Visualisations for Doing Good with Data

Ok, so I've boycotted the stupidly early 8.30 a.m. paper session at ECREA 2014. My day starts, consequently, with a paper by Helen Kennedy and Rosemary Hill, whose focus is on data visualisation. Data-driven decision-making remains out of reach for those who cannot work with the data directly, and data visualisation may help here.

There are now visualisation agencies such as Periscopic which specialise in data visualisation for social good. The idea here is to mobilise data graphically, to make data useful to a greater audience. Such visualisations include the work of The Guardian's Data Blog, for example, and may tackle serious as well as lighter themes.

But in order to make sense of such data-driven visualisation processes, we need to know more about the moment when a person encounters such visualisations. There is widespread discussion about how to optimise visualisations, but very little research into users themselves; much of this remains a debate by experts for experts.

The project addresses this lack by engaging with the 'consumers' of visualisations, variously across rural and urban, low and high migration areas. Getting respondents to engage was difficult, however, given the variations in interest in such questions.

Responses varied. One respondent disliked all visualisations outright, and also expressed a low confidence in his overall reading ability; there was some significant suspicion that graphs were used by the media to confuse viewers. Understanding the graphs felt like too much work to them, and the level of respondents' English language skills also seemed to matter.

But can data visualisations empower people, then? Confidence and (language and mathematical) skills will matter here, the need to commit time to understanding them, the level of pre-existing interest in a topic, and the degree of trust in the media's intentions also play a role.

So, engagement with visualisations is context- and content-dependent, and we need to think about visualisations as media forms and using media studies' conceptual resources. Identifying these issues is the first step to enabling people to engage confidently with visualisations; these are questions about civic participation, and as visualisations circulate more and more, people may develop more skills – but unless and until that happens, doing good with data will remain very difficult.