For a number of years, I’ve been researching through empirically based methods and methodologies what we’ve called the “attention-comprehension gap” (St.Amant & Meloncon, 2015). What this little theory says is that research has often focused to narrowly on “readability,” data visualizations, or document design, but has rarely brought these areas together along with a specific focus on how well users actually understand the information.
Since good research often takes time, I’ve not published much on the preliminary work, but what I do know is that pieces like the recent “provocation” about wanting to make COVID-19 visuals more “gross” are not particularly useful to advancing information design (here this term is meant to be a combination of data visualizations, words, and document design) around risk communication.
Why do I say it’s not useful? Well, because it’s not grounded in empirically tested research, and it ignores a whole body of scholarship that actually has looked at the efficacy of visuals and risk communication. So for example, take the review of literature that I did with Emily Warner. It concluded the following:
- Pictographs and icons are good: this sort of intersects with idea to humanize visuals that was cited by Moeller and Bivens, but what they didn’t point to nor does more recent research validate is that there were a number of critiques of the humanization of visuals and little to no research that shows it makes a difference for users.
- Bar graphs are good: these are easy for folks to see and understand and more so for users with low numeracy
- Keep the visualizations simple: include clear labels and explanations without a lot of extra information
- Use care with design features for the visuals, as well as other parts of the “document design”
Around the same time as I was completing my review (decidedly from a tech comm perspective), Garcia-Retamero and Cokely did a review focusing more on literacy. A big takeaway from their review essay is “visual aids generally offer a relatively efficient means of reaching diverse individuals regardless of their levels of numeracy and graph literacy as long as the visual aids include simple labels and explanations…” Then there is a growing body of research around whether or not comics can be effective in in risk communication
Using existing research, we designed our initial materials to test with folks on these guidelines and we did vary them up in some situations. When working with visualizations around the increase of STDs in college age students with college age students, we found that the “gross” or focusing on the extreme rise and the somewhat “gross” effects of STDs did not resonate with users at all. Instead, they preferred documents that presented the visualizations and related information in a more straightforward manner. That is, well designed, clearly articulated visualizations that presented the data showing their risks and what they could do about minimizing their risks. Some correlations can be made between individual risk factors between our work on STDs and the current pandemic. Both situations are grounded in risk tolerance and information design.
One of the reasons we posited the attention-comprehension gap was because of the need to focus on comprehension and cognition when it comes to visualizations and document design. Much of St.Amant’s recent work is focus on this cognitive aspect of technical communication and design and there is some interesting work coming out of psychology around comics and cognition. What we found in our preliminary study* is if the design is well liked, it may not improve comprehension of the information around risk (see Garcia-Retamero & Cokely). In other words, I am sure that gross visuals would have some shock value and recall value for users, but the bigger question would be if the shock and recall actually obscured what the visuals were trying tell people. So a good research question would focus on the impact of the shock value of “gross” and whether it actually impedes comprehension. My guess is that folks would remember the visual but not actually why it was important.
For many of us trained in technical and professional communication with an emphasis on user experience, in-depth audience analysis, and information design , some of these findings seem to fall into the “no, duh” category. But what they do that is of the utmost importance is validate our own knowledge and expertise and guide the development of data visualizations and risk communication in an evidence-based manner. They give us a starting point that we can with a pretty high level of certainty actually work.
All of this is to say is I appreciate a good provocation to encourage researchers to think and to do things differently. But, in this case, the provocation was certainly not provocative. Rather, it went for a wow factor at the expense of already existing research, which is decidedly not at all what researchers need to be doing at this moment in time.
So as folks in rhetoric, technical communication, health communication, and the rhetoric of health and medicine want to move into public scholarship and posit interesting theories, I would encourage us to ground the public scholarship in existing research. And the current research tells us that we need visuals that are accurate, relevant, and they follow some of the evidence based guidelines that exist.
Should there be a well-designed research study that points to the importance of gross visuals in the near future, then I’ll be happy to include that approach in my own research and my own practice. Until then, I’m going to follow what empirically based research has shown to be effective.
*the link reports on a small slice of the larger study and not what is referenced here. We are still writing up the results.