Software Engineer
February 1, 2016
As part of the Human Nature Lab, we created a simulated model of the spread of violent behavior on social networks. Working off of real data gathered about raiding parties in the Hadza tribes in Africa, we designed a coordination-like game between leaders and ordinary participants to test their propensity to join violent raids under different social conditions.
We compared the effects of population-level interventions, such as limiting the extent of leaders' influence across network ties or lowering the propensity for risk-taking across the entire population, to the effects of lowering the violent tendencies of just a few key individuals. We found that, depending upon the properties of the chosen individuals within the network structure, reducing those individuals' propensity to engage in risky violent behavior can have the same net effect as reducing violence population-wide.
Read the full article here