of the entire processes.
So usually when I work with somebody who's not really in the field, I usually start with this picture.
Niki:
Yeah, great.
[Ninth slide: Sketch of slag chart]
So making that slag chart went through many iterations before it was perfected. This was a new technology that I wasn't familiar with. So in order to explain the process to me, Alissa drew this sketch. So this is the first sketch that you gave me when I met you.
[Ninth slide: Zoomed-in sketch—top of sketch]
I've kept the sketch for many years cause I've always loved it. I just thought, “Wow, this is a beautiful sketch!” So in the most literal sense, a sketch is simply a set of marks laid out in a meaningful pattern like a map. So let's dissect what Alissa's sketch communicated to me.
The entry point of my gaze was at the very top since that object was centrally located and apart. Then it moved to the dotted lines around the three rectangular shapes on the right, and then to the bracket on the left and the objects within that bracket.
[Tenth slide: Zoomed-in sketch—bottom of sketch]
Finally, my gaze went down to two distinct sections at the bottom. Basically, I chunked the image into bite size sections in order to make the whole sketch easier for me to digest.
[Eleventh slide: Zoomed-in sketch—Green One Process]
Overall, the three rectangular shapes within the dashed lines are the largest and most detailed part of the whole sketch. Clearly, they were the focal point. Two rectangles contained textures while the third was blank. Textures indicated content within a liquid, depicted by the wavy line at the top. Right? That always depicts liquid. The textured contents of the first rectangle had a darker value, and they were bolder. This indicated weight or abundance and became lighter in the second rectangle.
The arrows moving from the left to the right conveyed direction, path, and process. This mechanism was going from dark to light or from more solid to less. Finally, there were arrows on top and on the bottom that indicated input or additions to the liquid and content.
[Twelfth slide: Audience, Context, Path, Actors, Clarity]
In order to clean up and refine Alissa’s sketch, I needed to consider these things: audience, context, path, actors, and clarity.
[Thirteenth slide: Audience]
So when creating visuals, it's important to know who the target audience is. Here are some questions to ask yourself. How familiar is your audience with the subject matter? Who is your audience? What is your desired outcome?
[Fourteenth slide: Sweet Spot between Simple and Complex Visuals]
So it's not an easy task for scientists to simplify or reduce complex research. Finding the sweet spot can be difficult because there is a possibility of either being too simplistic for experts or too complex for the general public.
Alissa, would you like to talk a little bit about– Is it simple for you to condense your research into something very simplistic that someone like I could understand?
Alissa:
I'm not sure. I mean, this is always a learning process, I think. What I realized, especially at Columbia, I end up interfacing or communicating with a wider range of audience, in my opinion, because even some of the scientific talks I give—this one is unique [laughs] because I don't usually do visualization talks—as well as a lot of our panel discussions and presentations on campus are open to public. And I think that's a good learning curve for us or learning experiences for us because at the end of day, science should be communicated among experts.
But in order to really have a high impact, we need to learn to communicate our science with not only public, but also policy makers. For example, a lot of environmental technology development—for example, climate change related carbon capture and storage in my group—we do want to talk to policy makers eventually or at least communicate with them so that they understand where the technology stands so that whether they can make a policy so that we can move the technology forward.
A lot of technology development is kind of pulled by three dimensions: Technology should be there advancing science and engineering, but there gotta be a market as well as policy. So I think Columbia is very unique in that even engineers have in mind or do we do really fundamental engineering and science research, we have a good communication with the other two pillars.
So for example, in this reactor system you're showing left as just a big black box versus right hand side with a lot of details. Definitely we teach our students and do research on every little bits and pieces of the right hand side. There's how to design the impeller, what dimension of this reactor has to be, how to introduce different reactors, and remove different products. There's a lot of complexity going into, and in fact even a little piece of that can be entire PhD thesis.
But today when you try to address the question what they really want to know, what are the grand challenges and how we are approaching the problem and what are the critical piece we really need to focus on. And I think that without losing the details of how we advance science, communicating that is very important and not straightforward.
Niki:
Yes.
[Fifteenth slide: Context]
So context is how content is packaged. For example, think of a book. It has a front cover and a back cover, and in between is the subject matter. Well, the same is true for picture. It has an entry point and an exit point, and in between is the important content.
[Sixteenth slide: Path]
Within the content is a narrative or a distinct path from entry to exit. Clear paths are crucial for directing the flow of information. So here are a few examples of common paths. There's linear, serpentine, convergent, circular, and radiating.
[Seventeenth slide: Example of a linear path]
Here is an example of a linear path. So as you can see, it is one straight line.
[Eighteenth slide: Example of a serpentine path]
And here is an example of a serpentine path. And it zigzags.
[Nineteenth slide: Example of a convergent path]
Here is another example from Alissa's lab—a convergent path. The arrow is converging there.
[Twentieth slide: Example of a circular path]
A circular path.
[Twenty-first slide: Example of a radiating path]
There's a radiating path. Of course you can have two paths within one, so this is a radiating and a circular path.
[Twenty-second slide: Actors]
Next is establishing the actors. Who was the story about? Who are the players that will travel along the path we have chosen? Just like real movie actors, who's the star, the co-star, the extras? Who is the villain? And what is the relationship between these actors? So the batch reactor is the star in this case, and the slag stockpile is the villain.
Is there anything you wanted to say about that? No.
[Twenty-third slide: Clarity]
Alissa:
[laughs] So should we call them villain? I'm not sure. Slags? [laughs]
Niki:
Yeah.
Alissa:
I guess in a sense the way we view things is that I was even telling my students in the class the other day that [a computer alert sounds], “Is coal a bad guy?” Well, depending on how you view it. Or do we put the gray color because I want to say that's the old story, a lot of times.
Coal made us to—or even petroleum—made us to build our modern society.
Niki:
Right.
Alissa:
So we should be thankful that we were able to find these high energy density materials. The problem was not about this material: These are just materials.
Niki:
Right.
Alissa:
The problem was how we use those materials. And I think that's why arrows and other boxes are very important. And putting them as a villain is fine, but in a sense that I would say the way we were using those materials were villains [laughs] more than just the material itself. So I think having those comments or those kind of ideas expressed well within the picture is very important.
Niki:
Yeah. No, but that's a very good point. It was the way we are using them.
Finally, clarity is achieved by not adding extra clutter to the diagram. Strong graphics are clean. This doesn't mean that the image should be stark, but rather that each element should have a distinct purpose and not be superfluous. Shapes, colors, and patterns and fonts should be used to drive the message along the path and not just to be pleasing to the eye.
[Twenty-fourth slide: Pitfalls that limit effective collaboration]
So collaboration between scientists and artists is tricky and requires clear communication.
[Twenty-fifth slide: Pitfalls]
Scientists work under tight deadlines and their budgets are capped. They often don't have the means, opportunity, or time to engage with an artist. The main thing that drives scientists is the love of discovery and sharing those discoveries with the people who can understand the intricacies.
