As families throughout the UK settled in for an evening of scary movies and calories this Halloween, we were interrupted by a new and far more horrifying tradition — the televised government COVID briefing. Here in Edinburgh, our family had a merciful sense of distance, because many of the aspects of the fight against the virus are devolved, meaning that the restrictions to be announced would apply only in England. Scotland, along with Wales and Northern Ireland, have their own rules.
What bothered me about this particular press conference was the quality of the presentation. A lot has already been written about how Boris Johnson’s government is handling things, look here, here and here for examples. Ministers have been accused many times of not ‘following the science’, so perhaps in an attempt to counter such criticism, Johnson handed the briefing over to Chief Medical Officer, Chris Witty and Chief Scientific Advisor, Sir Patrick Vallance to ‘present the latest data’.
Unfortunately, the Powerpoint slides that the UK government presented to us on Saturday, which we can only assume were intended to demonstrate their understanding of ‘the science’ were messy, overly complicated, disjointed, and confusing. Professor Alice Roberts of the University of Birmingham, among others called out this particular egregious example.
So, why am I writing about this? If the government’s Powerpoint game is weak, what does it matter? They’re not in power to make pretty graphs, they’re there to make decisions, right?
Well, no. It does matter; in fact, it matters a lot. There’s a lesson here for any organization that doesn’t take the time to properly process and understand the information that they have available before sharing it publicly. The real problem isn’t that the slides were poorly laid out and the axes were often missing. The biggest crime isn’t even that the text was too small to read, or frequently ran off the edge of the screen. The problem is that nobody had taken the time to understand all the different sources of data and synthesize them into a coherent story.
The art and science of data storytelling
We all know that telling stories is a powerful way to convey ideas. We’ve been doing it as a species ever since we developed language. Stories, particularly stories about people, are interesting to us. We connect with them emotionally and, whether it’s joy, love, surprise, anger, or sadness, there are deeply wired survival mechanisms in our brains that tell us that we should remember the things that make an emotional impression on us.
Data, on the other hand, has a reputation for being a bit boring because it’s about numbers, not people. It’s an ill-deserved reputation, however, because often, underlying all those numbers are real human stories. For instance, watch this data storytelling / standup comedy hybrid TED talk by Ben Wellington about what cycling accidents, parking tickets and subway fares can tell you about the cultural idiosyncrasies of New York City.
Sometimes, a lack of clarity or narrative can have serious consequences. An example from history was the disastrous decision to proceed with the launch of the Challenger space shuttle shuttle in 1986. In his book, Visual Representations, Edward Tufte makes the case that if data about O-ring damage caused by temperature changes had been presented graphically and in context, the launch would not have happened and seven lives would have been saved. This blog post shows both the slide that was presented as well as Tufte’s visualisation, the difference in clarity is striking.
The case of COVID data is certainly no less serious. There are many lives at stake; people, loved ones, communities, all affected by the pandemic. There is a complex series of risks to balance from the physical risk associated with the virus itself to the economic, mental health and other risks like increased domestic violence, posed by lockdowns and restrictions. If the government wanted to get us all on side with the new restrictions — as we need to be for them to work — all they had to do was take us through the journey of what those numbers mean about the lives of real people.
Take this data journalism piece from The New York Times, for example. Their animated data story manages to take us through how the US administration claimed to be handling the COVID outbreak and juxtaposes it against what was really happening, based on tens of thousands of data points. By combining context with visualization, and drawing out individual representative examples, Derek Watkins and colleagues guide us, the audience, through a complex emotional journey and deepen our understanding of what happened.
When Sir Patrick Vallance embarrassingly tried to pass the blame for his poor communication by saying, ‘this is a complicated slide, it’s from the NHS’, it was clear to me that he hadn’t been briefed well enough to know why that particular slide was in that particular place in the deck. In other words, there was no story, no context; it was just a bunch of numbers on a series of poorly laid out Powerpoint slides.
A lesson for all organizations
Good visualization is key to making sense of data. By understanding the grammar of graphics, we can learn how to encode data visually to make it both aesthetic and meaningful. If we want to go further than that and make meaningful decisions on the basis of our data, we must understand its context and that means telling stories about it.
So remember, if, when data is presented at — or by — your organization, it’s boring and hard to follow, if the stories that lie behind the numbers aren’t drawn out to make them meaningful and interesting, then nobody will understand the context or care about the data. And, if that’s the case then, like the UK government, your organization is likely to make poor decisions; and your stakeholders are likely to be as disengaged and dismissive of your data as the British public was of their government’s this Halloween.
If you haven’t had the pleasure and would like to see the original slides from the briefing, they’re available here, they’re in slightly better shape than when they were presented live, but not by much.
Thanks to fellow Chef, Alice Meadows for help editing this post.