Like most people you probably feel that adding charts and graphics to a white paper or presentation will make it more readable. Now you have actual research that backs up what your gut has been telling you all along.
A recent article in The Economist, “Graphic Details,” highlights a study conducted to determine if papers with more figures and graphics in them were more widely read, and thus more influential. Find the story here; subscription may be required. In the study, Bill Howe and colleagues at the University of Washington, trained a computer algorithm to identify different types of graphics (diagrams, equations, photographs, plots, and tables). Then, using another algorithm, similar to Google’s PageRank, they evaluated more than 650,000 biomedical research papers housed in PubMed Central, an online database for medical researchers.
The results: Overall, graphics improved readership and increased the chances of a paper being referenced, or cited, by other authors and researchers going forward. But, some types of graphics worked better than others and some didn’t work at all.
Inserting diagrams improved the number of future citations the most — the average paper received two more citations for every additional diagram per page. For every additional plot (basic graphs such as bar charts or line charts) on a page, the average paper received one more citation. While the exact cause of the results is unknown, speculation is that diagram-heavy papers typically introduce new concepts and, thus, receive more attention in general.
While graphics and plots improved citations overall, certain types of graphics had the opposite effect. The use of photographs and equations slightly decreased the chances of a paper being cited, according to the article. One possible reason: study subjects were biomedical researchers, and they have little use for math. The authors did not speculate on why the use of photographs decreased popularity.
So, now that you are pumped to add more graphics to your papers and presentations, you probably want them to have an impact. Another recent article, this one from Harvard Business Review, can help get you started. The article, “Visualizations that Really Work,” is by Scott Berinato and can be found here. Again, you may need a subscription.
In the article, Mr. Berinato stresses the importance of understanding what you are trying to achieve with a graphic before you start plugging away in Excel (a position we, at Vertical, fully endorse). He organizes the world of visualizations into four types:
- Idea Illustration — These graphics are less about the numbers and more about teaching a concept. More abstract in nature, their goal is to explain something, usually a process or framework.
- Idea Generation — These graphics are a relative of the whiteboard or “back-of-the-napkin” diagram. Their goal is to solve a problem or explain complex systems. These graphics, according to Mr. Berinato, should be all about function — they don’t need to be pretty.
- Visual Discovery — These are charts and graphics that employ more sophisticated techniques to make sense of big data. Their goal is to identify and illustrate trends in a way that words, or a table or spreadsheet cannot.
- Everyday DataViz — These are the typical charts and graphs you create in Excel and drop in a PowerPoint. While based on numbers, they rely on less data than Visual Discovery visualizations and are often used in a more formal setting, to prove or illustrate a point. The goal with Everyday DataViz graphics is to keep them as simple as possible.
Reading Mr. Berinato’s article before you bury yourself in bar charts can give your paper an advantage. Remember, you want your papers and presentations to be shared and cited, but not as an example of how not to do a graphic.
– Michael Sobczynski