Friday, July 14, 2017
Gilbert Scott Building - Room 132 (University of Glasgow)
This paper combines the strengths of two longitudinal data sets to produce a comparative study of follower behaviour in the social media platform Twitter. The first data set tracked the Scottish Independence referendum. The second tracked the UK’s referendum on EU membership. Together they constitute a unique resource for examining fluctuations in online user behaviour in the context of a constitutional referendum campaign. In relation to the Scottish Independence referendum, two of our authors found that, as individuals with different political interests prioritize campaigns variably, they tend to decide and to participate at different time points. This simple micro-level dynamic gives rise to an emergent regularity at population scale. Thus the number of new online participants was found to be inversely proportional to the corresponding time to the final voting day. That is, people’s online participation followed a power-law behavior. In this paper we compare and contrast the dynamics in the UK’s referendum on EU membership with these findings. Initial investigations find a similarly steep tailing in the closing weeks and days of the referendum campaign, but with an observed overshoot, with follower subscriptions peaking just after the UK’s EU referendum. In this paper we drill down into these data sets, seek to explain the similarities and differences in Follower-behaviour and explore the wider implications of our findings for the study of public attitudes as expressed in social media during constitutional campaigns.