Workshop Series: Andrew Stephen

20 August, 2010 - 14:00 to 15:30
Room 430 Joyce Ackroyd Building, St Lucia Campus

This paper examines how observable and measurable characteristics of the people who originally transmit information in online social networks affect how far that information spreads. Two characteristics are compared: a transmitter's connectivity (how well connected they are in the network) and activity (how frequently they transmit information over their social ties in the network). Despite extensive past research on connectivity (e.g., the literature on hubs), the role played by activity in driving diffusion is largely unexplored. Across three studies (an experiment, a simulation, and an empirical analysis of link sharing in Twitter) the authors find that

  1. a person's transmission activity positively influences diffusion,
  2. people who are exceptionally frequent content transmitters-pumps-have a large positive effect on information diffusion,
  3. when comparing the activity effect (cf. pumps) to that of connectivity (cf. hubs), activity is at least as strong a driver as connectivity, if not more under a variety of realistic conditions, and
  4. the transmitter activity effect on diffusion holds even after controlling for the information's quality and breadth of appeal.
Andrew Stephen, INSEAD, France