From the course: Data Visualization: A Lesson and Listen Series

Lesson: Complex visualizations

From the course: Data Visualization: A Lesson and Listen Series

Start my 1-month free trial

Lesson: Complex visualizations

“

(upbeat music) - I'm a big fan of simplicity. I talk about it all the time. My primary design advice for instance, can be boiled down to two words, do less. Strip away complexity, boil down everything you produce to its essence to give yourself the best chance to be heard and understood through the noise. But, there are exceptions. Sometimes what you're investigating is complex and you can't boil it down or aggregate the data up to one single average or key metric and still provide anything meaningful to your audience. Sometimes you have to share the complexity. Think of a huge issue like climate change. You could boil it down to an average temperature rise worldwide one number, but, that doesn't provide nearly enough complexity to explain how it's going to get colder and wetter in some areas, hotter and drier in others that there'll be catastrophic changes in ocean acidity and a host of other issues that require understanding to really see the issue clearly. And sometimes regardless of the presence of complexity there may be controversy. If the subject is controversial and you expect that you might face a hostile or skeptical audience, research has shown that when you simplify too much, your audience is actually less likely to be convinced by your data. So, when you need to approach a complex or controversial subject, what do you do? I have a few ideas. One thing is to bring on some expert help. If it's that complex, unless you're an expert in the field you might want to enlist the help of an expert to be sure you're getting the story and the data, right. You certainly don't want to make any mistakes that in some cases can have dire consequences. As an example, think about data about the efficacy and safety of a vaccine, where the implications of good and bad information could be literally life and death for thousands or even millions of people. Another idea is to do something I recommend frequently even for less complex stories. Even if you simplify the story you're telling to its barest components there's a strong argument for still showing more data. As an example, let's say you're investigating the correlation between several variables and the efficacy of a new medication. You could simply bubble that up to a single number. Say 92% of people with blood glucose level X, oxygen level Y and body weight Z were fully cured after the recommended end days of treatment. That's nice and simple messaging and maybe very helpful for say a presentation but, instead of just putting that on a slide how about showing me the bubble chart with dots for each of the thousand patients studied so I can see that strong correlation between those four variables as well as any interesting outliers. And one last idea is to skirt the complexity and controversy entirely by focusing on the story from a completely different angle. That's sort of what McKinsey and company did when they created a series of visualizations during the COVID-19 pandemic. Rather than look at the highly nuanced medical or public health data, they focused mostly instead on the economic implications which is more in their wheelhouse anyway. Next up, we'll be talking to Jason Forrest from McKinsey who headed up the effort and will tell us more about it.

Contents