Do you want to talk about some of the pitfalls of working to have graphics made with artists and non-scientists?
Alissa:
As I mentioned, we don't speak the same language. [laughs]
Niki:
Yeah.
Alissa:
So I think that language barrier is hard. But it's not only between scientists and visual artists. I think even when you try to do interdisciplinary research between scientists and policy makers or business school researchers or even Teachers College in education, there's a lot of difference in we don't speak the same language.
Niki:
Yeah.
Alissa:
So you need to really create something together creatively. You need to talk to each other enough length to understand each other's language—what we are emphasizing, how we are emphasizing, and how also that the visualization Niki prepared will be read by mostly engineers, for example. Then Niki needs to understand how engineers would understand the visual. So I think depending on that kind of audience is very important so she can learn from us how I would react to some of her visuals. And I think more learning we do, the process gets easier, I'm sure.
At Columbia, it was really nice because we had Niki's group to help us. I think it's getting more expensive, [laughs] so that's a challenge for us. Time, definitely, it does take more than you think because we have to go through a number of iterations. As I mentioned, one of the latest figures we use for paper just submitted took us 13 iterations. So I think start early.
And I think in my group, we also use this visualization exercise to kind of help us to solidify our vision—long-term vision or big vision—of our research itself and what is the potential impact. I think as we go through iterations, we even changed our mind on what component we want to add or what component we want to remove. Because as Niki mentioned, what's in the picture versus what is omitted—both of them—are very important, in a sense. So that decision was made kind of working with Niki, and it's a good learning experience for scientific researchers as well as a visual artist, in my opinion.
[Twenty-sixth slide: The main goal of the scientist is to discover the knowledge]
Niki:
For most scientists, art is an afterthought. The main goal of the scientists is to discover the knowledge. The dissemination is the second priority. Whereas the main goal of the artist is that of disseminating concepts, ideas, and narratives visually.
[Twenty-seventh slide: Many of the exciting discoveries are smaller steps along the way]
It's important to bring art into the process early on, as Alissa just said. Many of the exciting discoveries are smaller steps along the way that are hard to explain. Visualization of data mapping out research can lead to new hypothesis and insights for scientists themselves.
[Twenty-eighth slide: Benefits]
Not only that, but artwork can bring in additional funding. In other words, if the artwork enabled success at getting funds that wouldn't be obtained otherwise, the artwork would become the driver. Artwork needs to move from simply disseminating knowledge that is already codified to becoming part of the scientific process of model building.
[Twenty-ninth slide: Include Art at the Start!]
Finally, as with most things, the more attention and time given to its development, the better the results. So include art at the start.
[Thirtieth slide: Columbia University resources for graphics]
Here are our in-house Columbia resources. That's my email, and everybody's welcome to contact me. And there's also our team at Columbia Creative who can also lend a hand. As an external resource–
[Thirty-first slide: An external resource for creating science visuals]
There's an external resource for creating science visuals. There's an online course that's endorsed by the Medical Illustrator Association, and they are offering a special discount rate to Columbia community. So they're offering this course at $79. It was $199. And you can just log into [a computer alert sounds] their website there.
Okay. Well, thank you, Alissa. Thank you very much. And I have to say that of all the professors I've worked with, you are really one who's very keen on art and graphics, and you really promote that. And I really appreciate that about you.
Alissa:
Thank you. It's always a pleasure to work with you.
Niki:
Thank you, Alissa.
Michael:
Thank you both. That was fascinating. I mean, we work through that kind of thing every day, of course, at our publication, but it's really interesting to see how the process is there. So thank you again.
[Screen share stops]
So lastly, we have—or not lastly because she's awesome—we have Jennifer Christiansen from Scientific American. She's the graphics editor there. And she's going to be discussing how to develop scientific data visualizations for the non-specialist audience. This sort of dovetails well into Nicoletta's and Alissa's discussion.
So Jennifer.
Jennifer Christiansen:
Great, thanks so much. Let's see. I'm gonna try sharing my screen here.
[Screen share begins]
[Jen’s presentation: title slide]
And there we go. Can I have audio confirmation that you can hear me and see the slide?
Michael:
Yes.
Jen:
Excellent. Thanks so much.
Yeah, so thanks for including me this morning, and thanks to the folks listening in and watching for carving out some time to look at your screen some more today. It's really interesting to hear what other people have to say that are coming at things from slightly different angles and with slightly different purposes. But everything I've been hearing so far resonates a lot with the kinds of things that I think about and work with and do on a daily basis as well.
So we started out this morning talking a little bit more about conceptual art in the service of science with Michael's talk about kind of concept pieces and Kelly's discussion about developing covers. And then we just heard some more about creating schematics and what you think of when you think of schematic diagrams and graphics. I'm going to be speaking a little bit more solidly to data visualizations today. So we're going to take things into very kind of quantitative realm now.
[Second slide: Title slide with “data visualizations” highlighted]
So let's start here. What are data visualizations?
[Third slide: Data Visualization]
Let’s see. There we go. I'm gonna modify my screen a little bit here. There we go. So data visualization is essentially information that's chiefly numerical made visible. But that's a pretty broad definition.
[Fourth slide: Data Visualization with bar chart and dinosaur visuals]
Encompassing everything from bar charts to even dinosaur reconstructions, which are illustrations rooted in scientific measurements of things like bone length.
[Fifth slide: Data Visualization with square around abstract representations]
In general, however, folks usually think of data visualization as abstract representations of data rather than figurative illustrations, and that's what I'll be speaking to this morning.
[Sixth slide: Content]
Now for sure, data visualization is strongly rooted in content. That's the information that you're visualizing.
[Seventh slide: Content and Context]
But this morning, I'd like to encourage you to take a step back and also think about the ultimate context. For whom, and why?
[Eighth slide: Content]
The content is often fairly straightforward, especially if you're a scientist working with the data that you've collected.
[Ninth slide: Content—Data/Information]
It's the information that you're setting out to visualize.
[Tenth slide: Content with line graph]
From time series–
[Eleventh slide: Content with connected circles]
to interconnections–
[Twelfth slide: Content with bar graph]
–category comparisons, and beyond. The data set itself influences the form of your visualization.
[Thirteenth slide: Content with visuals and Context]
But I don't think that people think about context enough.
[Fourteenth slide: Content and Context definitions]
The intended audience and purpose should also be considered. Form follows function. I think it's more intuitive when people think of text. I think many people understand that there are different writing styles that are used for advertising, narrative non-fiction books, PR, scientific journals, and newspapers. You'll see different styles of writing in each of those different categories.
[Fifteenth slide: Developing scientific data visualizations for non-specialist audiences]
So this talk is divided into two sections.
