There’s been a significant push in Internet Studies over the last few years towards ‘big data’ studies, which aggregate huge volumes of information (such as tweets or website linking patterns) and subject them to analysis, often quantitative analysis. Much of this research provides valuable insights into how people are using the Internet and its impacts on society, politics, and economics. At IR13 last year there were plenty of projects which took a ‘big data’ approach to the study of social movements, particularly the Arab Spring and Occupy, and provided important analysis about how they organise and communicate. And, of course, my collaborator on the Mapping Movements project, Tim Highfield, is doing excellent work in the area.
However, I do think that there are important aspects of this shift towards big data that we need to maintain a critical approach towards. Part of the reason why ‘big data’ is so appealing is that it looks like Science: there are numbers! and statistical analysis! There have been claims that it will allow us to ‘do away with the need for hypothesis and theory’ (presumably ridding ourselves of the biases contained in these processes). It fits within our perceptions of what ‘proper’ science should be: more objective, less reliant on qualitative methods like participant observation and interviews. This notion of science has, of course, been critiqued from a number of perspectives. Emily Martin’s ‘The Egg and the Sperm‘, for example, provided an excellent demonstration of how profoundly even ‘hard’, supposedly objective, science, is shaped by cultural assumptions, including those surrounding gender.
The shift towards ‘big data’ is not only linked to the uptake of new analytical tools, it is also linked to our (gendered) ideas of what science should look like. As more funding becomes available for big data research, it is important to bear in mind the ways in which our assumptions structure the value we place on different research, and the ways in which access to different research fields is gendered. While many women provide vital contributions in STEM fields, there continue to be significant structural barriers to participation by women and minority groups in these areas. Devaluing qualitative research in favour of quantitative big data not only builds on misplaced assumptions about the value of ‘hard sciences’, it also adds to the factors excluding marginalised perspectives from academia.
This is not to say that we should abandon big data approaches. As I said, I believe that they provide many helpful insights. There’s also some fascinating work out there that uses big data in ways that undermine the assumptions that this research must be ‘objective’ – Zizi Papacharissi and Maria de Fatime Oliviera’s work on affective publics springs to mind here. Tim and I are approaching the use of big data by drawing together big data approaches and participant observation, interviews, and other qualitative methods. So the issue is not so much whether we use big data, as whether we remain aware of the ways in which its use is structured by our assumptions about what constitutes ‘science’, and of ways in which this may privilege some groups’ participation over others.