Jensen, E. A. (2017). Putting the methodological brakes on claims to measure national happiness through Twitter: Methodological limitations in social media analytics. PLOS ONE. DOI: 10.1371/journal.pone.0180080
With the rapid global proliferation of social media, there has been growing interest in using this existing source of easily accessible ‘big data’ to develop social science knowledge. However, amidst the big data gold rush, it is important that long-established principles of good social research are not ignored. This article critically evaluates Mitchell et al.’s (2013) study, ‘The Geography of Happiness: Connecting Twitter Sentiment and Expression, Demographics, and Objective Characteristics of Place’, demonstrating the importance of attending to key methodological issues associated with secondary data analysis.
With the rapid global proliferation of social media, there has been growing interest in using this existing source of easily accessible data to develop social science knowledge. The extraordinarily large sample sizes that are made possible by social media-based research have made such ‘big data’ studies particularly alluring to both scientific journals and news media. However, amidst the big data gold rush, long-established principles of good social research have too often been ignored. The consequences of flouting these principles have not been fully articulated to date. This article uses a published example by Mitchell et al. [1] to illustrate how methodological limitations can undermine big data research. The main issues identified are [1] inferential over-extensions resulting in over-claiming [2], limitations in the operationalization of key concepts, [3] de facto sampling bias, and [4] a failure to account for the inherent shortcomings of this form of secondary data.