[Sixteenth slide: Content and Context]
I'll provide some best practice tips and resources that relate to content. And then I'll move on to discussing the idea of context with the goal of getting you to think critically about how you can present your data to the public and engage audiences beyond your peer group. I should note that I'll be focusing on data visualizations that are developed to communicate results to an audience, not visualizations that are used as tools for analysis.
[Seventeenth slide: Organizing your content]
And rather than focus on the tools and the technical details related to building those data visualizations from scratch, I'll present strategies that should be helpful for a wide range of researchers regardless of your existing skill set. I'm operating under the assumption that some of you might rely on Excel for chart output while others might develop custom code for novel visualization forms. So for that reason, I'll focus on higher level strategies that should apply to folks across the board regardless of the tools that you use.
[Eighteenth slide: Organizing your content and customizing for a different context]
Then, I'll use a few before and after examples from Scientific American to provide you with concrete strategies for presenting your data to lay audiences, ranging from just kind of simple edits to moderate reworks to really kind of more complex and bespoke solutions.
[Nineteenth slide: Design tips]
Shortly, I'll share some references like books and blog posts that include some design best practices and tips. But first, I'd like to present one case study. So this example is courtesy of Robert Simmon. He's a data visualization engineer with Planet Labs. He wrote a blog post for the Scientific American website using an example rooted in sea surface temperature. You can read the full post at the URL shown here. I'll also have all of these URLs on my last slide so you can have a moment to catch them all if they buzz by too quickly here.
[Twentieth slide: NOAA Sea Surface Temperature Anomaly Chart]
So despite being relatively simple—this is just one variable changing over time—the data is a little bit difficult to read in this original chart by NASA’s Scientific Visualization Studio. Now, part of that might have to do with kind of their built-in defaults, so this isn't necessarily a critique of a particular studio; it's just using it as an example to show you how things can be fine-tuned and pushed further.
So Robert improved on it by creating visual hierarchy with subtle changes and typography, color, and arrangement.
[Twenty-first slide: NOAA Sea Surface Temperature Anomaly Redesigned Chart]
Here's the redesigned chart. As he wrote in that blog post, quote “A visual hierarchy brings the most important elements of a graphic into the foreground and pushes less important elements into the background.” End quote.
[Twenty-second slide: Original vs Redesign]
So, what did Robert change in order to shift attention from chart architecture to the data?
[Twenty-third slide: Original vs Redesign with removed frame]
Well, he removed the frame around the graph entirely. it flattens the graphic and kind of makes the page at the same level of importance, kind of flattens things out as I mentioned.
[Twenty-fourth slide: Original vs Redesign with adjusted color]
So he also adjusted the red, neutral, and blue colors so that they'd have equal visual weight. In the original, that white central zone is dominant and kind of glows and it really pops out. But in the revised version, the red and the blue tones don't fade into the background, and it's a little easier to read this as a continuous line as opposed to having your focus drawn to that central white band.
[Twenty-fifth slide: Original vs Redesign with removed many tick marks]
He removed intermediate tick marks and some of the labels, which made the remaining labels easier to read and reduced kind of the visual noise. Visual noise—you can think of it as clutter that doesn't really help you see the signal but just might distract from that signal.
[Twenty-sixth slide: Original vs Redesign with moved labels]
He shifted some of the labels off to the side of the chart entirely.
[Twenty-seventh slide: Original vs Redesign with URL]
Now you'll notice that these were all pretty subtle design tweaks. The chart form remains intact because a line chart is a great way to show change over time. His adjustments are all related to clarity and improved legibility.
[Twenty-eighth slide: Chart Choosers]
There are a lot of great guides to help you make chart type choices. Like I indicated in that last one, a line chart is a great way to show change over time but it's not necessarily the best way to show change differences in buckets from, like discrete categories. A bar chart might be your best choice. Again, there's lots of choosers out there that can help you kind of make these decisions.
Here's one from the journalists at the Financial Times. At the top of each column here, there's a different data relationship or goal that's described. Like the one here that I highlight speaks of the idea of flow. Below that description there are a series of recommended chart types that they suggest that you use. And this is kind of just to help with the initial brainstorming stage. This isn't necessarily hard and fast rules.
[Twenty-ninth slide: Chart Choosers—Effective Data Visualization book]
If you tend to stick to very kind of classic chart forms like line charts, bar charts, and the like, and if Excel is your primary charting tool, this book by Stephanie Evergreen is a great guide for getting you to think clearly about the point that you hope to make with your chart.
[Thirtieth slide: More Resources]
And these two titles from Alberto Cairo are particularly useful for journalists, but chapters on topics like perception science makes them useful to a much wider audience. The Functional Art also gets into kind of schematic graphics as well, but The Truthful Art and a lot of The Functional Art are focused on data visualization. But I kind of often recommend The Functional Art as your general, kind of all purpose, you want to get going, to understand some fundamentals across all types of graphics.
[Thirty-first slide: More Resources—two other books]
Now these two books go further into the types of design decisions that Robert Simmon exhibited with the sea surface chart makeover that I shared earlier—that temperature chart, excuse me. Both of these start with the basics. The Fundamentals of Data Visualization, I think, is particularly relevant for scientists.
Earlier, Alissa mentioned that sometimes the language of artists and scientists kind of can provide a block, but Claus—Claus who wrote the Fundamentals of Data Visualization—is a research scientist first, and so the language he uses and the examples he presents might feel more familiar to another research scientist.
[Thirty-second slide: More Resources—Andy Kirk’s book and website]
Andy Kirk's book and website are incredibly deep and wide resources for all things dataviz. If you don't know where to start, you can go to his site and kind of figure out what you need from there.
[Thirty-third slide: More resources: How to pick more beautiful colors]
And here's one of my favorite articles on color. It was just published a week or two ago. It's by Lisa Charlotte Rost. It walks you through common color mistakes and how to avoid them. It's a really kind of conversational and easy to understand and very kind of clear post that I think has a really wide relevance.
[Thirty-fourth slide: Customizing for a different context]
Okay. Let's move on to specific strategies for sharing your data with non-specialist audiences.
[Thirty-fifth slide: Customizing for a different context—simple edits/reframing]
I'll start with an example that simply reframes an existing chart. So perhaps you have a data visualization that was developed for a peer-reviewed article, and you're hoping to make it either more accessible to a broader audience or a quicker read for your peers in something like a poster presentation.
[Thirty-sixth slide: Reframing]
So this chart was provided as the reference material for an article on the rings of Saturn. The top chart shows how a star dimmed in a spotty manner over time. It's one of the early lines of evidence that suggested something irregular might be passing quickly in front of a distant star, dimming its apparent brightness as viewed from Earth. Other graphics in the article would get into the research that followed, but first we wanted to present the weird data that captured the scientist's attention.
