Scientific Poster Design

Selling Science


Recently, Eddie Lee posted a link on Twitter about a conference, the 2015 Computational Social Science Summit, where he was giving a poster on war and peace in a pigtailed macaque society. In taking a look at the poster, I was reminded again of the importance of good design in scientific posters.

Design as Communication

This topic has been bouncing around in my head since reading the Walter Isaacson biography of Steve Jobs a few months ago. Love him or hate him, Jobs brought a level of obsession to the issues design and aesthetic, and his thinking about what design means struck a cord. I am particularly interested in how design and aesthetic affect communication.

For instance, take advertising. We all know how advertisements use design to influence consumer behavior. Sometimes the tactic is obvious - bikini clad models selling beer - but, often, it is more subtle. People are influenced by sight and sound, and we immediately make associations to those sights and sounds based on past experience. You do not sell cookie dough by claiming it is delicious tasting, you sell it by drawing a connection in the mind of the viewer to sitting in the kitchen as a child while the cookies were being baked, with the anticipation of having one becoming almost too much to bear. You draw a connection to the idea of home, of love, of warmth, without ever making that connection explicit - the viewer does it themselves.

All of which goes to say that design and aesthetic are forms of communication. In the world of fashion, that statement is taken as self-obvious. How people wear their clothes, and what clothes they choose to wear, signals something about them, and accordingly, tells us how we should interact with them. Or, phrased another way, by dressing in a certain style, people cause us to make associations to them based on our own conceptions of people who wear that type of clothing. We then interact with those people in a certain way, not because we know of them and their life story and come to the independent conclusion we should interact with them in this or that way, but because of how they dress and carry themselves.

The important point is thus that design not just helps facilitate communication, but is itself a form of communication. It is not an explicit communication, but instead relies on the conceptions of the outside observer to make the communication happen.

Scientific Communication - A Spartan Approach

Communicating science often fails to recognize the above reality. Science is suppose to be “clean”, without all the fancy bells and whistles. Presentations and presenters are there to communicate data and information, and to do so in a straightforward way. Every presentation follows a linear track - What is the problem?, Who cares about it?, How did we solve it?. (Though sometimes the middle question is missing from the presentation!) Graphs, figures, and tables are there simply to present the data. Presenters are there to follow the path, give the talk, and answer the questions. Crucially, scientific presentations often do not capture audience attention because the emphasis is on the data being presented, not the people listening to the presentation. Phrased another way, scientific presentations tend not to use design to capture audience attention. As a result, most scientific presentations lack audience engagement.

What does poor design look like? I have seen talks in which bright green lines are used in graphs with white backgrounds, making them literally painful to read. We have all heard presenters announce “I know you cannot see this, but….” only to announce one of the more important points in the talk. Slides contain paragraphs of text, sometimes simply copied from the paper. Such “presentation sins” combine in such a way as to cause the audience to tune out the presentation. Why pay attention when the graphics are hard to read, the font size too small, and the content poorly laid out?

Now, it should be said one way to see why scientific presentations have been crafted in this way is to recognize the importance of “truth in advertising” to scientists. You cannot get up and give a technical talk filled with errors, no matter how much you try to dress it up. Scientists do not want to be seen as snake oil salesman by their colleagues, since that reputation would spell the death of funding, conference speaking opportunities, and overall career success. Consequently, presenting the truth in a straightforward way is a logical method for avoiding that reputation. However, we seem to have gone a bit too far in that direction.

Advertising Science?

Advertising design sells products. What does scientific design sell? Not science, but the scientist. More specifically, the ideas the scientist is presenting. Unfortunately, members of the audience often have little technical expertise in the specific subject matter being presented, so subtle association-based selling cannot work very well. Instead, ideas are sold via aesthetic by implying competency and professionalism on the part of the presenter. Certainly part of competency is reporting truth - you cannot lie to a scientific audience - but another part is how the presentation looks and flows. Good design communicates to the audience “I know what I am doing, and so listen to what I am saying”.

Again, such considerations are often not of importance to scientists, primarily because competency is established on the basis of the material being presented. However, as more and more science gets thrust into the public light, scientists should be aware of how design affects the reception and comprehension of their work. The interplay between science and the public is a fascinating one, but we will leave those considerations for another day.

The Macaque Poster

Take a look at Eddie’s poster. What do you notice? For me, the immediate reaction was “Ah, we have two different things going on here - war and peace.” Notice the use of red and blue to distinguish between the two without having to draw clunky boxes or the like. Further, when doing comparative analysis on the probabilities for duration of the two, the colors neatly come together and merge.

On the whole, this poster exemplifies how science and good design need not be enemies. Colors, fonts, and graphics are used in such a way as to convey a sound scientific point - “How are war and peace different in this society of macaques?”. Viewers could look at the middle column and reach two essential conclusions: 1) war and peace have different durations in this society, and 2) when looked at relative to the variability of the duration, both tend to follow a similar pattern: neither war nor peace last very long.

Technical Note

For those who are interested, Eddie informs me the plots were made using the Python packages matplotlib and seaborn. Check them out if you would like to make similarly good-looking plots!