Andy Kirk's View from the Summit of Data Visualization

These are heady times for data visualization, says Andy Kirk, one of the apostles of the movement. There was a burst of growth last year, a move to focus more on message than ornamentation, and appeals to do battle with the forces of ‘post-truth.’ In this interview, Kirk addresses these issues and more in an era he considers just on the lee side of a "golden age" for data viz.Kirk launched his visualisingdata.com website in 2010 and became a freelance data viz consultant the next year, offering his vision to global behemoths such as Disney, Intel, WHO, OECD and McKinsey. He spent 18 months working as a co-investigator on ‘Seeing Data’ research project, which explored visualization literacy among the general public; that project was funded by Britain’s Arts & Humanities Research Council and hosted by the University of Sheffield. He is also a visiting lecturer for master’s degree programs at the Maryland Institute College of Art and Imperial College, London.SAGE released his latest book, Data Visualisation: A Handbook for Data Driven Design, last year. The Financial Times soon after named it one of the "six best books for data geeks."Andy KirkWhat’s been the most interesting development in data visualization for you in the last year?I would suggest the trend of being more discerning with interactivity. This is a trend that wasn’t necessarily triggered by the New York Times but was certainly catalyzed by a talk given by one of their most noted senior graphics editors, Archie Tse. (Here are the slides from his talk: https://github.com/archietse/malofiej-2016/blob/master/tse-malofiej-2016-slides.pdf.) This is not suggesting that interactivity is not still a hugely important consideration for creating compelling digital visualizations, rather it's about avoiding feature creep, avoiding gratuitous 'bells and whistles’ and avoiding unnecessary obstructions to the process of a reader gaining understanding from a visual. Interactivity is an ally but shouldn’t be a device used to abdicate responsibility as a reporter/communicator on to the user -- “here, you find something interesting.”What one piece of advice would you give to students seeking to understand how effective visualization is created?My piece of advice would be to open your eyes as a reader. When you are working to make sense of a visualization or infographic, look beyond the surface and see the decisions the creators has made: the chart types used, the color choices, the layout, etc. Consider how you respond to these choices - do they aid or impede the communication? - and you will start to develop the eye to judge these things for yourself  in your own work.In this so-called “post-truth” era what role can data visualization play?This is a burning issue in the field. The fundamental principle of data visualization, in my mind is, a) to be truthful, and b) to be trustworthy. So what undermines these things? Bad quality data, incomplete data, incorrect statistical models, misunderstandings about probabilities and uncertainties, lack of clarity about the provenance of data and the processes it goes through before presenting, snapshot analysis that misses context, nuance and hidden assumptions, corrupt visual design techniques that distort and deceive. These are all practices we as an industry/field need to stop, call out and rally against.What do you see for the future of the field?I think we are probably just on the other side of a golden age - a period of rapid growth, of experimentation and of increased mainstream exposure - and because of what has happened over the past 12 months (and characterized by some of the things mentioned above) there’s probably a sense now of sense that quality over quantity is of greater importance, that explaining is perhaps more important than exploring, that rigor is more pressing than speed. What won’t change much is people, and whenever you have humans making and humans consuming these things perfection will never be achieved. Let’s see how we go!  

*** More on Data Viz: An archived webinar featuring Andy Kirk and Stephanie Evergreen. Click here to view.

Previous
Previous

Ethical Considerations in Case-Centered Qualitative Research

Next
Next

What Do We Mean When We Say 'Big Data'?