File S1 - Rising Tides or Rising Stars?: Dynamics of Shared Attention on Twitter during Media Events

Supporting figures and table. Figure S1, Tweet volume per minute. Number of tweets per minute in the 12 datasets. (a–d) The six hours during the four debate events (“DEB”). For other categories, we plot the six hour volume centering around the peak within the data range: (e–h) Normal period prior to the debate evenings (“PRE”). (i,j) National convention events including RNC and DNC (“CONV”). (k,l) Breaking political news events including Benghazi attack and Romney's 47-percent video (“NEWS”). Figure S2, Changes in communication volume. Diamond shapes indicate the mean value of each category. This figure shows the ratio of tweets mentioning a user to the total tweets at the peak hour. Figure S3, Lorentz curves for cumulative degree distributions of activity. Increasing equality converges toward diagonal line from the origin to the upper-right and increasing inequality converges toward a hyperbola rising to 100% of volume at the 100th percentile. Figure S4, Connectivity-concentration state spaces. For each of the twelve observed events, the Gini coefficient for the network's degree distribution is plotted on the -axis and the average degree of the network is plotted on the -axis. Table S1, Kolmogorov-Smirnov test (K-S test) for comparing the PRE curves with the remaining three curves in other conditions. (PDF)