Q&A with Alberto Cairo
Conversations with Data: #57
Do you want to receive Conversations with Data? Subscribe
Well-designed data visualisations have the power to inform and leave a lasting impression on audiences. From cutting through the clutter to simplifying complexity, visual storytelling is no longer an add on -- it's a necessity.
In this week's Conversations with Data podcast, we spoke with author, educator and data visualisation expert Alberto Cairo. He is the Knight Chair in Visual Journalism at the School of Communication of the University of Miami (UM) and also serves as director of the visualisation programme at UM's Center for Computational Science. He talks to us about the latest edition of his book "How Charts Lie", and provides some useful advice for journalists covering the US elections.
What we asked
Tell us about the start of your career. How did you first become interested in data visualisation?
Like most people working in data visualisation, I ended up entering this field by happenstance. As you know, I'm originally from Spain and I studied journalism with the idea of eventually working in radio. I really liked radio and I had an internship in Spain's National Public Radio, and I even read the news. In the last year of my journalism studies, a professor of mine saw that I could sketch things out. She recommended me for an internship, in the graphics department of a newsroom. I knew nothing about information graphics or explanation graphics at the time. But I got into that internship when I was 22 or 23 and I fell in love with it and I learned on the job. I had very good teachers in the newsroom and I stayed in the field ever since.
At the time, I barely did anything data-related. Instead, I was mostly designing explanatory graphics. This meant using illustrations, 3D models and animation to tell stories. I eventually became El Mundo's head of graphics. This shift towards data visualisation happened around 2009 when I began studying cartography. As it has a high quantitative component, I also started studying statistics. So cartography led me to data visualisation and that's how I began working in this field.
Your latest edition of "How Charts Lie" is coming out soon. Does it cover any new material?
The new paperback edition comes out on 13 October 2020. There's some new material including a new epilogue about COVID-19 and what has happened with graphics covering the pandemic. It is relatively short, about 10 pages or so. But that's the only addition to the book. The rest is identical to the hardcover edition, with the exception that we are using a different colour palette. The publisher decided to change the cover. Instead of being a blue cover, it's a yellow cover, which is very eye-catching. I really like it. It's much brighter and perhaps a little bit less serious. But I don't mind, as the book itself is written in a cheeky tone. The colour palette inside, instead of being based on red and greys, is more orange and reds. And I really like it. So it's essentially the same book, but with those extra 10 pages about COVID-19.
Remind us again what the focus of the book is?
"How Charts Lie" is a book for the general public. It's a book for anybody who wants to become a better reader of charts, meaning statistical graphs and data maps, etc. The title is a provocation that intends to attract people to the content of the book. But the book itself is obviously not a book about how to lie with charts. It's more a book about how to become a better reader of charts. And moreover, a better title of the book could have been, "How we lie to ourselves with charts". The book points out how charts that are otherwise correctly designed, correctly made, may still mislead you if you don't pay attention.
This ties in with the misconception that visualisations are like illustrations, that you can just take a quick look at it and move on. The point that I make in the book is that we need to stop treating maps, graphs and charts as if they were drawings. We need to start treating them as if they were text.
You served as artistic director for "At the Epicentre" -- a brilliant visualisation about the COVID-19 death rate in Brazil. Tell us more about the project.
"At the Epicentre" is essentially a data visualisation that asks the following question: "What if all the people who die of coronavirus in Brazil were your neighbours? How many people around you would disappear?" The visualisation traces a circle around you and shows you how wide that circle could be if all 100,000 people or so were your neighbours. It shows that human beings have a tough time understanding numbers unless those numbers put us at the centre.
I must clarify that I was the art director for this project, but I cannot take credit for the idea. The creators are responsible for the idea. We worked with a team of designers and developers in Brazil as well as Google News Initiative -- Vinicius Sueiro, Rodrigo Menegat, Tiago Maranhão, Natália Leal, Gilberto Scofield Jr., Simon Rogers and Marco Túlio Pires. The data visualisation is available in both Portuguese and English.
Finally, what advice do you have for journalists and designers covering the upcoming US elections?
I would say journalists should try to deemphasise horse race narratives in their reporting. We journalists tend to overemphasise the latest poll. For instance, one may show Trump is losing so many points in comparison to the previous poll. Another may show Biden is gaining so many points over an earlier one. This is just another mantra. Any single poll is always noise. What really matters is the weighted average of all those polls. Over a period of time, five or 10 polls all pointing in the same direction might be a pattern. But the outcome of a single poll itself means nothing. So if one poll shows Trump is gaining four points, that's meaningless because that may be just noise. Maybe it's a product of polling error. We don't know.
I wish that journalists would be a little bit more cognizant on how to think probabilistically. My advice is to learn a bit about probability. And that might be an idea for my next book. All of these polls and forecasting, are interesting to read. I'm not against publishing numbers, but I wish that we would emphasise even more than we do right now the level of uncertainty around those estimates that we're making.
Our next conversation
In the next episode of our Conversations with Data podcast, data designer Stefanie Posavec and data journalist Miriam Quick will talk to us about their new book, "I am a book. I am a portal to the universe." We will also hear about their visual inspirations and passion for pushing the creative boundaries in data storytelling.
As always, don’t forget to let us know what you’d like us to feature in our future editions. You can also read all of our past editions here.
Tara from the EJC Data team,
P.S. Are you interested in supporting this newsletter as well? Get in touch to discuss sponsorship opportunities.