[Thirty-seventh slide: Reframing—Perplexing Pattern]
So we presented the central part of the chart pretty much as is. But as a welcoming gesture to the non-specialist reader, we flagged the take home message clearly with a title to signal that this data presented scientists with an odd pattern, kind of a riddle to solve. Then we shook the jargon out of the labels and added two annotations or notes tied directly to the chart patterns as seen in the tan circles here. That would help explain clearly what people are looking at without requiring them to kind of bounce back and forth between, like, a really long caption and the image.
[Thirty-eighth slide: Reframing—Original chart and Perplexing Pattern visual]
Here's a side-by-side look. Now, I'm not trying to suggest that the bottom version should replace the top version in all cases. You'll note that we removed some pretty critical information–
[Thirty-ninth slide: Reframing—Original chart and revised one with squares]
–such as actual flux values, which are boxed here in red, and specificity with regards to time, which is boxed in orange. But because that top version exists in the primary literature, I have the freedom of stripping out some of that content here knowing that a source citation will lead a reader that needs to know more to the more complete story.
[Fortieth slide: Reframing—Monthly track directions of larger insects]
Now, many scientists use some really sophisticated and complex visualization strategies when analyzing their data. Kelly noted a few of those that actually then end up turning into covers with really kind of rich and engaging forms. And often those visualizations can cross over into being really engaging and effective communication tools as well.
Here's one example. It's the sort of visualization form that takes a lot of effort to figure out how to read, quite frankly. But the payoff is worth it because once you learn how to read it, you're kind of rewarded with the ability to discover all sorts of interesting patterns within the image.
[Forty-first slide: Reframing—Trillions of Insects Migrate]
So here's the version as it ran in our magazine. So this Scientific American rework clearly includes an aesthetic shift. There's been a change in color palettes and typography, for example. Data designer Jan Willem Tulp made a lot of subtle design changes, but the core of it is pretty darn faithful to the original. The critical change here lies in the addition of some playful insect illustrations that provide a bit more context for an otherwise really abstract set of charts. And then along with the title, those illustrations sort of help signal that these charts represent data that is related to migrating insects.
There's a fine line sometimes between adding kind of illustrative details like this and kind of tipping over into—what Edward Tufte has kind of coined the term as being—“chartjunk.” But I think there's a there's a really kind of a lovely use for them and kind of helping engage a reader who might otherwise be put off by the abstract nature of things and to help people kind of remember, you know, this in their mind's eye when they're thinking back on what they have read in the past.
[Forty-second slide: Reframing—Trillions of Insects Migrate with red square]
We also added a clear “How to Read This Graphic” panel. Kind of a welcoming gesture. It's a clear starting point for the reader. If you turn to the page and you don't know what you're looking at and you're trying to figure out where to start, we kind of clearly flag that as a “Start here. We're going to walk you through how to read this graphic in a very kind of a casual and kind of narrative way, much in the way that you might talk a friend through reading it.” And that kind of sets up the reader with the tools to then dive into the actual graphic itself.
[Forty-third slide: Reframing—Original graphic and revised one]
So here are those two versions side by side.
[Forty-fourth slide: Customizing for a different context—Moderate reworks]
Now sometimes, a bit more work is involved in customizing an existing data visualization for a broader audience. Often, information needs to be added to help the reader more fully understand what they're looking at, or the information needs to be kind of reorganized for a clearer look at the broader story being told.
[Forty-fifth slide: Rework—star brightness chart]
Here's another chart that shows an apparent change in brightness of a star as viewed from Earth. The source of this particular light pattern is Boyajian's Star, named in honor of astrophysicist Tabetha Boyajian.
[Forty-sixth slide: Rework—Enigmatic Light Patterns]
Here's the final magazine graphics box that was rooted in that chart.
[Forty-seventh slide: Rework—Zoomed-in right side of Enigmatic Light Patterns]
The right hand side is similar to some of the examples I showed earlier. Jargon has been replaced with plain descriptive language, and a few on-art labels point out the critical details.
[Forty-eighth slide: Rework—Zoomed-in left side of Enigmatic Light Patterns]
The key addition here is the diagram on the left. With this schematic, we've added some background information for the non-specialist reader, specifically an indication of why the apparent brightness of stars often changes and what that sort of pattern usually looks like.
[Forty-ninth slide: Rework—star brightness chart and Enigmatic Light Patterns]
Armed with that information, the reader can then see why the pattern on the right is unusual. Again, here's the original at the top and the magazine version on the bottom.
[Fiftieth slide: Customizing for a different context—custom solutions]
Finally, custom solutions are an option. We heard about some of those from Niki and Alissa earlier. So sometimes a visual isn't necessary for analytics purposes, but it's a handy way to communicate findings to broader audiences. Often, that can cost time and money, but not always a lot. [laughs] There's kind of our scale there.
[Fifty-first slide: Custom—Chart on author position on rosiglitazone safety]
For example, here's a table from a paper on conflicts of interest in publishing on cardiovascular risk and specific medications. So this table is perfectly suited for its purpose—a research paper in the primary literature. However, a non-specialist reader might get lost, kind of unsure of the take-home message, which numbers should they really be focusing on here, and compared to what.
[Fifty-second slide: Custom—Chart on author position on rosiglitazone safety with red box]
So in this case, a simple visualization rooted in a small portion of that table can help make the primary point clear.
[Fifty-third slide: Custom—Case Study: Conflicting Interests]
Here's the visual we ran in the magazine. Now here we show authors of scientific papers that declared a conflict of interest with a particular class of drugs in red. Authors with no financial conflict of interest are shown in blue. As it turns out, those same authors—the ones in red here with the conflict of interest—were more likely to publish favorable conclusions related to that class of drugs. You can see that in the top row here where most of those published favorable accounts or have a conflict of interest and are shown in red.
[Fifty-fourth slide: Custom—Chart and Case Study]
Here are the two items side by side. Now this chart wasn't complicated to build. Nor, frankly, is it as rich as the original table, which contains more information. But it does provide an engaging entry point for an audience that might not be willing to wade through a table, or really have the background to understand what they're looking at in that table.
[Fifty-fifth slide: Custom—Cumulative contact networks or individuals]
Not all custom projects are so straightforward. In some cases like this one, you may have a clearly identified and discrete data set, and you may have developed some serviceable visualizations for publication. But it may be worthwhile to hire a data visualizer to take things to the next level with a different audience in mind.
This study found that nurses interact with the widest variety and largest number of individuals across the hospital. Now, that's information that could provide useful in developing strategies to contain disease outbreaks within hospitals. The research scientists used visualization as a tool for analysis and published the results in this form.
[Fifty-sixth slide: Custom—Tag—You’re Sick]
If the goal is to then communicate those results to a general audience in an aesthetically pleasing and engaging manner, you can consider hiring a professional. In this case, I obtained the data from scientists and asked data designer Jan Willem Tulp to visualize that information for a lay audience. Here, each line represents a face-to-face interaction among individuals in a hospital as detected by RFID tags. Each gray node shows an individual's total number of interactions. The colorized nodes show the number of interactions with a specific group.
