Skip to main content


Evolution of vertical social learning

Research Achievements

Evolution of vertical social learning

Trainees R. Kyle Bocinsky, Bryan Roosien, and Daryl Trumbo developed an agent-based simulation extending a model of social learning model developed by McElreath and Strimling (2008). The Bocinsky et al. model allows study of the evolution of vertical (learning from parents) social learning among humans and other hominids. Their research was stimulated by two IPEM seminars, one by Richard McElreath, the other by Pontus Strimling. Unlike McElreath and Strimling's analytic model, Bocinsky et al. allowed individuals to learn and disperse on a landscape, and restricted the radius around which an agent may learn. They found that vertical learning performed best in social environments with low dispersion and low learning radii, and in stable physical environments. This research is reported in a poster entered in the trainee poster competition, Spring 2011.