Big Data and Social Movement Research

By Fabio Rojas

“Big data” sounds fun and exciting but it has also been heavily criticized. But now, it’s time to step back and treat “big data” as we would treat any other form of data. We should identify its strengths and weaknesses and ask how it can help us with our own specific research goals. So let’s start with an obvious, but under-appreciated, point about empirical research: there is no such thing as perfection in data. Every method for generating and collecting data has strengths and weaknesses. Thus, we should be interested in data collection methods where the positive points outweigh the negative points. For example, experimental data has a great virtue – those who receive the treatment are randomly selected, thus eliminating bias. Experimental data also has a serious drawback. Experimental settings may not reflect “real world” processes and are often not generalizable. This is a serious problem for biomedical research, for example. A drug tested in a highly controlled environment may work differently than in the actual setting of a hospital. Yet, we value experiments because they do one thing exceptionally well – they eliminate selection bias and address the issue of confounding variables.