The Politics of Data Visualisation

On the Internet ‘big data’ is popularized by data visualisation (datavis) that makes raw data accessible and meaningful to wider audiences. Here big data is harnessed to address society’s problems by illustrating trends and debunking assumptions that contradict the data. Despite the value of this work, the process of collecting and visualising data is never entirely neutral and never complete and so data visualization cannot capture every relevant fact. Something will always be missing. Data visualizations conceal more complicated realities. The decision to collect data and how this data is represented all reflect ideological assumptions and often unstated political agendas. Data displays embody values. These are reflected in which data is selected, as well as the methods, media and styles used to communicate information. Big data driven data visualisation on highly complex and political issues all too often result in a reduction of the complexity and a flattening out of phenomenon to what can be captured with numbers. Purely quantitative approaches to data visualization driven by big data are inadequate on politicised issues as they typically fail to capture power relations, ideology, attitudes and behaviors that cannot be reduced to a number.