The Process of Building Science-Centric Graphics
By Jen Christiansen / 6 minute read
As a science-graphics editor at a magazine, I specialize in making advances in science and technology accessible to nonspecialist audiences. That includes developing illustrated explanatory diagrams and data visualizations that explain the latest research findings in depth and place those findings in context of the larger research arc. That may sound straightforward, but many newsworthy topics in science stand on a foundation of lots of incremental research findings, most of which I can’t assume that readers already know.
When developing science-centric graphics, background research is key, as is consulting with content experts, such as research scientists, to make sure that information is portrayed correctly. During my time as a graphics editor at Scientific American, and my past experiences as a researcher and art director at National Geographic, I’ve developed several strategies for illustrating technical topics for broad audiences, on deadline. What follows is an idealized version of my process for working on graphics that are rooted in cutting-edge science.
First, I should note that many graphics in Scientific American are developed by freelance artists. The eight stages below apply in all cases, but the details of each stage reflect my own approach to pieces that I render from start to finish. Others should have the freedom within this framework to adjust the details to work best for them.
Also, I recognize that different organizations have different timelines and levels of resources. These steps work regardless of deadlines or organization size, or even story scope. Those differences just indicate whether one can work through these steps in hours or in weeks.
Stage 1: Establish the Basic Goal of the Graphic
In general at Scientific American, text manuscripts take the lead. Upon reading a draft story, I identify which concepts I think would be best served with an illustration. My animating question is: Would a graphic be useful in helping to convey this information?
When are graphics useful?
- When images can tell the story more efficiently, effectively, or completely than words can. Take the iconic Feynman diagrams, which show subatomic particles’ movement, collision, and energy transfer. In these cases, the visuals stand in for abstract formulas.
- When the narrative involves complex and intertwining relationships, and an image map can help the reader track connections. An example is a diagram that explains the intricacies of photosynthesis.
- When the reader might benefit from seeing and exploring trends and patterns of the complete data set rather than be served up a few key numbers in the text, such as a chart showing the daily number of new Covid-19 infections.
- When a direct and immediate visual comparison is useful in highlighting change, or differences between states, such as competing hypotheses or before-and-after views.
After I identify potential graphics topics, I confer with the text editor to make sure we’re in agreement, and to confirm that the intended text and image paths haven’t diverged.
Stage 2: Research
Whenever possible, I begin with the primary source: ideally the key journal paper that describes the latest research, or direct communication with a scientist. Then I expand from there, starting with papers that catch my eye in the primary paper’s citations, and basic searches on the lead scientist and his or her collaborators. Then I move on to the bigger picture: How does this latest finding fit in with other research in this area? More often than not, I need to do some basic searches on core concepts, to make sure that I’m not misinterpreting things, and that I have a basic understanding of the terminology. Keyword-driven Google image searches help me sort out what other graphics have already been produced on the topic, and often help me to identify weaknesses or holes in the broader coverage — helping to focus my line of thought on how we can add something new to the conversation.
Assistance from the text editor and writer is useful at this point, as they often already have key references in hand, along with a list of possible sources. This stage is perhaps the most variable of all in terms of time span. If the graphic will be focused and narrow in scope, this stage can be quick and efficient, taking no more than a few hours. Larger, epic projects that aim to bring many lines of research together — like the National Geographic projects described by the art researcher Amanda Hobbs in her post “My answer to the question: ‘So, what do you do?’” — take more time; think many weeks.
Stage 3: Concept Sketch Development (Gestural Pencil–Sketch Equivalent)
Now it’s time to put that research to use and translate my written notes and doodles into a cohesive sketch. I begin with articulating what, exactly, the graphic aims to explain, starting with the wide view. Am I comparing and contrasting competing hypotheses? If so, two side-by-side panels might make sense. Showing change over time? A linear or cyclical step-by-step approach may be useful. Showing how something works? The physicality of the subject may help inform the layout. This step often comes together over the course of hours or days, since it’s rooted in the thought and work already completed in Stage 2.
Once I’ve spent some time thinking through the basic form, I develop a rough layout, with straightforward and descriptive headlines and subheads, so that I can better communicate my intentions to my colleagues and expert consultants.
All the while, I’m thinking in terms of:
- centering the new research finding;
- supporting the new information with broader context;
- considering what additional details might help engage a reader who is initially unfamiliar with the broader topic;
- shaking out label jargon (by using plain language); and
- shaking out visual jargon (by avoiding icons familiar only to scientists in a specific discipline).
Stage 4: Concept Sketch Review
Once I’m happy with the concept sketch, I share it with colleagues to make sure that preliminary plans for the text and graphic are still cohesive. Coordination between the text editor and graphics editor is critical at this point, since this is the time for changes in the graphic’s concept. The graphics don’t need to echo the text, but the pieces should be complementary. Then the concept sketch gets sent out to an expert, such as a research scientist, for review. Feedback loops with outside experts can be slow, so I try to accelerate the process by including questions designed to prompt immediate and focused attention on the validity of the overall goal of the graphic, as well as anything that I’m particularly hesitant about.
In some cases, if my initial interpretations don’t pass muster with the content experts, I’ll need to loop back to Stage 3. If revisions are needed, I’ll ask the content expert for more reference material, to help guide the revised plan.
Stage 5: Tight Sketch Development (Detailed Pencil Drawing, Rough 3D Render, or Equivalent)
Once the concept sketch is approved by colleagues and content expert, the next stage is all about folding specific notes and change requests into a tight sketch. At this point, I sharpen the illustrative details and flesh out the labels. Meanwhile, the text editor drafts captions.
Stage 6: Tight Sketch Review
Once I’m happy with the tight sketch, I send it out for another round of reviews. Feedback that focuses on specific details of the illustration are fine at this stage, but the composition and overall plan should not be changing. If there was a fundamental problem with the content — and therefore a fundamental problem with the composition — that should have been flagged at Stage 4. To avoid the need to backtrack, I find that being transparent with outside reviewers and colleagues about the review process is key. I generally let them know that they’ll see three rounds, and that each review step is designed to build on the previous step. If folks know that the first review may be their only opportunity to flag a fundamental problem with the conceit of the illustration, they’re more likely to review that step with the big picture in mind, as opposed to getting lost in reviewing smaller details from the start.
Stage 7: Final Graphic
Once the technical details are in place, the focus shifts to the final rendering and final captions. In my experience, the text editor often writes captions, although in some cases — and in some newsrooms — captions are the responsibility of the writer or graphics editor.
Stage 8: Final Graphic Review
The final rendering is sent around for a last look by colleagues (including a copy editor and a fact-checker) and the expert consultant to ensure that no errors were introduced in the final rendering stage. At Scientific American, we’re often working with scientist authors, so an official fact-check at this late stage is generally sufficient. But other publications, with different production work flows, may want fact-checkers to weigh in earlier, perhaps at Stage 4.
By working through these stages, with concept sketches that start as broad-stroke composition guides that are rooted in the concept being explained, I find that I’m forced to think through the content before getting distracted by drawing details. If the organization of the graphic is solid, then the illustrative details can develop organically within that framework. In the spirit of an anatomical artist, I strive to get the bones organized properly before fleshing things out.