[Fifty-seventh slide: Custom—both graphics]
So as an art director, I hire other artists to develop solutions like this all at the time. And that option is also open to research scientists, assuming you build in the budget to do so.
[Fifty-eighth slide: How to proceed?]
So how do you proceed?
[Fifty-ninth slide: How to proceed? Build in time for many iterations]
Well first of all, build in time for many iterations, as Alissa noted. Even straightforward text-based modifications can take some time, and many of the examples I've shown were not the first attempt at that solution.
[Sixtieth slide: Rework—Profile of a paleo-orogen]
For example, for a piece on reconstructing the Sierra Nevada mountain range based on ancient rainfall data, we started with the data from the chart on the left. Data designer Tiffany Farrant-Gonzalez broke things down into separate panels so that we could walk people through the different time periods one at a time. But this was not the first attempt at a solution.
[Sixty-first slide: Variations of charts]
Here are some of the steps along the way in which Tiffany explored different ways of displaying the same data. The decision to show each time period as its own chart was made pretty early on, so we broke it down into three pretty quickly. But lots of subtle variations were explored before we honed in on the final version.
[Sixty-second slide: How to proceed? Look at your content with fresh eyes]
Perhaps the best piece of advice I can give you is to simply look at your material with fresh eyes. Let go of the jargon, and think about how you can use plain, descriptive language to approximate the message for an audience that doesn't speak your jargonese.
This goes for visual jargon as well. Does your field rely on a certain chart form because they're simply a part of the lexicon? So rather than just kind of flock to that default, think about the story that you have to tell with your data and how you can best feature that story. Look at your material through the eyes of your intended audience.
[Sixty-third slide: UNCG Research pages and chart]
For example, here's a chart that appeared in a university research magazine. It's the same version of the chart that appeared in the peer-reviewed journal, but the audiences are not the same. The university went to the trouble of using a professional photographer to feature the scientists in action, but they didn't give the graphic that same level of consideration. And that does a disservice to the scientists, their work, and your audience.
[Sixty-fourth slide: Zoom in on previous slide]
Here's a closer look. If you hope to engage the public, it's important to look at the figure with fresh eyes. Why are you including it in your article, and how can you help your readers interpret it? At its simplest, this can simply mean translating the caption at the bottom and creating smaller notes tied directly into the graphic with leader lines.
[Sixty-fifth slide: Distinct Urinary Metabolic Profile chart]
And shaking the jargon out of the labels. You simply can't present the public with access labels, such as tPS[1].
[Sixty-sixth slide: Detecting cancer through urine tests]
But I'd advocate making the time to take things to the next level. Set up a title and an introductory text block. Then deconstruct the graphic and walk people through the results. Here in this mock-up, I simply separated out the samples used to construct the model and the samples used to test the model, which was one of the points they were making with the original. The end goal is to provide an engaging, clear, and informative graphic.
And you don't need to, like, water anything down or simplify in your efforts to clarify. Just approach the source material with a fresh eye and think about how you can best guide a non-specialist reader through the content.
[Sixty-seventh slide: How to proceed? Remember to ask]
So always remember to ask, “For whom and why?” and allow those answers to inform the shape and details of your final visualizations.
[Sixty-eighth slide: URLs]
So here in the blue bar is a link to a Google sheet that I maintain that includes more resources, including the names and information about the books and posts that I had earlier. But those book and article links are also in teal below on this deck.
All right, let's see if I can stop sharing my screen.
[Screen share stops]
And hand things back over.
Michael:
Thank you, Jen. That was awesome. I have to say, I'm lucky enough to work with Jen on a daily basis. And this particular presentation is always interesting to me. You know, I do a little bit of this, but I rely on her much smarter look at these kind of things.
So we've arrived at the Q&A period of this. And if you have questions or comments, please put them into the chat box. We've gotten a few questions and a few clarifications. So I'm gonna open this up. It'd be great if Kelly, Niki, and Alissa might join us back so you can see them.
So one question is [coughs]—it’s really for everyone—is, “Can the panel talk about the notion of dumbing down a graphic?” A frequent complaint and source of problems from researchers—and we have the same problem—they don't want to dumb their ideas down. So readers can't all be expected to be specialists, of course, in any particular field, but that doesn't mean that they're dumber than you.
You folk, just think about the question for a minute. And then you might want to just jump in and see and tell us how you might address that. I think the best–
Alissa is raising your hand. Please go ahead and talk.
Alissa:
I would not call it dumbing it down because in my opinion, making your science communicated very concise manner—either verbally or written form or graphic—is harder than explaining in pages. It's always harder to do a concise manner. So I think that makes you think harder about what are your key messages. A lot of times when you're doing research— especially graduate student—you're so deep in the details, which is required, that you lose track of what is the big message you're trying to have.
I think graphics cannot pack all the information, and that's not the purpose of visualization. [A phone chimes] It's more about showing the vision or direction you're going and then start the conversation. So I think as long as you understand the purpose of the particular graphic you're trying to show, and I'm sure that there are some graphics you're trying to really illustrate molecular level interaction of something, and that may be something you can't see by photograph.
So there are different purposes, but a lot of graphics we show or we are trying to show here is that just because you have more information, it does not mean that's better communication. So it's not dumbing it down. It's a harder decision making as a communicator of science and engineering, “What is the most important thing we need to communicate?” in my opinion.
Michael:
Jen, go ahead, and you speak to that as well.
Jen:
Yeah, no, I'd love to build on that. Oops, am I mute? No I'm not, excellent. Nigel Holmes, is a designer for magazines for decades, and Alberto Cairo speak to this idea in terms of, you don't want to simplify; you should focus on clarifying. So I often, like, try to keep that mentality. It's like, what are you clearly trying to state? And as Alissa was alluding to, you kind of have to edit some things out of that, but the goal isn't to, like, simplify the process; it's to clarify and kind of understand what it is you're trying to show and show that clearly.
I like to think, also, as graphics editors as being translators. So, you know, you're going from jargon to plain language, and that happens with words and with imagery. And jargon is really useful because it sums up a lot of really specific things in one word. So sometimes using plain language, you might lose a little bit of that nuance, but probably only for other specialists. That level of nuance might not be critical to that kind of central point you're trying to make.
Niki:
I'd like to add to that that I like to chunk things up into bit pieces because I find that when I get these enormous concepts—scientific concepts—that are really hard to digest, I start to work on them in smaller ways to organize them into something that makes sense to me in a cleaner, more streamlined fashion. And I found the way to do that is chunking them into little parts and, you know, sewing them together in something bigger that's clean.
Michael:
That point—there's a comment in the box—because people worry about including enough information. Would you recommend having two figures as opposed to one figure if it helps to communicate it better rather than condensing it into one?
