A great post by Jen Schradie on pitfalls of using big data from online sources for social movements research. She lists five limitations of making inferences from this kind of data, and these serve as rather important points of caution when consuming this kind of work as well.
To sum up, the five:
- Hashtag data are often cherrypicked (case selection biased towards high levels of social media use)
- Big Data is too small (underrepresentation of poorer, less educated people and marginal groups online creates important bias)
- Privileging the online often ignores the offline (“we should not be fetishizing online spaces as where everything happens”)
- So much data, too little qualitative methods (the ease of obtaining quantitative data from online sources makes contextual understanding form qualitative methods even more necessary)
- Online data can ignore societal structures (technology needs to be better understood not only as a factor producing outcomes, but as a medium whose use is subject to social structures)