One of the first things you learn when studying statistics is how easy it is to get them wrong – you get taught a long list of fallacies and mistakes and told to avoid repeating them. Then you learn, probably though harsh experience, how easy it is to get the statistics right, yet leave anyone reading your work confused, unclear and turned-off. Somewhere in this process, if you’re lucky, you will be introduced to the art of data visualisation. For me and many others, this moment arrived through viewing Hans Rosling’s famous 2006 TED talk "The best statistics you've never seen"

It was exciting, colourful and dynamic – he described it like a sports commentator, and no-one could avoid becoming engaged by what he was showing. Then you realised that not only was it engaging, but that the underlying data was deep and complex with all the foibles of real data and despite this he was able to put across a few clear messages in a way that his audience will not be able to forget. It was in this moment that I realised the true significance of the communication of data and statistics – not just getting it right, as that achieves little, but communicating the truths revealed from data so that people truly ‘get it’.

When I tried to produce similar engaging representations of data, it soon became clear that it was not as easy, or effortless, as he made it seem. A lot of work goes into giving a viewer both access to data while communicating your message in a way which leaves it stuck in their mind. Then I realised the same challenges arise in related fields such as designing good user interfaces or intuitive control systems. It’s not just enough to provide something for a user  – you’ve got to engage them, ideally so they want to use your tool, believe your message or trust your data.

The death of Hans Rosling today means we’ve lost a true educator and entertainer. Much of his work will live on, through his gapminder website which helps people produce similar visualisations, and many of us will try to produce exciting ways for people to interact with data and communicate what is reveals. But today is a sad day for anyone with an interest in communication, or of the millions who were excited by his talks.

Paul Baxter

Paul Baxter is a specialist in statistical processing, data analysis and signal processing algorithm design, especially robustness and applicability to real-world situations with interests in data visualisation and sensor behaviour.