the kdmcinfo weblog

Better timelines through editing

Some really smart programmer/journalists at ProPublica recently released an open source interactive timeline tool called TimelineSetter. It does many things very well, including automating timeline creation from a csv file. But there are things journalists should do to take better advantage of the tool. 

A timeline must illustrate time

Illustrating the relationship between time and events is the reason why you should choose to create a timeline. If it's not a factor, a chronology is easier for the reader to consume. I wrote more about that on my personal site last year. 

For example, I produced the graphic below in 2005 at the Sun-Sentinel, when more hurricanes and tropical storms formed in the Atlantic than ever before recorded. It illustrates the frequency of storms and the scarcity of days without any storms. The graphic is about hurricanes, but it resonated with readers because it comes close to quantifying fatigue. 

 busiest hurricane season

Click image for large version

It's hard to explain the stress associated with being under a storm watch for five months straight. The constant state of readiness, not knowing if you need to hunker down or flee, takes a physical and psychological toll. Reactions to the graphic reflected that. "Oh! That's why I'm so tired!" and "Good, I'm not crazy." were common responses.

Applying the principle to the TimelineSetter tool

Creating effective timelines requires tough editing. Let's examine this example produced by NPR. The headline states it's a history of international law but the explainer text suggests it seeks to illustrate the evolution of civil rights for women under international law from 1945 to 2010. We'll assume for the sake of discussion that the story that this timeline should tell is about women's civil rights. 

When the graphic loads it displays of the period between 1945 and 1967. The easiest way to navigate the information to use the Previous and Next buttons.

npr timeline 1

But this subset of the information doesn't illustrate the events' relationship to time. To see all the entries, we have to use the magnification glass to zoom out. The resulting timeline is too visually dense to understand without reading each callout item. 

npr timeline zoomed out


So what's the solution? Tough editing. Strip away anything that is not absolutely essential. Start by removing the dates in the bar. They are impossible to read and important dates show up in the callout boxes anyway. All the reader needs is the range. 

full view of npr timline

That's better but the data doesn't clearly reenforce the statement made in the explainer text. Let's move the women categories to the left and turn off the other two. The resulting image shows that there is a greater density in the latter part of the timeline. 

npr timeline women only

But we can still edit more tightly. If you turn off everything except Women's Rights you see that there isn't a dramatic shift.

npr timeline only womens rights

But if we show only the Women, War & Peace category we see a dramatic shift over the last few decades. 

npr timeline only Women, War & Peace

Tough editing has taken us from an overly dense visualization that is impossible to comprehend on first glance to a visualization that illustrates the data. Explainer text should be rewritten to directly emphasize the point that's now illustrated and invite readers to compare/explore the other categories (which need more explanation). 

The martini glass approach

Editing data to focus a story and then give users a way to explore additional information is sometimes called the martini glass approach. Imagine your reader enters your visualization through the small end of the glass where they get the information that you as a journalist reports as the most significant.

This reduces the chance that a reader is overwhelmed by the information and leaves your page prematurely. It also ensures they have the best chance to understand the most important bits.

Martini glass approach

As they finish the story you have laid out for them, they begin to explore the additional data. This causes a deeper engagement with the data than simply clicking through a chronological slideshow. Users engage in play and exploration. They are rewarded when they learn on their own. 

And there are additional benefits. Users are also more likely to associate a positive interactive experience with you, your site and your brand. This provides added incentive for them to return in the future.