How do we judge the resonance of data visualisations?

Posted on July 23rd, 2013 by Chris Twigg

How do we judge the resonance of data visualisations? Your comments and suggestions are invited! 

Data visualisation and resonance may not be the most obvious bedfellows. Resonance, in a communications context, is about the extent people connect with, care about and invest in an idea. Opposed to this, data visualisation is really about data analysis – a means to an end. But the obvious move of data visualisation into popular culture, which is a result of more and more data that needs to be distilled for people to understand, suggests data visualisations need to resonate to successfully communicate to audiences who are not subject experts!

I like to think data visualisation can add value beyond being a way to deal with complex stuff. How can data visualisations really connect people with data and make it matter to them?

The appeal of subject matter provides a clue. Andy Kirk wrote about ‘Appreciating the critical role of subject matter‘ in particular how folks can sometimes be indifferent or disinterested in data visualisations. Why is this? Most likely those people didn’t connect with the content. The purpose of the visualisation was out of tune with the context and needs of the audience, so no matter how well executed it was the visualisation just wasn’t interesting to them.

Technological innovation, functionality and the craft of data visualisations alone aren’t likely to inspire resonance amongst non expert audiences, at least not in an enduring way. Yes, we can see the shapes and colours the data make – but is there something significant about that? Yes, we can compare a data set – but is there a reason to care? Questions like these are fundamentally about communication. So the purpose of a data visualisation, it’s context and the needs of the audience are foremost in the question of achieving resonance.

Let’s consider audience, context and purpose just a little more. The literature on data visualisation(1), or even just a good look at examples around the web, suggests two poles. On the one hand there are exploratory visualisations like those used for scientific analyses. They can represent the complexity of big data sets. Their purpose is to allow data exploration. Yet it’s difficult to imagine exploratory visualisations resonating very well beyond special interest or expert user groups who have prior knowledge and interest in the subject matter – and actually want or need to engage with it.

On the other hand there are the kind of explanatory visualisations for example in newspapers and magazines. Journalists/storytellers distill data first then make data visualisation part of a story so people can understand it. Visualisations like this are supposed to resonate with a non-expert audience – which is the majority of people. They need to explain and appeal unless the audience is to pass them by.

Resonance (simply, being interesting) needn’t be at odds with principles of data visualisation (like being objective and using effective ways to see and understand complex things). I think there’s room to match good data visualisation (that’s useful, accurate and well crafted) with good communication (where the visualisation needs to appeal to a non expert/mass audience).

So, right now I’m looking for particular ways of judging how data visualisations resonate with non expert audiences. Do you have ideas or suggestions, can you add or suggest otherwise to this list?

Judging how data visualisations resonate with an audience, some ideas for measures:

  • Sense of engagement, interest?

  • Motivation to work through it?

  • Recall/things remembered?

  • Empathy or care for subject(s) involved?

  • Mis/understanding of the subject?

  • Belief/trust/credibility?

  • Take away message?

  • Would recommend to others?

  • …?





(1) See for example Illinsky and Steele’s Designing Data Visualisations (O’Reilly). They make the distinction between exploratory and explanatory visualisations, as well as touching on infographics vs data visualisation in this context.


Leave a Reply