Jen:
I think that can be a good strategy, yeah. It's kind of, “Here's some background information, and now here's the nugget of what's new.” I do that a fair amount for articles in the magazine.
Was Kelly getting in there? No, sorry.
Kelly:
I was gonna say–
[Jen laughs]
It is a thing that you do, like, to say, Jen, it is all about context, and sort of what we're talking about. Right? So I think it's not necessarily always about the level of comprehension where someone is reading. I know that we're talking about dumbing down, which I assume that we're talking about language and education level, but also the idea of simplifying and clarifying, and I think you have to really think about your audience.
So is your audience– Is it policy makers? Because they're actually very time poor. They may be very intelligent, but they don't have a lot of time. So, you know, you may have to think about simplifying just to the very basic message that you want to get across. Or thinking about the ecosystem in which your graphic is going to live. So will it be in a printed report where someone can kind of spend time with? Will it be in a Twitter feed where it's just going to kind of roll around, where you'll need to have some different visual techniques? So it's just sort of mapping out all of these things.
I hear that a lot as well from scientists, you know. When we talk about Nature methods, as a journal, has a column called “Points of View,” which speaks specifically to scientists communicating to other scientists, so not necessarily the sort of context of a wider audience. But even in that context, I think that you get a lot of, you know, “We've spent years collecting this data. We want to use it all.” [laughs]
So I think, you know, that in a scholarly context, there's always extended data. So you know, you can put your main points in the main paper, and use extended data for that kind of bit.
Michael:
Yeah, thank you. That's great. [coughs] Kelly, while we have you on the screen, a number of questions have come in directly related to what you do.
I think that this is a particularly good question because it affects how we communicate in general. The question is, “How do you balance the connotations or multiple meanings of metaphor for readers in different geographical locations and cultures given that Nature is read by people around the world?”
Kelly:
Yeah, that's a great question, and it's something that we think a lot about and sometimes not always enough about. You know, so things like very simple metaphors. Like I showed in the presentation, there was a lock. And sort of like, well, you know you make these assumptions that are sort of “lock and key are universal.” But actually when you really dig deep into it, they're not necessarily the way they look is universal.
When we use sort of metaphors, very often we will sort of shop them around to editors around different parts. So we have editorial offices around the world, and so we will actually kind of try to get feedback. It's hard to know what you don't know, but it takes a lot of work and a lot of effort to make sure that we're actually reaching everyone and that it's not a kind of western-centric, like U.S. or, you know, Anglophile-type thing.
I think also, you know, some of our cover lines can be very sort of culture centric, sort of Anglo-centric, American culture type things, and I think we work really hard to flag those up. But you know you're always working within your own context, and so it does take extra effort to reach out and make sure that the things that we're using– It's not just kind of symbols and shapes; it's also colors. Different colors have different connotations in different cultures, and so, yeah, we try to be aware of that as we can.
Michael:
And I'd like to add on to that, because we deal with that same problem at Scientific American very often, and I think that all the communicators here have that issue. I think the biggest thing that you have to think about is to just ask the question. Right? Again, it's about thinking about the audience rather than just thinking about what it is you're producing, and to who that audience is going to be. It maybe seems simplistic, but, you know, just by stepping back and saying, “What does this mean?” is a good place to start.
So another question for you, Kelly, is– [coughs] The person asks, “I'm curious how the Nature team decides when to assign an in-house illustrator for cover design versus using submissions, and how many submissions do you usually get for any given issue?”
Kelly:
Okay, I can sort of answer [laughs] the question. So we get a lot of submissions. I wouldn't know how many we get. And this is a question that someone's probably going to ask anyway, so I'm just going to sort of digress slightly. You know, people ask how do we select what goes on the cover. For Nature, it's complicated because we're a multidisciplinary journal. We want to make sure that we are, you know, very balanced in the fields that we're showing from week to week.
Also, a lot of it comes down to just kind of scheduling. So this has particularly been made more complex over the years as some papers go online way ahead of when they go into print. And so if they go, you know, let's say a month or so before they're scheduled into a print issue, then they won't likely have the cover. So to get back to that question, I don't actually see all of the cover submissions because some of the papers are not an option for me because of scheduling. So I only look through a sort of subset.
So I would say that if—as a piece of advice, and I think we could probably at Nature do a better job of publicizing this—but if you as a researcher value very highly having your paper on the cover more so than, say, speed of publication, because different fields are have different levels of competitiveness, then flag that to your editor early in the process. Because some people are not aware of the fact that if they go on a super advanced publication schedule where they go, you know, online very quickly that they have near zero chance of a cover. So that's something that has not been asked here, but is asked of me a lot.
In terms of how we decide about commissioning illustration versus using submissions as is, it is really a mixed bag. I think there are different kinds of covers. Some of the covers go with special issues that are planned well in advance, and for those we have the luxury of time. And also we are covering a lot of different pieces and maybe a lot of different research groups that are maybe competing and are in the same issue, or maybe it's a combination of journalism and peer-reviewed research. And so we would need some other kind of third-party visual for that, so that's usually when we commission.
Whereas if we're showing a particular bit of research on the cover and the group has sent in a really lovely submission, and either it's just beautiful as is, which sometimes happens, or, you know, we can see potential in it, and we can go back to their artists and say “Can you move this here?” or “What do you think about showing that?” Or sometimes I'll look in the paper at some of the figures and I'll say, “What do you think about maybe building visualization around this?” It tends to be a bit obvious by the submission when the group is working with a really good artist, and we're able to work with them.
So it really is kind of all over the place, and also we do sometimes do our own cover. So our artists and designers will illustrate our own covers. Thank you.
Michael:
Yeah, thank you, Kelly. I appreciate that. A number of questions have come in about tools. So I know that people who are [coughs] doing that, and I will respond to another question in the chat box momentarily, but the number of people are asking about “How do you create the visuals?” So notably, do you use PowerPoint or Excel? Are there free or low cost options for people? Do people use storyboards to think these things through?
I think these are all pertinent questions. And maybe Nicoletta, you can start by speaking to that and then Jen perhaps, please.
Niki:
I'm sorry, I was answering a question online [laughs] and I apologize. Do you mind just–?
Michael:
The question is what tools you might suggest. A number of people–
Niki:
Asking for tools, yes.
Michael:
Do you suggest using PowerPoint or Excel?
Niki:
Yes, I saw those questions in the chat, and that's exactly what I was looking at. Using PowerPoint is not ideal. It really isn't ideal. And I would suggest that there's a number of free, very good courses on Illustrator that you can get online. I think even LinkedIn does a whole tutorial on it. But Illustrator would really be the best way to go, and then obviously having some Photoshop skills.
