Of supranodes and socialwashing: network theory and the responsible innovation of social media platforms

Figures & data

Figure 1. IRA supranode presence on Twitter, Facebook and Instagram during the U.S. 2016 presidential elections.

Figure 2. IRA supranode network theoretic implications.

Table 1. Network structural measures and IRA supranode impacts

Network characteristic	Description	Impact
Degree	Number of ties to a node	Increase
Density	Percent of potential ties actually present	Increase
Average path length	Average of all shortest path lengths in the network	Decrease
Diameter	Maximum shortest path length in the network	Decrease
Cluster coefficient	Ratio of the actual triangular node connections in a graph to the total number of possible triangular node connections	Increase
Assortativity	Pearson correlation between adjacent node degrees	Increase
Global efficiency	Average of the inverse of all shortest path lengths	Increase
Eigenvector centrality	Connection by a node to highly influential nodes increases the connecting nodes centrality	Increase
Transitivity	Proportion of three nodes where if A connects to B and B connects to C then A will also connect to C	Increase

Table 2. Type of network ties and role in generating cascades

Tie	Exemplars	Role in generating a cascade
Proximates/ Similarities	Location, membership (Borgatti et al., 2009; Kane et al., 2014)	Developmental—shaping opinions Instrumental—diffuse information through bridging
Relations	Kinship, affective, cognitive (Borgatti et al., 2009; Kane et al., 2014)
Interactions	Sex with, talked to (Borgatti et al., 2009; Kane et al., 2014)
Flows	Information, beliefs (Borgatti et al., 2009; Kane et al., 2014)
Bots/aliases	Issue magnifiers	Instrumental—magnify support toward adoption threshold within clusters Instrumental—diffuse information through bridging

Table 3. Supranode sample actions and impacts on network structure and outcomes

Social media network features	Drivers	Supranode actions	Network structural impact	Outcome impact
Digital profiles; homogenized network connections	No cost Ease of use No managerial controls Allow group formation	Bots, aliases, collective action	More nodes Larger personal networks Supranode formation	Cascade formation
Network Structural transparency	Targeted advertising Support power users	Target nodes	Agency driven node networking	Target node interaction Cascade formation
User creation of digital content	Targeted advertising Find nodes	Target nodes Target messages	Agency driven information flows Larger information flows	False information flows Cascade formation

Table 4. Supranode archetypes, exemplars, drivers, and pressures

Archetypes	Exemplar	Drivers^1	Pressures^2
Private actors	Political campaign social media manipulation efforts	Power, Money	Mimetic
Government	Chinese government social media controls; Russian Internet Research Agency	Power	Coercive; mimetic
Platform	Facebook; Twitter	Money	Normative
Consultants; marketing agencies	Cambridge Analytica	Money	Mimetic
Standards bodies	IEEE	Glory	Normative

Malone et al. (2010)

Suchman (1995)

Borgatti, S. P., Mehra, A., Brass, D. J., & Labianca, G. (2009). Network Analysis in the Social Sciences. Science, 323(5916), 892–895. https://doi.org/10.1126/science.1165821

 PubMed Web of Science ®Google Scholar

Kane, G. C., Alavi, M., LaBlanca, G., & Borgatti, S. P. (2014). What’s different about social media networks? A framework and research agenda. MIS Quarterly, 38(1), 275–304. https://doi.org/10.25300/MISQ/2014/38.1.13

 Web of Science ®Google Scholar

Malone, T., Laubacher, R., & Dellarocas, C. (2010). The collective intelligence genome. MIT Sloan Management Review, 51(3), 21–31. sloanreview.mit.edu/article/the-collective-intelligence-genome/

 Web of Science ®Google Scholar

Suchman, M. (1995). Managing legitimacy: Strategic and institutional approaches. Academy of Management Review, 20(3), 571–610. https://doi.org/10.2307/258788

 Web of Science ®Google Scholar