10 key principles for data-driven storytelling

Over the summer, I tweaked our findings – and conclusions from our classroom conversations – into a short book chapter , featured in a new book called ‘Data Journalism: Past, Present and Future’ .
Based on our reading across these three thematic areas, and detailed research into coverage provided by seven different journalistic outlets, we were able to identify a series of important principles which all journalists using data should seek to abide by.

Many of these values are not just applicable to data storytelling; indeed, they should already be an established part of the journalist’s craft, but these principles are worth reiterating nonetheless, not least because – based on our reading and research – these core values are all too often overlooked.

Here are the 10 fundamental principles we identified:
Don’t beat readers over the head with numbers. Data should be used to illustrate key points and help bring a story alive. The data itself is seldom the story, the implications of it need to be unpacked and explained, like in this piece from ProPublica about the complex way that Charles and David Koch spread their wealth in politics.

Decide which tools allow you to tell a story best. Journalists have access to a plethora of means, including graphs, infographics, charts, tables and interactives to help tell a story. The value derived from each of these tools will vary depending on the story you wish to tell and the key data that you want to focus on.

Determine if you should develop a house style. When looking at sports stories, we found that each publication favoured a different way of incorporating data. When looking at basketball and the success of the Golden State Warriors, the San Jose Mercury News tended to weave stats (and in this case video) into stories to help make key points, SB Nation showed a preference for bar graphs, whilst FiveThirtyEight had a larger propensity for tables and complicated graphs.

Determining your approach may be influenced by both the story you want to tell and the data literacy and preferences of your audience. Either way, your data-driven stories should be well designed so that audiences do not struggle to understand what’s being shown or how to interact with the data. For complex visualisations that may mean including some instructions (an idea too often overlooked by journalists), as well as links – including downloadable data files – to the original sources used in these efforts.