But making graphs in PowerPoint is really difficult. It's like drawing with a brick, in my opinion, because it's a very templated software. If you want to have nuances and be able to do more detailed things, Illustrator is really the easier choice. And you have layers in Illustrator, and you have ways of making masks and cutting and stuff that you don't have in PowerPoint. PowerPoint is really a place where you want to assemble all of those things for your presentation. It's not the frying pan that you would cook things; it's more the table that you would put the food on.
So I would say Illustrator and Photoshop, for sure. And again, LinkedIn and lynda.com are great resources.
Michael:
What do you think, Jen? Because you deal with this on a daily basis.
Jen:
Yeah, I mean the tools of the publishing industry tend to be the Adobe Creative Cloud tools, so I use Photoshop, Illustrator. But those are expensive packages. I don't know how they work with university licenses. So it could be worth–
Niki:
Well–
Jen:
–just looking to see. I don't know.
Niki:
You can get them monthly now. So if you have a graph that you need to do, you can register with Adobe Creative Cloud for just simply a month to use their package, but it's true they are expensive. I'm sure that there's educational discounts, probably, through Columbia. We could probably look into that and find out. But there are ways to sign on for a month if you just need to get certain graphics out and then come back to it, and I think it's something like 30 bucks.
Jen:
So for other options, I put in the chat box RAWgraphs.io is a great alternative to– Well, it's a kind of a halfway point between something like an Excel and coding things, so if you're comfortable with spreadsheets but not comfortable with coding visualizations, you can output things for download as SVGs and PNGs. It's a way to get to very sophisticated data visualization output for somebody who might not consider themselves a data visualizer or coder for that. So I recommend checking that out. It's free.
Niki:
There’s–
Jen:
If you stop using the Excel defaults—Excel defaults kind of make everybody's look the same and aren't always the best defaults—one of the books I suggested by, I think it was Story with Data, she goes through how to break away from those defaults. So I'm sorry, Niki. You were–?
Niki:
Oh, I was just going to mention that there is a new web resource. It's relatively new. It's called BioRender. I believe it's by render.com. It's a bit cookie cutter, but if you're, you know, in dire straits. It's a bit of a drag and drop. They have all sorts of icons that are already prefabricated that you can drag and drop and make, you know, a graph out of. But again, this is if you're really in dire straits and you need to get something out pretty quickly, but it's a good resource.
Jen:
So if I could just– I'm sorry
Niki:
Yes, go ahead. [laughs]
Jen:
I'm just going to add one more thing.
Niki:
Yeah.
Jen:
I almost always start with pencil and paper quite often.
Niki:
Yes, absolutely. Yes.
Jen:
And your phone has a camera in it, and so I'm often taking, you know, taking photos of sketches. I have tracing paper that I overlay–
Niki:
Yes.
Jen:
–because even though I use digital tools all of the time, it's still faster and more intuitive for me to– And I feel like I'll get less caught up in all the details. I think ideas flow a little bit more seamlessly.
Niki:
Yeah, nothing like pen and pencil. That is definitely the number one go-to. And all your ideas, you should—no matter how badly you think they look—you should just scratch them out with pen and paper. They always come out better if you start out that way.
Michael:
I would echo that too. I think that it's really important to– We're visual communicators. Right? But it's the easiest way to sort of disseminate a message. Sometimes, you know, you can even think through keywords and then start sketching around that. It's a very easy way to solve a problem.
Niki:
Sketching gets the juices going in your visualizing, and it really helps to map out and clarify in your mind what you're trying to say. And obviously when we make these sketches, they're not perfect; they're not great. They're just sketches, and that's the whole point. They're exercises to really try and get you to understand in a non-written way what you want to say, and they're very important.
Michael:
So another couple of questions [coughs] are how to do things. One question was if “I'm making a flowchart, how many words or descriptions should I include for an explanative purposes?” And somewhat related, I think, “If someone's to split one big figure into several smaller figures, in what situation would it be efficient to do that? What's the reasoning behind doing that?”
Niki:
Sorry, who–?
Michael:
[indiscernible]
Niki:
Okay. So the first question was how much text, basically? Is that–?
Michael:
Yeah. In a flowchart, about how many words might you use?
Niki:
Yeah. I would try and keep the language as concise as possible and to the point. And as I think Jen has pointed out in her presentation, the graphic should lead you around, and so the text is almost just a label or a title. So you don't want to overwhelm the image with text. Everything needs to be as clean as it possibly can. And I'm sorry, what was the other question?
Michael:
The other question is, “In what situation is it efficient to split the one big figure into several smaller figures?”
Niki:
See, that's a tricky question because I need to look at the figure. Each figure really has to be addressed individually. It's kind of hard to say that without seeing a figure.
Michael:
Yeah, I'd agree with Niki on all of these too. I think it's important that to understand, seeing whether or not it makes sense to do that. Right? Because sometimes it doesn't, and you have to see what the message is. I think Jen, you might be able to speak to that too.
Jen:
Yeah, I think I'd echo probably what you all are saying, but I also see Kelly leaning to– [Kelly laughs] Kelly, did you want to take over on that one?
Kelly:
I was just going to say a very simple rule is when you start to have overlapping data sets, when you start to, you know, on one framework have a bunch of different data sets—that's more of a dataviz thing than a sort of figurative thing—but that's where it becomes difficult for someone unacquainted with your research to sort of– It takes a really long time to figure out [laughs] what you're doing. So I think that's what I would say.
Niki:
There's a question, “Is there a distinction to be made between labels and narrative?” And going back to that first question, I would say that there is. I would say that the narrative should be within the image itself. The narrative—you know, we're talking about visual language now, so the narrative is not in written words.
The words that you have should be assisting the visual, the graphic. And the graphics should be telling the majority of the story. And again, we come back to path because the path is what the narrative is. It’s very important to when you're doing those pencil sketches in the beginning to come up with what you want your path to be, and that's where you clarify your thinking in a visual way.
Jen:
Yeah, I'd add to that by saying that yeah, I definitely agree that the form of your graphic should be echoing and reinforcing that narrative. But sometimes I find that, like, “step one,” “step two,” “step three” kinds of labels that might be more like annotations or notes are useful to help walk somebody through that narrative.
Niki:
Absolutely, yeah.
Jen:
But in that case, creating a hierarchy so that your labels that are very descriptive and just kind of like name tags for different objects are kind of at one level, whether they're like a slightly lighter font or a little bit smaller, and that your narrative-based notes kind of exist on a different plane of some sort, whether they're bold or in a circle, so that way somebody knows that they're reading like signposts versus, like, little–
Niki:
Right. The chunked narrative notes are usually, you know, off in the side or in a corner. I mean, they can be within the body of it too, but there is a distinction between the labels and a smaller explanatory function.
Michael:
So there's been a lot of discussion here about the language between scientists and engineers and art folk. You know, I'm gonna just speak to one quick thing. You know, I have fond memories about putting scientists and data visualizers together and having them talk because I find them not being very dissimilar in a lot of ways that they think.
Jen mentioned a data visualizer named Nigel Holmes earlier, and earlier in my career I got Nigel Holmes and Michio Kaku in the same room. Michio was doing a very simple– He was trying to explain relativity to a very lay audience, and we hired Nigel to do the work.
How do you guys feel about that kind of interaction? You know, the direct interaction between a data visualizer or artist and scientist?
Kelly, what do you think about that?
Kelly:
So that's pretty much what we do all the time. [laughs] Nature does obviously have some journalism where it's more akin to the work that Scientific American does where, you know, you have a sort of professional science journalist, but most of our content is either primary research or expert authored. While we may not be dealing with the author of the paper, we'll be dealing with someone who may have peer-reviewed it or someone in the circles or whatever.
So I think to me, that is sort of where things become the most fruitful because– I mean, I know we have said over the course of this event that there are sort of different languages, but I find that people treat scientists as if they're not creative or they don't, you know, get art or design or whatever, and I have found that to be not true at all.
Niki:
[laughs] I agree with that. I completely agree with that, Kelly.
Kelly:
And sometimes they have artistic references that actually are new to me or whatever. And I think also we have these sort of professional editors as translators as well, which is sort of a great thing about working for our publication is that, you know, you have more people to bounce ideas around.
And I would say that generally, no matter what you're making and whoever the audience is and whatever you're doing, bring someone in that's fresh that doesn't know what you've been working on or– Always have a second pair of eyes or a third pair of eyes and that will improve everything you're doing, whether it's a poster for a session or whether it's something for a press release or whether it's something that you'd like to have published. This is my biggest bit of advice, usually, to people.
Michael:
Jen, what do you think? I mean, you tend to do this too, I know..
Jen:
Yeah, I mean I think I don't know if I can beat Kelly on that one. [Kelly laughs] That was a good way of articulating it all. But yeah, like it's fun and surprising to see the kinds of results you can get when you have collaborations with almost anybody. But, you know, when you have a collaboration with a scientist who's very excited and passionate about the content, it's fun to see conversations and results that kind of come out of that.
Michael:
Yes. And as Catherine Mc—pardon me if I pronounce your name incorrectly—McRobie points out in the chat box that, you know, freelance scientific illustrators exist and are usually delighted to work with scientists in any field. I will echo that that is 100% true. And in fact, most artists are very scientific in a lot of ways, and they like to take those kind of conceptual ideas and help scientists and engineers to show things like this.
So we're coming down to the last seven or so minutes of our time here.
Another question that came in and I think is important to the folks who do graphics is, “Can you talk about the role of visual analogies between science concepts and more general real-life images, i.e. a motherboard equaling a subway map or stuff along those lines?”
Niki:
I'm not clear: motherboard equaling subway map?
Michael:
I think–
Niki:
Oh, I see.
Michael:
You know, the idea that, like, how do you see–
Niki:
Right. Got it. Yeah. Actually, I would say that that's something that Kelly touches on a lot with the covers—these analogies and these conceptual ideas, you know.
Kelly:
Yes.
Niki:
Yeah.
Kelly:
I'm sorry to interrupt, but it is an interesting point that like when we do things that—like let's say the microfluidic device that was a Swiss Army Knife—sometimes when we do things and it's clearly an object-based metaphor, we have to be careful that we don't make them into, like, a photorealistic kind of style that people think they exist.
Niki:
Absolutely.
Kelly:
So then we'll just choose a sort of illustrative style rather than a photorealistic kind of 3D-rendered style for certain things. Sometimes it's fine but, you know, I remember we had a paper on time crystals, and we sort of made the cover– It looked like these pieces of glass and the fog and things and with a few in-jokes with the numbers.
But, you know, we talked about, “Oh, well maybe we will animate this for, you know, social media or whatever.” But then we were like, “If we do that, it will look real and we don't want people to think that this is what they look like.” I mean, they are real, but it's like they don't actually look like this. You know? So that's a good point.
Niki:
Yeah, that's a good point.
Michael:
That point too, you know, we also did a cover about time crystals. The difference between, I think, Kelly’s publication and my publication is more nuanced because we're talking to a very educated lay audience. Part of the reasoning behind the way that we do things as opposed to Kelly's team, I think in some cases, is because we are trying to grab people's attention. When there were newsstands in pre-COVID times, I think that [Niki laughs] we are trying to grab the attention of people as they pass them and getting them to pick those things up, which is not a problem that everybody on this call will really need all the time. Again, it comes down to who your audience tends to be.
And then another question came in related to that, so I'm gonna just finish up on this point, but it’s “How do we at Scientific American choose covers and choose, like, what stories end up in things?”
Similar to Kelly's point, I think it's a little all over the place. We have an editorial board, and we schedule stories far out in time. Right? So between two and six months out. We do a lot of online stories too, but the cover stories tend to be, you know, things that we think are gonna pull in more people, are gonna relate to the subscribers of the magazine better.
And those stories are sometimes written by journalists, but about 50% of those things are written by scientists and other science-related people. If you submit something to Scientific American or if somebody comes to you from Scientific American, that may be on the cover if we find it seems like it's interesting and it'll reach a large audience for us.
Kelly, I don't know. You don't have to worry that much about that, I don’t think. Right? [coughs] But your magazine ends up in everybody's, like, scientists’ inbox all over the place.
Kelly:
[laughs]
It's true we don't have to worry about the newsstand. [laughs] So changes the process quite a bit.
Niki:
I can imagine, yeah. Well–
Michael:
[Coughs]
So I think coming close to the end of this, and I want to just say thank you to everybody who attended and certainly to the speakers. I really appreciate everybody who's been here, and to Robert Hornsby who's been behind the scenes sort of helping this whole time. I want to just say thank you.
I want everybody to know that we've been recording the full session and a video is going to be available afterwards. I think that we're going to email that to all the people who have registered to this. And we're also going to save the chat with all the links so you can see– Like there are a number of helpful links. I noticed the Adobe Suite things popped up in a couple of places and some other helpful tools.
I'd suggest if you have the time to go back and, you know, go and find the person that you think is going to really be helpful. I know Jen's presentation had a number of helpful links in it, and I think that I've used about half of them. So, you know, even if we work on a daily basis together, I usually forget these things. So these are all useful tools.
But yes, again, thank you to Jen, Kelly, Nicoletta, and Alissa for being on this panel and really being very helpful. And I hope that this has been a good help to everybody else who's been on this call.
Niki:
I'd like to say thank you to you, Michael, and I appreciate having you and Jen and Kelly. So, awesome. Thank you for coming to Columbia and speaking.
Michael:
Our pleasure.
Kelly:
Thanks for having us.
Niki:
Okay, well, goodbye. [laughs]
Unidentified attendee #1:
Thank you all.
Unidentified attendee #2:
Thank you.
Unidentified attendee #3:
Thank you.
Unidentified attendee #4:
Bye.